Nephrotoxicity associated with exposure to chemicals, principles and methods for the assessment of (EHC 119, 1991) IPCS INCHEM Home


    INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY



    ENVIRONMENTAL HEALTH CRITERIA 119




    PRINCIPLES AMD METHODS FOR THE ASSESSMENT OF NEPHROTOXICITY
    ASSOCIATED WITH EXPOSURE TO CHEMICALS




    This report contains the collective views of an international group of
    experts and does not necessarily represent the decisions or the stated
    policy of the United Nations Environment Programme, the International
    Labour Organisation, or the World Health Organization.

    Neither the Commission of the European Communities nor any person
    acting on behalf of the Commission is responsible for the use which
    might be made of the information contained in this report.

    (EUR 13222 EN)

    Published under the joint sponsorship of
    the United Nations Environment Programme,
    the International Labour Organisation,
    and the World Health Organization, and on
    behalf of the Commission of the
    European Communities





    World Health Orgnization
    Geneva, 1991

          The International Programme on Chemical Safety (IPCS) is a joint
    venture of the United Nations Environment Programme, the International
    Labour Organisation, and the World Health Organization.  The main
    objective of the IPCS is to carry out and disseminate evaluations of
    the effects of chemicals on human health and the quality of the
    environment.  Supporting activities include the development of
    epidemiological, experimental laboratory, and risk-assessment methods
    that could produce internationally comparable results, and the
    development of manpower in the field of toxicology.  Other activities
    carried out by the IPCS include the development of know-how for coping
    with chemical accidents, coordination of laboratory testing and
    epidemiological studies, and promotion of research on the mechanisms
    of the biological action of chemicals.

    WHO Library Cataloguing in Publication Data

    Principles and methods for the assessment of nephrotoxicity
    associated with exposure to chemicals.

          (Environmental health criteria: 119) (EUR ; 13222)

          1. Kidney diseases - chemically induced
          2. Kidney neoplasms - chemically induced
          3. Kidney - drug effects   I. Series   II. Series  EUR; 13222

          ISBN 92 4 157119 5         (NLM Classification WJ 300)
          ISSN 0250-863X

          (c) World Health Organization 1991
          (c) ECSC-EEC-EAEC, Brussels-Luxembourg, 1991

          Publications of the World Health Organization enjoy copyright
    protection in accordance with the provisions of Protocol 2 of the
    Universal Copyright Convention.  For rights of reproduction or
    translation of WHO publications, in part or  in toto,  application
    should be made to the Office of Publications, World Health
    Organization, Geneva, Switzerland.  The World Health Organization
    welcomes such applications.

          The designations employed and the presentation of the material in
    this publication do not imply the impression of any opinion whatsoever
    on the part of the Secretariat of the World Health Organization
    concerning the legal status of every country, territory, city, or area
    or of its authorities, or concerning the delimitation of its frontiers
    or boundaries.

          The mention of specific companies or of certain manufacturers'
    products does not imply that they are endorsed or recommended by the
    World Health Organization in preference to others of a similar nature
    that are not mentioned.  Errors and omissions excepted, the names of
    proprietary products are distinguished by initial capital letters.

    CONTENTS

         PRINCIPLES AND METHODS FOR THE ASSESSMENT OF NEPHROTOXICITY
         ASSOCIATED WITH EXPOSURE TO CHEMICALS

     1. SCOPE OF THE HEALTH SIGNIFICANCE OF NEPHROTOXICITY

     2. NEPHROTOXICITY

         2.1. Target selectivity
         2.2. The dynamics of renal injury
         2.3. Classification of renal disease
         2.4. The epidemiology of nephrotoxicity
         2.5. Risk factors for toxic nephropathies
               2.5.1. Factors related to renal function
               2.5.2. Clinical risk factors
               2.5.3. Extrapolation of animal data to man
               2.5.4. Risk assessment from nephrotoxicity
                       studies in animals
               2.5.5. Special risk groups in humans
               2.5.6. Multichemical exposure
               2.5.7. Renal functional reserve
               2.5.8. The effects of chemicals on kidneys
                       with pre-existing renal lesions
                       2.5.8.1   Nephrotoxicity in the presence
                                 of renal and extrarenal disease

     3. KIDNEY STRUCTURE AND FUNCTION

         3.1. Renal anatomy
               3.1.1. Histology
               3.1.2. Enzyme histochemistry and quantification
               3.1.3. Immunohistochemistry
         3.2. The renal blood supply
               3.2.1. Renal haemodynamics
         3.3. The nephron
               3.3.1. Cellular heterogeneity and cell-cell
                       interaction
               3.3.2. The glomerulus
               3.3.3. The proximal tubule
               3.3.4. The medulla 58
                       3.3.4.1   The loops of Henle
                       3.3.4.2   Collecting ducts
                       3.3.4.3   The distal tubule
                       3.3.4.4   The countercurrent multiplier
                                 system and urine concentration
                       3.3.4.5   The interstitial cells
         3.4. Species, strain, and sex differences in renal
               structure and function
         3.5. Renal biochemistry

               3.5.1. Biochemistry and metabolism in the cortex
               3.5.2. Biochemistry and metabolism in the medulla
                       3.5.2.1   The biochemistry of renal
                                 prostaglandins (PG)
                       3.5.2.2   Lipid metabolism
                       3.5.2.3   Carbohydrate metabolism in the
                                 medulla
                       3.5.2.4   Medullary glycosaminoglycan (GAG)
         3.6. The metabolism of xenobiotic molecules in the kidney
               3.6.1. Oxidases
                       3.6.1.1   Cytochrome P-450-dependent mixed-
                                 function oxidases (monooxygenases)
                       3.6.1.2   Prostaglandin peroxidase-
                                 mediated metabolic activation
               3.6.2. Conjugation
                       3.6.2.1   Glucuronide conjugation
                       3.6.2.2   Sulfate conjugation
                       3.6.2.3   Glutathione conjugation
                       3.6.2.4   Mercapturic acid synthesis
                       3.6.2.5   Amino acid conjugation
               3.6.3. Other enzymes involved in xenobiotic
                       metabolism

     4. THE MECHANISTIC BASIS OF CHEMICALLY INDUCED RENAL INJURY

         4.1. Immunologically induced glomerular disease
         4.2. Direct glomerular toxicity
         4.3. Tubulointerstitial disease
               4.3.1. Acute interstitial nephritis
               4.3.2. Acute tubular toxicity
               4.3.3. Chronic interstitial nephritis
         4.4. Mechanisms of cellular toxicity
         4.5. Factors that modify cellular injury by toxins
               4.5.1. Cellular transport and accumulation
               4.5.2. Metabolic degradation
               4.5.3. Intracellular protein binding
               4.5.4. Membrane reactions and pinocytosis

     5. THERAPEUTIC AGENTS AND CHEMICALS THAT HAVE THE POTENTIAL TO
         CAUSE NEPHROTOXICITY

         5.1. Therapeutic agents
               5.1.1. Analgesics and non-steroidal
                       anti-inflammatory drugs (NSAIDs)
               5.1.2. Paracetamol and  para-aminophenol
               5.1.3. Antibiotics
                       5.1.3.1   Aminoglycosides
                       5.1.3.2   Cephalosporins
                       5.1.3.3   Amphotericin B
                       5.1.3.4   Tetracyclines

               5.1.4. Penicillamine
               5.1.5. Lithium
               5.1.6. Urographic contrast media (UCM)
               5.1.7. Anticancer drugs
                       5.1.7.1   Cisplatin
                       5.1.7.2   Adriamycin
               5.1.8. Immunosuppressive agents
                       5.1.8.1   Cyclosporin A
               5.1.9. Heroin
               5.1.10. Puromycin aminonucleoside
         5.2. Chemicals
               5.2.1. Ethylene glycol
               5.2.2. Organic chemicals and solvents
                       5.2.2.1   Volatile hydrocarbons
                       5.2.2.2   Chloroform
                       5.2.2.3   Halogenated alkenes
                       5.2.2.4   Hydrocarbon-induced
                                 nephrotoxicity
                       5.2.2.5   Bipyridyl herbicides
         5.3. Mycotoxins
         5.4. Silicon
         5.5. Metals
               5.5.1. Lead
               5.5.2. Cadmium
               5.5.3. Mercury
               5.5.4. Gold
               5.5.5. Bismuth
               5.5.6. Uranium
               5.5.7. Chromium
               5.5.8. Arsenic
               5.5.9. Germanium

     6. RENAL CANCER

         6.1. Renal tumour classification
         6.2. Renal adenocarcinoma
         6.3. Upper urothelial carcinoma (transitional
               cell carcinoma)
         6.4. Experimentally induced renal adenomas and
               adenocarcinomas
               6.4.1. Background incidence of spontaneous
                       tumours in experimental animals
               6.4.2. Inorganic compounds
               6.4.3. Organic molecules
                       6.4.3.1   Nitrosamines and related
                                 compounds
                       6.4.3.2   Morphological changes
                       6.4.3.3   Biochemical changes in cells
                       6.4.3.4   The mechanistic basis of renal
                                 carcinoma
         6.5. Experimentally induced upper urothelial
               carcinomas (transitional cell carcinomas)

     7. ASSESSMENT OF NEPHROTOXICITY

         7.1.  In vitro studies
               7.1.1. Choice of chemical concentrations
                       for  in vitro studies
                       7.1.1.1   Proximate and ultimate
                                 nephrotoxicants  in vitro
               7.1.2.  In vitro investigations of
                       nephrotoxicity
                       7.1.2.1   Perfusion and micropuncture
                       7.1.2.2   Renal cortical slice
                       7.1.2.3   Isolated nephron segments
                       7.1.2.4   Primary cell cultures
                       7.1.2.5   Established renal cell lines
                       7.1.2.6   Subcellular fractions
         7.2.  In vivo experimental studies
               7.2.1. Methods for assessing chemically reactive
                       nephrotoxic metabolites in animals
               7.2.2. Evaluation of glomerular function
               7.2.3. Evaluation of tubular functions
               7.2.4. Proteinuria
                       7.2.4.1   Total proteinuria and
                                 electrophoretic pattern
                       7.2.4.2   Urinary excretion of single
                                 plasma proteins
                       7.2.4.3   Enzymuria
                       7.2.4.4   Immunoreactive tissue
                                 constituents
                       7.2.4.5   Urinary excretion of
                                 prostaglandins
               7.2.5. Clinical context
               7.2.6. Radiological techniques
               7.2.7. Other non-invasive renal assessment

     8. DETECTION OF NEPHROTOXICITY IN HUMANS

         8.1. Markers of nephrotoxicity
               8.1.1. General requirements
               8.1.2. Diagnostic value
               8.1.3. Prognostic value
         8.2. Screening for nephrotoxicity in humans
               8.2.1. Glomerular filtration
               8.2.2. Tests designed to assess selective
                       dysfunction
               8.2.3. Tests designed to assess tissue damage
                       8.2.3.1   Enzymuria
                       8.2.3.2   Immunoreactive tissue
                                 constituents
         8.3. Clinical investigations
               8.3.1. Invasive techniques
                       8.3.1.1   Biopsies from humans
                       8.3.1.2   Autopsy in humans

               8.3.2. Tests designed to assess glomerular
                       filtration and renal blood flow
               8.3.3. Proteinuria
               8.3.4. Tests designed to assess selective
                       damage

     9. SUMMARY AND CONCLUSIONS

    10. RECOMMENDATIONS

    REFERENCES

    RESUME ET CONCLUSIONS

    RECOMMANDATIONS

    RESUMEN Y CONCLUSIONES

    RECOMENDACIONES
    

    WHO/CEC TASK GROUP ON PRINCIPLES AND METHODS FOR THE ASSESSMENT OF
    NEPHROTOXICITY ASSOCIATED WITH EXPOSURE TO CHEMICALS

     Members

    Professor E.A. Bababunmi, Biomembrane Research Laboratories,
       Department of Biochemistry, University of Ibadan, Ibadan, Nigeria
        (Vice-Chairman)

    Dr P. Bach, Nephrotoxicity Research Group, Robens Institute of Health
       and Safety, University of Surrey, Guildford, Surrey, United Kingdom

    Professor G. Baverel, Department of Pharmacology, Alexis Carrel
       Faculty of Medicine, Lyon, France

    Professor W.O. Berndt, University of Nebraska Medical Centre, Omaha,
       Nebraska, USA  (Chairman)

    Dr G. Duggin, Toxicology Unit, Royal Prince Alfred Hospital,
       Camperdown, New South Wales, Australia

    Dr H. Endou, Department of Pharmacology, Faculty of Medicine,
       University of Tokyo, Bunkyo-ku, Tokyo, Japan

    Professor R. Goyer, Department of Pharmacology, University of Western
       Ontario, Health Science Centre, London, Ontario, Canada

    Dr M. Robbins, Tissue Radiobiology Research Unit, Churchill Hospital,
       Headington, Oxford, United Kingdom  (Rapporteur)

     Observer

    Dr C. Cojocel, European Chemical Industries Ecology and Toxicology
       Centre, Brussels, Belgium

     Secretariat

    Dr J.C. Berger, Health and Safety Directorate, Commission of the
       European Communities (CEC), Luxembourg

    Dr E. Smith, International Programme on Chemical Safety, Division of
       Environmental Health, World Health Organization, Geneva,
       Switzerland

     Consultants representing the CEC

    Dr A. Bernard, Unit of Industrial Toxicology and Occupational
       Medicine, Catholic University of Louvain, Brussels, Belgiuma

     Secretariat  (contd.)

    Dr P. Druet, National Institute of Health and Medical Research
       (INSERM), Broussais Hospital, Paris, Francea

    Professor A. Mutti, Institute of Clinical Medicine and Nephrology,
       University of Parma, Parma, Italya

                 
    a  Attended 6 December 1989 only.

    NOTE TO READERS OF THE CRITERIA DOCUMENTS

         Every effort has been made to present information in the criteria
    documents as accurately as possible without unduly delaying their
    publication. In the interest of all users of the environmental health
    criteria documents, readers are kindly requested to communicate any
    errors that may have occurred to the Manager of the International
    Programme on Chemical Safety, World Health Organization, Geneva,
    Switzerland, in order that they may be included in corrigenda, which
    will appear in subsequent volumes.

                                  *   *   *

         A detailed data profile and a legal file can be obtained from the
    International Register of Potentially Toxic Chemicals, Palais des
    Nations, 1211 Geneva 10, Switzerland (Telephone No. 7988400 or
    7985850).

    PREFACE

         The preparation of this monograph was undertaken jointly by the
    International Programme on Chemical Safety (UNEP/ILO/WHO) and the
    Commission of the European Communities.

         A joint WHO/CEC Task Group on Principles and Methods for the
    Assessment of Nephrotoxicity Associated with Exposure to Chemicals met
    at the National Institute of Public Health and Environmental
    Protection, Bilthoven, the Netherlands, from 4 to 8 December 1989. The
    meeting was opened by Dr K.A. van der Heijden on behalf of the
    Netherlands and the host institute. The Secretariat responded and
    welcomed the participants on behalf of the three cooperating
    organizations of the IPCS (UNEP/ILO/WHO) and the Commission of the
    European Communities. The Task Group reviewed and revised the draft
    criteria document and prepared a final text.

         The drafts of this Monograph were prepared by DR P. BACH,
    Guildford, United Kingdom, PROFESSOR W.O. BERNDT, Omaha, USA and
    PROFESSOR R. GOYER, London, Ontario, Canada. During the preparation of
    the monograph, many scientists made constructive suggestions and their
    contributions are gratefully acknowledged. The photographs in Figure
    2 were supplied by Dr N.J. Gregg, in Figure 4 by Professor W. Guder,
    and in Figure 14 by the Department of Toxicology, Institute for
    Medical Research and Occupational Health, University of Zagreb.
    Following the Task Group, Dr P. Bach and Dr M. Robbins collated the
    text for IPCS. Dr E. Smith and Dr P.G. Jenkins, both of the Central
    Unit, IPCS, were responsible for the overall scientific content and
    technical editing, respectively.

    ABBREVIATIONS

    ADH       anti-diuretic hormone

    AIN       acute interstitial nephritis

    ARF       acute renal failure

    BEA       2-bromoethalamine

    BEN       Balkan endemic nephropathy

    BUN       blood urea nitrogen

    CIN       chronic interstitial nephritis

    DBCP      1,2-dibromo-3-chloropropane

    DCVC       S-(1,2-dichlorovinyl)-L-cysteine

    DTPA      diethylenetriamine pentaacetic acid

    EDTA      ethylenediaminetetraacetic acid

    ESRD      end-stage renal disease

    GAG       glycosaminoglycan

    GBM       glomerular basement membrane

    GFR       glomerular filtration rate

    GSH       glutathione-SH

    H & E     haematoxolin and eosin

    HCBD      hexachloro-1,3-butadiene

    HPLC      high-performance liquid chromatography

    LDH       lactate dehydrogenase

    3MC       3-methylcholanthrene

    NAC        N-acetyl cysteine

    NADPH     reduced nicotinamide adenine dinucleotide phosphate

    NNM        N-nitrosomorpholine

    NSAID     non-steroidal anti-inflammatory drugs

    PAH        p-aminohippurate

    PAS       periodic acid Schiff stain

    PCBD       S-(1,2,3,4,4-pentachloro-1,3-butadienyl

    PG        prostaglandin

    PoG       proteoglycan

    RPN       renal papillary necrosis

    TEA       tetraethyl ammonium

    UCM       urographic contrast medium

    UDP       uridine diphosphate

    1.  SCOPE OF THE HEALTH SIGNIFICANCE OF NEPHROTOXICITY

         Over the last 20 years it has become increasingly obvious that
    the kidney is adversely affected by an array of chemicals. Man is
    exposed to these as medicines, industrial and environmental chemicals,
    and a variety of naturally occurring substances. The level of exposure
    varies from minute quantities to very high doses. Exposure may be over
    a long period of time or limited to a single event, and it may be due
    to a single substance or to multiple chemicals. The circumstances of
    exposure may be inadvertent, accidental, or intentional overdose or
    therapeutic necessity. Some chemicals cause an acute injury and others
    produce chronic renal changes that may lead to end-stage renal failure
    and renal malignancies. The extent and cost of clinically relevant
    nephrotoxicity has only started to become apparent during the last
    decade. However, the full extent of the economic impact of chemically
    induced or associated nephropathy is difficult to define because the
    diagnosis of early injury and the documentation of the cascade of
    secondary degenerative changes have not been adequately identified.
    Instead most chemically associated renal disease is only identified as
    an acute renal failure or as chronic renal failure at a very late
    stage when therapeutic intervention is impossible. More importantly at
    this stage, the etiology may be obscured by lack of reliable
    information on the likely causative agents, the levels and duration of
    exposure, and other possible contributing and exacerbating factors. At
    present, epidemiological evidence indicates that nephrotoxicity
    leading to acute and/or chronic renal failure represents a substantial
    financial burden to society (Nuyts et al., 1989). Indeed, there is
    some indication that chemical exposure could play a much greater
    influence in the very high incidence of end-stage renal disease
    encountered in nephrology and dialysis clinics than is currently
    considered to be the case.

         There are already several examples of this type of chemically
    associated disease that went unrecognized for some time. These include
    those nephropathies caused by cadmium, other environmental heavy
    metals, and, more recently, the organo-metallic compounds used as
    therapeutic agents, anti-cancer drugs, cyclosporin, analgesic abuse,
    and antibiotics.

         Owing to its diverse functions and small mass in relation to the
    resting cardiac output that it handles, the kidney is a target both
    for chemicals that are pharmacologically active and for toxic
    material. The nephron and its related cells perform a diversity of
    physiological functions.  It is the major organ of excretion and
    homeostasis for water-soluble molecules; because it is a metabolically 
    active organ, it can concentrate certain substances actively. In
    addition, its cells have the potential to bioconvert chemicals and
    metabolically activate a variety of compounds. There are a number of
    other processes described below that establish the potential for
    cellular injury. Specific physiological characteristics are localized
    to specific cell types. This makes them susceptible to, and the target
    for, chemicals. The effect of any chemical on a cell may be

    pharmacological, in which case the effect is dose related and occurs
    only as long as the concentration of the effector is high enough to be
    active. Alternatively, the chemical may cause damage to the cell. The
    cell responds to injury by repair and the kidney responds to cellular
    lesion by renal and extrarenal adaptation to compensate for loss of
    that cell function. Although there is a substantial capacity within
    the kidney for repair, there are also several circumstances where
    damage may be irreversible. In general, the proximal and distal
    tubules and urothelia can be repaired, but the glomeruli and medulla
    may have a significantly lower repair facility. It is, therefore,
    possible to initiate a series of degenerative changes as a result of
    interfering with one or more of the normal physiological processes.

         The Environmental Health Criteria monographs normally focus on
    industrial chemicals, but at present most of the experimental and
    human information on nephrotoxicity is based on therapeutic
    substances. These data are most useful because there are animal and
    human comparisons for specific chemicals where the levels of exposure
    and the nephrotoxicological  consequences are well documented. From
    these data it has been possible to glean some understanding of
    mechanisms of primary injury and the long-term consequences and health
    significance.  Thus, these compounds are generally well studied, and
    the more rational understanding of the mechanism of their
    nephrotoxicity in animals and man provides the basis for validating
    extrapolation between species and making rational risk assessment.

         Most risk assessment decisions are currently based on information
    concerning the aminoglycosides, halogenated anaesthetics, several
    heavy metals, and lithium, where there is an excellent concordance
    between animal data and findings in humans exposed to these agents
    (Kluwe et al., 1984; Porter & Bennett, 1989). This has provided some
    predictive indication of what will take place in humans exposed to
    analogues of these compounds. On the other hand, the demonstration
    that the occurrence of light hydrocarbon-related adenocarcinomas is
    specific to male rats shows that there are examples where the
    molecular understanding of a renal lesion in animals is irrelevant to
    humans.

         There are also therapeutic agents where attempts to extrapolate
    from animals to man have not been as successful. These include
    compounds such as cyclosporin, analgesics and non-steroidal
    anti-inflammatory agents. It has, however, been possible to develop
    some model lesions that parallel those in humans using these
    compounds. Generally, different protocols have had to be used, such as
    water deprivation and renal injury, but these have in turn provided
    the basis for developing improved screening methods for such chemicals
    and also for probing the molecular nature of the lesion. There are,
    however, a number of chemicals, such as renal carcinogens, mycotoxins,
    other natural toxins, and anti-cancer drugs, and some types of lesion,
    such as the immunonephropathies, where it has been difficult to
    establish good models in animals. A host of chemicals alter glomerular
    filtration rate (GFR) or some other aspect of renal function, but the

    long-term health significance is still not known and it is uncertain
    how to extrapolate such data to man.

    2.  NEPHROTOXICITY

         Nephrotoxicity can be defined as renal disease or dysfunction
    that arises as a direct or indirect result of exposure to medicines,
    and industrial or environmental chemicals. It is well established that
    toxic nephropathies are not restricted to a single type of renal
    injury. Some chemicals target one discrete anatomical region of the
    kidney and may affect only one cell type. Chemical insult to the
    kidney may result in a spectrum of nephropathies that are
    indistinguishable from those that do not have a chemical etiology.

    2.1  Target selectivity

         It has become increasingly apparent that there are a number of
    chemicals that may adversely affect one or more of the anatomical
    elements of the kidney, such as the glomerulus, proximal,
    intermediate, and distal tubules, and medullary, endothelial, and
    urothelial cells. Although some of these cell types (such as the
    proximal tubular cells) have a marked ability to repair damaged
    regions, others, such as the glomerular epithelium and the "type 1"
    medullary interstitial cells, do not. It is for this reason that the
    dynamic process that follows any renal injury can affect the outcome
    of the chemical insult.

    2.2  The dynamics of renal injury

         The renal response to injury is dynamic, and the kidney adapts to
    maintain homeostasis during the cascade of repair and recovery that
    follows the primary insult (Bach, 1989). Depending on the type and
    frequency of the damage, and the region of the kidney that is damaged,
    the organ can respond by a recovery, a reduced functional reserve, or
    by a progressive degenerative change. A reduced functional reserve may
    play a very important role in sensitizing the kidney to subsequent
    renal injury, and an initiated degenerative cascade may either
    stabilize or progress to acute or chronic renal failure. It is not
    possible to differentiate between a kidney that has totally recovered,
    one with a reduced functional reserve, and an organ with early
    progressive degenerative change, except in animals where function and
    morphology can be assessed under well controlled conditions.

    2.3  Classification of renal disease

         Classification of renal disease can be based on clinical
    manifestations, pathological changes, or etiological agents.  WHO has
    prepared a number of detailed and illustrated publications in recent
    years on the classification of renal disease (WHO 1982, l985, 1987,
    1988). The general approach is to subdivide the kidney into major
    anatomical components (i.e. glomeruli, tubules and interstitium, and
    blood vessels) and to relate these to the major clinical syndromes

    characteristic of renal diseases. Table 1 contains a modified
    classification of renal disease that focuses on major disorders of the
    kidney that may be associated with nephrotoxins.  This classification
    is consistent with previous WHO publications and textbooks of
    nephrology and pathology and provides a framework for discussing the
    mechanisms and pathology of nephrotoxicity. It must be appreciated
    that nephrotoxic agents may have multiple anatomical targets and that
    toxicity manifests itself in more than one clinical syndrome.  Further
    discussion of renal effects due to specific agents are discussed
    below.

    2.4  The epidemiology of nephrotoxicity

         Putting the health significance of nephrotoxicity into
    perspective is difficult because of the diverse array of chemicals
    that target different parts of the kidney, the spectrum of disease
    consequences, and the many interacting factors. There is also
    uncertainty in assessing changes in renal function before they reach
    the point where preventive medicine can no longer be practised and
    therapeutic intervention may be appropriate.

         Many industrial and environmental chemicals and therapeutic
    agents have been shown in experimental studies and from acute toxic
    exposures to be nephrotoxins, but the extent of their contributions to
    the overall incidence of chronic renal failure is not known. Data
    extracted from the European Dialysis and Transplant Association
    Registry identified only about 4% of patients starting renal
    replacement therapy in 1984 as having drug or chemical associated
    renal disease. However, nearly 50% of these patients were considered
    possible (but not diagnosed) cases of toxic nephropathy (Dieperink,
    1989). Of those patients identified as having chemical-related renal
    disease, analgesic nephropathy is the most important recognized
    outcome. In an analysis of the European Dialysis and Transplant
    Association Registry (1986), the prevalence of analgesic nephropathy
    was found to vary greatly between countries.  It is highest in
    Switzerland (18.1%) and Belgium (11.8%) and accounts for over 4% of
    patients in Denmark, Germany, Czechoslovakia, and Austria.  In 20
    countries the prevalence is lower than one patient in every hundred
    (European Dialysis and Transport Association Registry, 1986; Wing et
    al., 1989). Other specific drug nephropathies recorded less frequently
    include those due to cisplatin ( cis-platinum) and cyclosporin A. A
    small number of patients had other specific drug or chemical-related
    nephropathies.

         The role of the toxic agents that may contribute to the 50% of
    cases of chronic renal failure of undiagnosed etiology is less
    certain. There is opportunity for exposure to a number of chemicals in
    the workplace or ambient environment (drugs included) that are
    possible nephrotoxins. It has been estimated that there were nearly
    four million workers in the USA with potential occupational exposure
    to known or suspected nephrotoxins in the 1970s (Landrigan et al.,
    1984). The major occupational exposure is to workplace solvents, but

          Table 1.  Classification of renal disease due to nephrotoxins in humans
                                                                               
    1. Immunologically mediated

           Antibody mediated

              Membranous glomerulonephritis or immune complex type disease

              metals (gold, mercury)
              D-penicillamine
              drugs responsible for a lupus-like syndrome
              (hydralazine, procainamide, diphenylhydantoin)

           Anti-glomerular basement membrane antibody mediated

              organic solvents
              hydrocarbons

    2. T cell mediated (?)

           Nephrotic syndrome with minimal glomerular changes

              lithium salts
              non-steroidal anti-inflammatory agents
                                                                               
    
    toxic metals and organic compounds, including pesticides, are of great
    concern. The well documented occurrence of subclinical nephropathies
    in subjects occupationally exposed to nephrotoxins such as lead or
    cadmium (see section 5.5), the excess of mortality for renal diseases
    in cohorts of workers with previous exposure to these two heavy metals
    (Bennett, 1985; Bernard & Lauwerys, 1986), and, more recently, the
    suggestion that subclinical renal effects caused by cadmium are early
    signs of an accelerated and irreversible decline of renal function
    (Roels et al., 1989) should all be noted.  The linkage of these risks
    to the actual occurrence of chronic renal failure has not, however,
    been possible. A review of the end-stage renal disease (ESRD)
    population in the USA has shown that at least 19% have a renal disease
    of unknown or non-specific etiology (Burton & Hirschman, 1979). If
    other diagnostic groups that have uncertain etiologies, such as the
    30% with glomerulo-nephritis, 5% with interstitial nephritis of
    suspected etiology (lead, analgesics, etc.) and possibly the 8%
    diagnosed as having pyelonephritis, are added, it becomes apparent
    that the etiology of a large portion of patients with chronic renal
    failure is unrecognized or undefined. Environmental factors may have
    a previously unrecognized role in the etiology of these lesions.

         There are many reasons for the failure to recognize toxic
    etiologies (Sandler, 1987). A major reason is that chronic renal
    failure develops slowly over a number of years, so that retrospective

    identification of a toxic agent requires knowledge of lifestyles,
    therapies, or workplace environments that might provide risk factors.
    However, such data are generally not available. Unless the drug or
    toxic chemical is persistent in tissues, it is usually not possible to
    confirm or quantify exposure. The inability to recognize multiple
    etiologies or confounding factors adds further complexity to the
    problem. There is also lack of uniformity in clinical and pathological
    diagnoses. A further complexity is that there is a tendency to
    categorize chronic renal failure by a mixture of pathological and
    etiological classifications.  For example, in one instance a patient
    may be classified as having chronic interstitial nephropathy on the
    basis of a renal biopsy, whereas another patient with the same
    pathology might be classified as having toxic nephropathy due to
    exposure to a nephrotoxic chemical because of knowledge of exposure.
    In a survey of patients requiring dialysis in Israel, Modan et al.
    (1975) found diagnostic inconsistencies between hospital diagnosis,
    autopsy reports, and diagnosis made by the study reviewers.
    Disagreement was most often seen for chronic glomerulo-nephritis,
    chronic pyelonephritis and nephrosclerosis. The stage of the
    pathological process or severity influences classification.
    Interstitial nephritis tends to be diagnosed more frequently in the
    early stages of chronic renal failure, whereas glomerulonephritis is
    a more common diagnosis for patients undergoing dialysis.  There may
    be a rational basis for this in that the scarring in persistent
    interstitial nephritis does impede blood supply to the glomerulus.
    This could lead to glomerular disease and interstitial nephritis
    despite the fact that there are different etiologies or risk factors
    for the two conditions.

         The identification of chronic pyelonephritis is made more precise
    by following established criteria. These include the presence of gross
    irregular scarring, inflammation, fibrosis and deformity of calyces
    underlying parenchymal scars, predominant tubulointerstitial
    histological damage, and relative lack of glomeruli. There is evidence
    that some of the chronic interstitial nephritis that is labelled
    chronic pyelonephritis is due to something other than bacterial
    infection.

         Environmental Health Criteria 27: Guidelines on Studies in
    Environmental Epidemiology (WHO, 1983) provides guidelines for
    obtaining human data concerning the health effects of exposure to
    chemical agents. For agents that produce acute renal failure,
    long-term follow-up may identify those instances where chronic renal
    disease has persisted.  For agents that give rise to accidental
    poisoning, clinical case reports can provide important information. In
    the case of agents where exposure to larger population segments
    occurs, information may be obtained by using statistical and
    epidemiological methods to investigate possible nephrotoxicity from
    such exposures (as compared to a non-exposed control group).  Specific
    segments of the population that might be at higher risk to a potential
    nephrotoxic drug or workplace chemical should be particularly closely
    monitored for renal effects.

         There are marked differences between the incidence of
    analgesic-associated ESRD in different countries and within the same
    country (Table 2). This varies from up to 22% in Australia (in 1982)

               Table 2.  National prevalence of analgesic nephropathy
                  in patients with end-stage renal failure
                                                                        
                               %                                     %
                                                                        
    South Africa               22      Scandinavia (1979)            3
    Switzerland (1980)         20      France (1979)                 2
    Belgium (1984)             18      USA                           2
    Australia (1985)           15      United Kingdom (1979)         1
    Federal Republic of        13      Italy (1979)                  1
    Germany (1983)                     Spain (1979)                  0.4
    Canada (1976)               3
                                                                        
    
    and in parts of some of the European countries to as low as 0.2% in
    the USA. It is generally considered that the withdrawal of phenacetin
    has lead to the disappearance of the high incidence of renal papillary
    necrosis (RPN) in Scandinavia, Canada, and Australia, but a high
    incidence remains in Switzerland, Belgium, and the Federal Republic of
    Germany (Gregg et al., 1989). Specific geographical locations may have
    analgesic abuse problems such as the Winston-Salem area (USA).
    Worldwide variability in the prevalence of analgesic nephropathy has
    long been recognized. The correlation of the incidence of this disease
    with local analgesic consumption has been demonstrated. However, the
    relation between both phenomena is not well established since
    comparable consumption data, focussed on the sales of analgesic
    mixtures, are not available in most countries.  The high frequency
    abuse area in Belgium is situated in the north (Fig. 1a), where up to
    51% of dialysis patients are analgesic abusers, but this is markedly
    lower in the south (Elseviers & De Broe, 1988).  In Germany (Fig. 1b)
    the highest prevalence is in West Berlin (up to 50%), Hamburg, and
    Bremen (Pommer et al., 1986). These data indicate that the prevalence
    of this disease has been underestimated on a national basis.  There
    are indications that the overall prevalence of this disease has also
    been underestimated in several other countries. Local well-conducted
    studies of the prevalence of analgesic nephropathy showed higher
    prevalences than the European Dialysis and Transport Association
    Registry. In the Federal Republic of Germany, a prevalence of 13% of
    analgesic nephropathy in dialysis patients was found, while the
    appropriate European Dialysis and Transport Association data was never
    more than 6% (Pommer et al., 1986). In Belgium, the prevalence of
    analgesic nephropathy was 18%, whereas the European Dialysis and
    Transport Association registered a prevalence of 12% (Elseviers & De
    Broe, 1988). The percentage of nephropathies of unknown etiology and
    of pyelointerstitial nephritis may also indicate an underestimation of
    analgesic nephropathy. These percentages are low in countries with a

    high prevalence of analgesic nephropathy (Switzerland and Belgium) and
    are high in countries such as Italy and Spain (Wing et al., 1989)
    where analgesic nephropathy is considered to be rare. Analgesic
    nephropathy progresses silently over a long period, and so the
    diagnosis is difficult. In addition, most patients deny being
    analgesic abusers, which further confounds diagnosis. Symptoms are
    nonspecific until the degenerative cascade affects the cortex, when
    renal failure occurs. Moreover, even when the renal failure is
    recognized, the diagnosis of analgesic nephropathy remains difficult 
    unless diagnostic criteria are established.

    2.5  Risk factors for toxic nephropathies

         The risk of developing a clinically significant nephrotoxicity
    depends on pre-existing clinical conditions and may be identified in
    specific patient populations. Hypertension, diabetes, cardiovascular
    disease, etc. are all thought to have the potential to exacerbate
    nephrotoxicity, but many of these conditions have not been
    systematically investigated for all types of nephrotoxicity. There are
    examples where both chemicals and other disease factors cause a
    lesion. For example, sickle cell disease and diabetes can cause renal
    papillary necrosis, a condition that is also common in individuals who
    abuse analgesics.  The risk factors that predispose individuals to
    renal papillary necrosis are not clear, and it is also unclear whether
    diabetics are at greater risk of developing the lesion if they take
    high doses of analgesics. This question  cannot be resolved until
    better diagnostic criteria are developed to identify the lesion before
    it involves the cortex. Risk may also vary for different nephrotoxins.
    While Bence-Jones protein excretion considerably increases the risk of
    radiocontrast-induced renal injury, the effects of other types of
    chemicals in patients with multiple myeloma are not clear. Therefore
    recognition of the risk factors is necessary for the understanding and
    prevention of renal damage.

         There are also many examples of animal data that have not yet
    been translated into risk terms in humans.  For example, the immature
    kidney may be resistant to amino-glycosides (Marre et al., 1980) and
    cephalosporins (Tune, 1975), but the reverse is true for other
    chemicals such as hexachlorobutadiene (Hook et al., 1983).  Other
    factors, such as electrolyte and volume changes or an alteration in
    the renin-angiotensin system, may affect some types of drug-induced
    acute renal failure (Bennett et al., 1983).

         Risk factors as a measure of vulnerability to potential
    nephrotoxicity that could be caused by drugs and chemicals are not
    well defined at present. Clearly, wide variation exists among
    individuals and even groups of people. Multiple exposure to toxic
    agents and multiple drug usage are certainly factors, but the ability
    to estimate risks to multiple exposure is limited. Factors intrinsic
    to the nature of renal function and risk factors presented by clinical
    disease have been reviewed (Porter, 1989) and are discussed below.

    FIGURE 1a

    FIGURE 1b

    2.5.1  Factors related to renal function

         The vulnerability of the kidney to toxicity from exposure to a
    particular drug or chemical is the product of several groups of risk
    factors. In any one person, multiple factors may be operative. The
    nature of normal renal function in itself contributes to the
    vulnerability to toxins. The intimate association of the capillary
    endothelial surface during the process of ultrafiltration provides
    opportunity for direct toxicity.  A further contribution to the
    glomerular capillary vulnerability is the positive hydrostatic
    pressure required for producing the plasma ultrafiltrate. Adding to
    this vulnerability is the "hyperfiltration injury" hypothesis proposed
    by Brenner (1983) and Brenner et al. (1978, 1982), who showed that
    when a nephron ceases to function, the remaining nephrons hypertrophy
    and the flow rate per functioning nephron is raised, thus increasing
    the exposure to the drug or chemical. A further aspect is the concept
    of renal reserve, i.e. fewer functioning nephrons further increase the
    vulnerability of the kidney to toxicity.

         Another aspect of the structure of the glomerular endothelial
    cells that can lead to injury to the kidney is the negative charge of
    the filtration membranes.  Positively  charged ligands can  become
    electrostatically attached and alter the permeability coefficient of
    the glomerulus. In addition, cationic proteins can be sequestered in
    the glomerulus and act as "planted antigens", and a circulating
    antibody can attach to such antigens resulting in an in situ immune
    complex formation; hydro-carbons from petroleum products can act in a
    similar way (Ravnskov, 1985). This is but one of a wide variety of
    immunologically mediated glomerular injury patterns that have been
    identified. This variety is not surprising when one considers the
    heterogeneity of the biochemical composition of the glomerulus and the
    wide spectrum of antigenic compounds to which the body is exposed
    (Glassock, 1986).

         Tubular vulnerability to nephrotoxins is related to the nature of
    normal tubular function.  The medullary countercurrent multiplier
    system provides a mechanism for eliminating body waste products while
    minimizing body water loss. A consequence is the reabsorption and
    recycling of compounds of low relative molecular mass, in particular
    urea and neutral toxicants and/or their metabolites, which can
    accumulate in the medullary interstitium. Depending on their chemical
    properties, they may initiate an inflammatory response through
    activation of mediators, a factor that may be relevant in the
    pathogenesis of analgesic nephropathy (Mudge, 1982).

         The organic acid and base transport systems of the proximal 
    tubule serve to excrete  certain molecular species.  Several commonly
    used drugs, including the organic acid penicillin, utilize this
    mechanism of transport.  Toxicants that are involved in these systems
    might induce renal injury directly because of high cellular
    concentration or by acting as competitive inhibitors to block the
    elimination of endogenously produced toxic metabolites.  Tubular

    mechanisms for acidification may play a part in tubular injury by
    drugs or chemicals that induce an acidification defect, e.g., lithium
    (Batelle et al., 1982).  Drugs or chemicals that are absorbed by
    pinocytosis become concentrated in lysosomes where they are subjected
    to digestion by hydrolytic enzymes.  Some toxicants, however, may
    inhibit the hydrolytic process, resulting in drug accumulation and
    tubular cell toxicity that may resemble lysosomal storage disease (as
    occurs in aminoglycoside nephrotoxicity).

    2.5.2  Clinical risk factors

         The application of multivariate analysis for investigating
    clinical risk factors in the onset of acute renal failure (Rasmussen
    & Ibels, 1982) may provide some insight into factors that may increase
    vulnerability to nephrotoxicity from drugs and chemicals. The risk
    factors summarized in Table 3 show that multiple risk factors coexist
    in the majority of patients with acute renal failure. Although age has
    been recognized as a factor in a number of studies (Porter, 1989;
    Porter & Bennett, 1989), it may simply be a convenient marker for the
    change in renal vulnerability that relates to the decline in
    glomerular filtration rate (GFR) occurring beyond the age of 50
    (Davies & Shock, 1950). The pathological basis for this decline is not
    certain but may be related to vascular changes that accompany aging
    (Avendano & Lopez-Novoa, 1987). Another possible explanation is that
    the kidneys of people over 50 years of age no longer respond to
    hypertrophic growth factors. Renal donors aged 50 or more show little
    or no functional increase after the loss of one kidney (Boner et al.,
    1972). Indeed, renal function reserve declines linearly with time
    after 30 years of age (Anderson & Brenner, 1986). Age may also reflect
    a loss of the ability of the renal tissue to repair. In young
    individuals nephrotoxicity in terms of tubular necrosis may be
    compensated for by constant repair, while in older patients this
    repair capacity may be diminished, resulting in the clinical
    expression of renal injury (Laurent et al., 1988).

         Pre-existing renal disease is an obvious risk factor predisposing
    to abnormal accumulation and excess blood levels of many nephrotoxic
    drugs and chemicals. It is not clear whether sex is a predisposing
    risk factor in humans, but male rodents are considerably more
    susceptible than females to nephrotoxicity and carcinogenicity from
    many environmental toxins (NTP, 1983, 1986, 1987).

         Factors such as short-term and high-dose exposure versus chronic
    and/or low-dose exposure influence vulnerability via the mode of
    metabolism, rate of excretion, etc. Long-term, low-dose exposure to
    substances that have a long biological half-life, such as lead or
    cadmium, increases risk from these nephrotoxins, but their role as
    co-risk factors is not known.


                     Table 3.  Frequency of combined risk factors in 143 patients
                           with acute renal failure (ARF)a
                                                                                                 
                                  Age      Hyper-     Gout/       Diabetes    Renal     Diuretics
                                           tension    hyper-                  disease
                                                      uricaemia
                                                                                                 
    Age (> 59 years)               30
    Hypertension                   29          4
    Gout/hyperuricaemia            21         18         4
    Diabetes                       11          6         4          1
    Renal disease                  18         12        12          6           4
    Diuretics                      29         27        21          8          13          0

    Multiple risksb               108         63        37         14          13          0
                                                                                                 
    a   Modified from: Rasmussen & Ibels (1982).
    b   Significant risk contribution to ARF based on discriminant multiple linear
        regression analysis.
    
    2.5.3  Extrapolation of animal data to man

         The use of animals has been essential to help define the
    molecular basis and the progression of model nephropathies, but it may
    be inappropriate to extrapolate animal toxicology data directly to man
    because of marked species, strain, dietary, and sex differences. In
    addition, there may be differences in dosing levels and regimen and in
    the absorption, distribution, metabolism, and excretion of potential
    nephrotoxins. There are also very significant differences in renal
    structural and functional characteristics in the common laboratory
    species used for risk assessment.

         Chemical safety assessment has generally been undertaken in
    relatively few strains of animals, such as the Sprague-Dawley, Wistar
    and Fisher-344 rats. There is limited information on inter-species
    comparisons.  In addition to assessing the renal differences in each
    of the species or strains used, it is necessary to examine extrarenal
    differences. Thus, for example, there are marked species differences
    in the hepatic handling of chemicals (Smith 1974; Testa & Jenner,
    1976) and the metabolic capacity of each of the major organ systems
    (Litterst et al., 1975a; Kluwe, 1983). This will have profound
    consequences on the amount of a parent chemical and the pattern of
    metabolites that reach the kidney. Dietary factors such as
    carbohydrate, lipid, and protein intake alter renal function, and the
    presence of contaminants and natural toxicants may add to the toxic
    burden of the kidney (Bridges et al., 1982).

         If the mechanistic basis of a renal injury is clearly
    established, it is easier to assess the risk of chemical injury in
    man, but such data are at present only available for a few chemicals.

    There are relatively few examples of nephrotoxic chemicals where there
    is a full profile of information from experimental animals and man,
    and in the vast majority of cases data are available only in rats.

         In order to extrapolate animal data for risk assessment, each
    screening procedure should cover a sensible level of exposure and a
    comparable condition to that found in man. The experience gained
    should provide a foundation from which a rational basis can be
    developed to identify potentially exacerbating risk factors and from
    which nephrotoxicity can be reduced. Some lesions can only be induced
    in rats with difficulty, and there may be a need to use sensitive
    species or strains and/or to adapt certain experimental manoeuvres to
    produce a lesion similar to that which occurs in man.

    2.5.4  Risk assessment from nephrotoxicity studies in animals

         Risk assessment from nephrotoxicity studies in animals has been
    best defined for therapeutic agents. Many have been widely tested in
    animals as a pre-clinical safety evaluation or used to study the
    mechanism of renal injury where there are adverse reactions caused by
    these compounds in clinical usage. The risk assessment for a number of
    workplace or environmental chemicals has been developed from animal
    models that have been used to study the mechanisms of these effects,
    especially those of the heavy metals and some of the industrial
    organic chemicals.

    2.5.5  Special risk groups in humans

         The marked variability in the response of any study population to
    potentially nephrotoxic compounds establishes clearly that there are
    groups at risk. There are a number of factors that could be
    responsible for increasing the risk of nephrotoxicity. These include
    existing renal disease, loss of renal parenchyma, high protein diet,
    chemical exposure, predisposing factors, multiple myeloma, and other
    conditions where there is an added level of protein excretion when the
    kidney is under an additional work-load.

         So far there has been relatively little interest in the
    individuals that do not appear to be at risk from exposure to
    potential nephrotoxins. While this is generally assumed to be the
    result of an absence of predisposing factors, there may well be other
    chemical, dietary, or disease considerations that provide a protective
    effect. There is experimental evidence to suggest that a pre-existing
    streptozotocin-induced diabetes and also poly-aspartic acid protect
    against aminoglycoside-induced renal injury, and that fish oil diets
    (high in omega-3 polyunsaturated fatty acids) reduce cyclosporin-A
    nephrotoxicity. These factors could well be used to reduce the health
    impact of nephrotoxicity.

    2.5.6  Multichemical exposure

         At present, there is virtually no information on the effects of
    multichemical exposure in man and very little data on the effects of
    more than one chemical administered simultaneously  in animals.
    Simultaneous  exposure to several chemicals represents a major
    toxicological problem, as man is generally exposed to more than one
    substance in medicines, in food, and from environmental factors.
    However, most experimental studies have investigated only single
    chemicals.  Interactions have been studied between mercuric chloride,
    potassium dichromate, citrinin,  and hexachloro-1,3-butadiene (HCBD)
    in vivo and in vitro using a rat model (Baggett & Berndt 1984a,b;
    1985).  Dichromate potentiates the mercuric  chloride effect, i.e. the
    effects produced by the combination of metals are always greater than
    the sum of the individual effects. There appears to be no simple
    kinetic explanation, i.e. no enhanced renal accumulation of the
    mercuric ion. The plasma membrane may be a site for the interaction of
    these metal ions and could be the preliminary step that leads to
    overall renal dysfunction and an ultimately enhanced acute renal
    failure. Dichromate-citrinin and dichromate-HCBD interactions have
    been demonstrated by alterations in urine flow, glucose excretion, and
    transport processes.  Some experimental data suggest that a
    synergistic interaction may occur in analgesic nephropathy. The
    mechanisms that underlie these interactions are not understood, and at
    present there is no rational basis to predict them.  Experimental
    studies have shown that tubular cell injury, induced by
    trichloroethylene and carbon tetrachloride, is potentiated by exposure
    to polyhalogenated biphenyls, e.g., polychlorinated biphenyls (Kluwe
    et al., 1979).

    2.5.7  Renal functional reserve

         The concept of renal functional reserve is a simple one in which
    not all of the nephrons nor all of the cellular functions in a single
    nephron are available or used at any one time (Friedlander et al.,
    1989). Thus there is a buffering capacity in the kidney that can cope
    with short-lived or protracted demands on function that exceed the
    normal level. Part of this functional reserve is used to meet the
    response to perturbation of the homeostatic system by water or
    electrolyte loading. Most of the studies on and understanding of renal
    functional reserve relate to changes in glomerular filtration rate and
    renal blood flow. It is likely that additional approaches are needed
    to test for other types of functional reserve.

    2.5.8  The effects of chemicals on kidneys with pre-existing renal lesions

         Although it is generally acknowledged that there are several
    types of renal lesions that exist as a nephropathy in the general
    population at a low but significant level (e.g., nephrotic syndrome),
    little is known about how these pre-existing lesions affect the
    response of the kidney to subsequent nephrotoxic insults.

    2.5.8.1  Nephrotoxicity in the presence of renal and extrarenal disease

         Safety screening is conducted on young, disease-free animals,
    housed under optimal conditions, fed contamination-free, high-protein
    food.  By contrast, man is exposed to a variety of dietary and
    environmental chemicals and to a poly-pharmacy of both prescribed and
    self-administered medications over many years. In addition, screening
    is generally undertaken using normal experimental animals. This may be
    inappropriate because, with the exception of occupational and
    environmental exposure, man is exposed to potentially nephrotoxic
    therapeutic substances to treat disease. Pre-existing diseases can
    have a profound effect on the direct or indirect response of the
    kidney to handling chemicals (Bennett, 1986). There is an increasing
    wealth of animal data to demonstrate that common clinical conditions
    in man, such as hypertension, renal compromise, and renal ischaemic
    injury, exacerbate cyclosporin nephrotoxicity and bacterial endotoxins
    in animals and that systemic infection increases the sensitivity of
    the kidney to aminoglycoside toxicity (Bergeron et al., 1982).

         The role of pre-existing renal lesions on nephrotoxicity is
    important, but there are few clear indications of what can be
    predicted from existing clinical data and from animal studies.
    Diabetes is generally associated with reduced renal function, diabetic
    nephropathy, and renal papillary necrosis, and might be expected to
    exacerbate chemical-associated nephrotoxicity. Untreated
    streptozotocin-induced diabetes, however, protects rats against
    gentamicin, low-dose cisplatin, and uranyl nitrate nephrotoxicity
    (Teixeira et al., 1982; Vaamonde et al., 1984). Recent studies on the
    acute effects of intravenous radio-contrast media on anaesthetized
    diabetic rats were inconclusive (Reed et al., 1983, Golman & Almen,
    1985, Leeming et al., 1985).

    3.  KIDNEY STRUCTURE AND FUNCTION

         An in-depth review of kidney structure and function is beyond the
    scope of this monograph. Only sufficient information will be given to
    provide a general background against which nephrotoxicity can be
    framed. A fuller insight into the complexities of the kidney in
    health, disease, and nephrotoxicity has been described by Valtin
    (1973), Orloff & Berliner (1973), Hook (1981), Porter (1982), Bach et
    al. (1982, 1989), Bach & Lock (1982, 1985, 1987, 1989), Seldin &
    Giebisch (1985), Brenner & Rector (1986).

    3.1  Renal anatomy

         The two kidneys are situated retro-peritoneally, on either side
    of the vertebral column, and process 25% of the resting cardiac output
    via an arterial blood supply. Much of the fluid and most of the
    solutes in blood are filtered through the glomeruli into the proximal
    part of the nephron (the functional unit of the kidney) from which
    essential small molecules are reabsorbed. Numerous macro-molecules are
    reabsorbed into the tubular cells by an endocytotic process and are
    digested in tubular lysosomes. Many organic acids and bases (including
    many drugs) are secreted (and reabsorbed) by carrier-mediated
    processes located principally in the proximal tubule. There is some
    secretion, mainly of waste solutes, from the blood into the distal
    part of the nephron, and much of the water in which they are dissolved
    is subsequently reabsorbed.

         Each kidney is made up of a large number of nephrons, groups of
    which unite to continue as collecting ducts or tubules, and these in
    turn combine to make up the ducts of Bellini, which exit around the
    papilla tip. The papilla opens into the calix, which is in continuity
    with the renal pelvis, a funnel-shaped area that narrows to the
    ureter.  The continued production of urine, together with peristalsis
    of the ureter, carries excreted waste to the bladder. The
    morphophysiology of the kidney varies markedly between species.
    Therefore, a generalized description will be provided, and only the
    important differences between the rat (and other common laboratory
    animals) and man will be described (Moffat, 1979).

    3.1.1  Histology

         Renal lesions occur in discrete anatomical regions. This
    highlights the need to understand changes in terms of the biochemical
    properties of the specifically affected region and its adjacent cells.
    While haematoxylin and eosin staining and a number of other routinely
    used staining procedures identify nephropathies and renal
    degeneration, these are generally based on a relatively non-specific
    assessment.  The non-specificity of  routine histopathology has, in
    fact, been the strength of these methods  in the preliminary
    assessment of chemically induced nephropathies. It may, however, miss
    some types of lesions and generally gives little information that can
    help identify the mechanistic basis of a lesion.

         Histochemical techniques can provide insight into primary and
    secondary cellular mechanisms. One aspect of "histochemistry" is the
    use of frozen segments of the nephron (Bach et al., 1987) and the
    application of fluori-metric or radiochemical assays to measure the
    activities and distribution of specific biochemical characteristics.
    Microdissection generally fails to give a detailed localization of
    these properties in relation to specific or individual cells. In
    addition, the technique is difficult to apply to injured renal cells.

         The most widely used histochemical approach is based on obtaining
    frozen, fixed frozen, or fixed embedded sections of the kidney that
    are then used with chromo- or fluoro-phores.  These react with a
    selected type of material. The types of materials that can be
    visualized in section depends on their chemical structure (e.g.,
    carbohydrate), enzymic activity (e.g., lactate dehydrogenase),
    antigenicity (e.g., specific molecules), or  physicochemical
    properties (e.g., lipophilicity), some of which are shown in Fig. 2.
    This approach also includes the distribution or incorporation of
    radiolabelled molecules by their interaction with a photographic film
    laid over the section (Bach et al., 1987). Immunohistochemical
    techniques using labelled antibodies permit antigens to be localized
    at the light microscopical and ultrastructural levels.

         These microscopic histochemical techniques provide information on
    the distribution at, or within, specific cells and their relative
    activities, and have been used to define a variety of characteristics
    of the kidney. It is important to stress that each method has its own
    inherent strength and weakness, and that it may be difficult to relate
    data from tissue sections to absolute biochemical measurements. This
    is a consequence of the complex mixture of materials that are present
    in tissue sections and the chemical changes that may take place in
    these sections, particularly once they have been fixed. While the
    biochemical characteristics of tissue in frozen sections are least
    likely to be adversely affected, subtle and misleading alterations in
    chemical properties can still occur.  In addition, treatments to
    conserve morphological features (fixation and embedding) alter these
    values. It may also be difficult to be certain whether an increased
    intensity of staining for a certain substance represents  de novo
    synthesis, unmasking, or the loss of factors that suppress staining.
    These two techniques have proved to be very powerful in providing
    information on the biochemical characteristics of cells and the
    changes associated with renal injury. The in situ hybridization
    techniques, using specific labelled nucleotidic probes, permit the
    detection of protein synthesis at a subcellular level.

    3.1.2  Enzyme histochemistry and quantification

         Histochemical heterogeneity is evident in each part of the
    nephron and varies between different species. In addition, the profile
    of characteristics within any region of the kidney and nephron may be
    related to sex and age. Antibodies are directed against unique or
    novel characteristics along the nephron that are present on enzymes,

    FIGURE 2a

    FIGURE 2b

    FIGURE 2c

    FIGURE 2d

    FIGURE 2e

    FIGURE 2f

    FIGURE 2g

    FIGURE 2h

    FIGURE 2i

    glycoproteins, or other molecules associated with membranes or soluble
    cytosolic constituents. The biochemical characteristic can be
    visualized using enzyme, fluoro-phore, or radioactive labels (Bach et
    al., 1987). Individual microdissected nephron segments have been
    studied to determine the quantitative distribution of selected enzymes
    (Table 4). Table 5 lists biochemical and function parameters for the
    different nephron segments, showing their individual receptor
    sensitivity.

    3.1.3  Immunohistochemistry

         The use of antibodies raised towards enzymes can help in the
    study of isoenzyme distribution and factors affecting changes in their
    distribution. For instance, aldolase-B monomers increase in the
    proximal tubules of rats during renal maturation, but not in the
    distal tubules. In contrast, aldolase-A monomers increase in the
    distal tubules but not in the proximal tubules.

         Immunohistochemical techniques demonstrate many functional
    proteins at discrete locations in the kidney, including the glomeruli
    and the tubular basement membrane, but there are few data on changes
    in these proteins as a result of nephrotoxicity. A variety of
    immunodeposits are associated with glomerulopathies including those
    caused by heavy metals, but these appear to be T-cell mediated. They
    are assessed by immunofluorescent monitoring in the glomerulus, but
    this acts as a passive sieve and may be involved as a secondary
    consequence of immunodeposition.

         Histochemistry can also be used to show the distribution of a
    number of oxidative enzymes. Cytochrome P-450 mixed-function oxidase
    activities have been shown immuno-histochemically to be localized in
    the proximal tubule, particularly in the S2 and S3 segments in the

                            Table 4.  Distribution of enzymes in individual nephron segments
                                         in various animal species
                                                                                                 

          Enzyme                                      Relative activitya            Animal
                                                                                                 
    Specific to the glomerulus

    Adenosine deaminase                                                             rat

    Specific to the proximal tubule

    Glucose-6-phosphatase                             S1>S2>S3                      rat
    Fructose-1,6-bisphosphatase                       S1<S2>S3                      rat
    Phosphoenolpyruvate carboxykinase                 S1>S2>S3                      rat, rabbit
    Fructokinase                                      S1=S2<S3                      rat
    Fructose-1-phosphate aldolase                     S1=S2>S3                      rat
    Glycerokinase                                     S1=S2>S3                      rat, rabbit
    Glycerol-3-phosphate dehydrogenase                S1=S2<S3                      rat
    Glutamine synthetase                              S3                            rat
    Alanine aminotransferase                          S1=S2<S3                      rat
    Gamma-glutamyltraspeptidase                       S1<S2<S3                      rabbit
    Gamma-glutamyl-cysteine synthetase                S3                            rat
    Glutathione-S-transferase                         S1<S2<S3                      rabbit
    Cytochrome P-450                                  S1=S2>S3                      rat, rabbit
    Alanine aminopeptidase                            S1<S2<S3                      rat
    Alkaline phosphatase                              S1=S2=S3                      rat
    Leucine aminopeptidase                            S1<S2<S3                      rat
    D-Amino acid oxidase                              S1=S2<S3                      rat
    L-Hydroxy acid oxidase                            S1=S2<S3                      rat
    Fatty-acyl-CoA oxidase                            S1=S2<S3                      rat
    Choline oxidase                                   S1=S2<S3                      rat
    25(OH)-D3-1alpha-hydroxylase                      S1<S2=S3                      rat
                                                                                    (D3 deficient)
                                                                                    rabbit (fetus)

    Table 4 (contd).
                                                                                                 

    Enzyme                                            Relative activitya            Animal
                                                                                                 
    Relatively specific to the proximal tubule

    Glutamate dehydrogenase                           S1=S2>S3                      rat, rabbit
    Malic enzyme                                      S1=S2<S3                      rat
    Trypsin-type protease                             S1>S2>S3                      rat
    ß-D-Galactosidase                                 S1=S2<S3                      rat
    N-Acetyl-ß-D-glucosaminidase                      S1=S2>S3                      rat
    Xanthine oxidase                                  S1>S2>S3                      rat
    Superoxide dismutase                              S1>S2>S3                      rat

    Specific to the lower nephron

    Hexokinase                                  MTAL=CTAL>DCT>CCD=MCD               rat, rabbit
    Phosphofructokinase                         MTAL=CTAL=CCD=MCD                   rat, rabbit
    Pyruvate kinase                             MTAL=CTAL<DCT<CCD=MCD               rat, rabbit
    Kallikrein                                  CNT                                 rabbit

    Relatively specific to the lower nephron

    Fructose 1,6-bisphosphate aldolase          MTAL=CTAL>DCT                       rat
    Glycerol dehydrophosphate                   CCD<MCD                             rabbit
     dehydrogenase

    Table 4 (contd).
                                                                                                 

    Enzyme                                            Relative activitya            Animal
                                                                                                 
    Relatively specific to the
     lower nephron (contd.)

    Lactate dehydrogenase                       MTAL=CTAL-DCT=CCD                   rat, rabbit
    Aspartate aminotransferase                  MTAL=CTAL>DCT                       rat
    Citrate synthase                            MTAL=CTAL=DCT>CCD                   rat
    Isocitrate dehydrogenase (NAD+)             MTAL=CTAL=DCT=CCD                   rat
    Na,K-ATPase                                 MTAL<CTAL<DCT>CCD                   rat
                                                                                                 

    a     CCD = cortical collecting duct; CNT = connecting tubule; CTAL = cortical thick ascending
          limb of Henle's loop; DCT = distal convoluted tubule; MCD = medullary collecting duct;
          MTAL = medullary thick ascending limb of Henle's loop;
          S1 = early proximal tubule; S2 = middle proximal tubule; S3 = late proximal tubule.
    
    rat and rabbit. Large numbers of peroxisomes containing D-amino acid
    oxidase and catalase are localized in the S3 portion of the proximal
    tubule, but they are absent from the glomerulus and the distal
    nephron. There is little immuno-histochemical data on the distribution
    of molecules that are likely to protect renal cells from the effects
    of reactive intermediates. Ligandin or glutathione- S-transferase B
    is located in the proximal tubule of both animals and man and in the
    thick limb of the loop of Henle in man. Catalase activity is greatest
    in the proximal tubule (where it is localized in the peroxisomes),
    less in the distal tubule, and very low in the glomerulus. Glutathione
    has been shown by histochemistry to be localized in the proximal
    convoluted tubule. However, there are some uncertainties as to what is
    being assessed, since the reaction measures sulfhydryl groups and not
    only glutathione. The distribution of at least one superoxide
    dismutase isoenzyme shows a marked species difference between the dog
    and rat, but is localized in the proximal tubules in both (Bach et
    al., 1987).

    3.2  The renal blood supply

         Each kidney is supplied by a renal artery (a branch of the
    abdominal aorta), which divides to form several interlobar arteries
    (Fig. 3). These in turn give rise to the arcuate arteries, which run
    between the cortex and medulla parallel to the kidney surface.  Many
    cortical radial arteries arise from the arcuate vessel and pass
    through the cortex. Here a small amount of blood reaches the surface
    to supply the kidney capsule, but most of the blood flow is directed
    through branches that form the afferent arterioles to the glomeruli.
    Each afferent arteriole breaks up to form the capillary plexus of the
    glomerulus; this is drained into the efferent arteriole. The efferent
    arterioles form two types of capillary networks

    *    In the "mid" and "superficial" cortical regions, they form the
         peritubular capillaries surrounding proximal and distal tubules
         (in the superficial regions some peritubular capillary networks
         interlace the nephron from which they were derived, but such an
         association appears to be the exception rather than the rule).

    *    In the juxtamedullary region (and some mid-cortical areas in man)
         each efferent arteriole is directed into the medulla, where it
         branches into the vasa recta bundles. Each bundle consists of up
         to 30 descending vessels, the peripheral vessels of which give
         rise to a highly branched capillary network in the outer medulla. 
         The core of the vasa recta bundle continues to the inner medulla
         where it terminates in a capillary network (Beeuwkes, 1980).

         The walls of all peritubular capillaries in the kidney are made
    up of a thin fenestrated endothelium resting on a basal lamina. The
    capillaries in the cortex generally open into the cortical radial
    vein, from which blood flows via the arcuate vein to the renal vein
    and finally to the inferior vena cava. The capillary plexuses in the


                            Table 5.  Summary of nephron heterogeneitya
                                                                                                 

    Anatomical regionb               Biochemical featuresc                   Functional features
                                                                                                 
    Glomerulus SF<JM              Renin:(SF>JM)                           GFR:(SF<JM)
      epithelium                  Adenosine-AC                            Mesangial contraction
                                                                          (AII, histamine)
      endothelium                 Histamine-AC (intra-mesangium)
      mesangium                   Serotonin-AC (extra-mesangium)
                                  ANP-GC                                  ROM generation (intra-
                                                                          and extra-mesangium)
                                  ET-PGE2 (intra-mesangium)
                                  AVP-PGE2 (intra- and extra-mesangium)
                                  AII-PGE2 (intra- and extra-mesangium)

    Proximal tubule               Glucose carrier:                        J-glucose:
      S1, S2, S3                  (brush border)                          (S1>S2,S3)
                                  Gluconeogenesis
                                  (S1>S2>S3)                              J-V:(S1>S2>S3)
                                  Cytochrome P-450:                       1,25(OH)2D3 synthesis
                                  (S1<S2>S3)
                                  NADPH-cytochrome c reductase:           J-V decreased by ANP
                                  (S1<S2>S3)
                                  Ammoniagenesis                          P-Cl/P-Na:(SF>JM)
                                  (S1>S2>S3)
                                  PTH-AC(S1>S3)                           PAHsecretion:
                                                                          (S1<S2>S3)
                                  Adenosine-AC                            J-aminoacid:
                                                                          (S1>S2,S3)

    Henle's loop
      Thin: DTL<ATL               AVP-AC(+ATL,-DTL)                       P-water:(DTL>ATL)
                                                                          P-NaCl:(DTL<ATL)
                                                                          P-urea:(DTL<ATL)

    Table 5 (contd).
                                                                                                 

    Anatomical regionb               Biochemical featuresc                   Functional features
                                                                                                 
      Thick: MTAL>CTAL            AVP-AC:(MTAL>CTAL)                      AVP stimulation of
                                                                          J-Cl: (MTAL>CTAL)
                                  PTH-AC:(MTAL<CTAL)                      PTH stimulation of
                                                                          J-Ca: (MTAL<CTAL)
                                  SCT-AC:(MTAL>CTAL)
                                  Tamm-Horsfall glycoprotein
                                  PGE2 synthesis                          J-NaCl:(MTAL>CTAL)
                                  (MTAL>CTAL)                             PG inhibition of J-Na:
                                                                          (+MTAL, -CTAL)
                                                                          EGF synthesis
                                                                          J-Na decreased by ANP

    Distal tubule
      DCT:single cell type        SCT-AC:(+DCT)                           SCT suppression of
                                                                          Vt:(DCT)
      CNT:multiple cell           PTH-AC:(+CNT)                           PTH stimulation of
      types                                                               J-Ca: (CNT)
                                  AVP-AC:(+CNT)                           AVP suppression of
                                                                          Vt:(CNT)
                                  ISO-AC:(+CNT)                           ISO suppression of
                                                                          Vt:(CNT)
                                  Kallikrein:(+CNT)
                                  Aldosterone binding                     Vt:(DCT < CNT)
                                                                          K-secretion:(DCT<CNT)?
    Collecting duct system
      two cell types              AVP-AC:(CCD > OMCD)
      CCD (P.cell>I.cell)         ISO-AC:(CCD > OMCD)                     Vt:(CCD > OMCD)
                                                                          J-Na and J-V decreased
                                                                          by ANP

    Table 5 (contd).
                                                                                                 

    Anatomical regionb               Biochemical featuresc                   Functional features
                                                                                                 
      OMCD (P.cell>I.cell)        PG-AC:(CCD < OMCD)                      P-urea:(CCD,OMCD<IMCD)
                                  PGE2 synthesis
                                  (CCD<OMCD<IMCD)
                                  Aldosterone binding
                                  Adenosine-AC (CCD>OMCD)
                                  ANP-GC (CCD<OMCD<IMCD)
      IMCD                                                                J-V decreased by ANP
                                                                                                 

    a    Based on data obtained from the rabbit and rat kidney.
    b    Parts of the nephron: ATL = ascending thin limb of Henle's loop; CCD = cortical collecting
         duct; CNT = connecting tubule; CTAL = cortical thick ascending limb of Henle's loop; DCT =
         distal convoluted tubule; DTL = descending thin limb of Henle's loop; I.cell = intercalated
         cell; IMCD = inner medullary collecting duct; JM = juxtamedullary nephron; MCT = medullary
         collecting duct; MTAL = medullary thick ascending limb of Henle's loop; OMCD = outer
         medullary collecting duct; P.cell = principal cell; S1 = early proximal tubule; S2 = middle
         proximal tubule; S3 = late proximal tubule; SF = superficial nephrons.
    c    Hormone effects on nephron receptors: AII = angiotensin II; AC = adenylate cyclase; ANP =
         atrial natriuretic peptide; AVP = arginine vasopressin; GC = guanylate cyclase; EGF =
         epidermal growth factor; ET = endothelin; ISO = isoproterenol; PG = prostaglandin; PTH =
         parathyroid hormone; ROM = reactive oxygen metabolites; SCT = salmon calcitonin.
         Transport substances: PAH = para-aminohippurate; J-x = flux of substance x; J-V = flux of
         fluid volume; P-x = permeability for substance x; Vt = transcellular voltage.
    
    FIGURE 3

    medulla drain into the ascending vasa recta, which join the arcuate
    veins. The arterial branches are terminal, without anastomoses. In
    contrast, the veins are richly anastomosed.

         There is a well-defined structural relationship between the vasa
    recta bundles and the nephrons in the outer medulla, at least in the
    animals that have been studied. A central core (consisting of
    descending and ascending vasa recta) is surrounded by a peripheral
    layer consisting of a closely intermingled ascending vasa recta and
    the descending thin limbs of the loops of Henle. Between these bundles
    are the thick ascending limbs of the loops of Henle, some descending
    limbs, and the collecting ducts.  Within the bundles both ascending
    and descending vasa recta are in intimate contact with each other
    (rather than with the same type of vessel). There are more ascending
    vessels, all of larger diameter, than descending ones, and this
    increased volume capacity relates to the removal of excess water from
    the interstitium and the maintenance of the medullary osmotic gradient
    shown in Fig. 4.

         Many of the "major" and "minor" blood vessels in the kidney have
    either smooth muscle cells as an integral part of their structure, or
    other cells that may have a contractile function. Thus most of the
    intrarenal vascular system has both adrenergic and cholinergic
    innervation. Intrarenal blood flow, the factors which alter it, and
    its effects on renal function are poorly understood. Although there
    appear to be the facilities for a direct and effective perfusion of
    the medulla from the arcuate artery, this does not seem to occur
    (Moffat, 1979; Beeuwkes, 1980).

    3.2.1  Renal haemodynamics

         The measurement of total blood flow through the kidneys can be
    measured relatively easily using modern techniques (Grunfeld et al.,
    1971; Pearson, 1979). Defining the zonal blood flow has given some
    conflicting results, but assessing regional blood flow within the
    kidney is fraught with difficulties and is subject to varied
    interpretations (Grandchamp et al., 1971; Grunfeld et al., 1971;
    Pearson, 1979; Aukland, 1980; Knox et al., 1984). There are, however,
    consistent data (derived from a number of  fundamentally different
    techniques) to show that intrarenal blood flow is greatest in the
    cortex (80-85% of total renal flow) and that it decreases through the
    juxtamedullary region to less than 10% of the total renal flow in the
    medulla. It must be stressed that, although the medulla is poorly
    perfused in comparison to the rest of the kidney, it is, nonetheless
    (because of the 25% resting cardiac output and therefore abundant
    renal blood flow), a well-perfused tissue. According to Thurau (1964),
    the medullary blood flow is about 15 times that of resting muscle and
    the same as that of the brain.  In addition, the capillary volume
    fraction of the medulla is more than twice that of the renal cortex
    (Beeuwkes, 1980). Despite this, there is considerable variation in
    tissue pO2 in the kidney; a marked decrease in pO2 levels is seen
    with increasing tissue depth (Brezis et al., 1984).

    3.3  The nephron

         The kidney is divided into three main regions, cortex (outer),
    medulla (inner), and pelvis (Fig. 5). Within the cortex arise the
    renal corpuscles, defined as superficial, midcortical or
    juxtamedullary depending on the anatomical location of the renal
    corpuscle in the cortex. The nephron is the functional unit of the
    kidney and consists of a continuous tube of highly specialized
    heterogeneous cells, which show sub-specialization along the length of
    nephrons and between them. There are marked structural and functional
    differences between the nephrons arising in the cortex and those
    arising in the juxtamedullary regions. The total number of nephrons
    varies between different species and within any one species as a
    function of age. The macroscopic differentiation of the kidney into
    distinct zones arises not only from the regional vascularity but also
    from the way different functional parts of the nephron are arranged
    within the kidney. A more detailed account of the ultrastructure of
    the morphologically definable regions of the nephron and their
    functional inter-relationship has been provided by Moffat (1981,
    1982), Bohman (1980), and Maunsbach et al. (1980). Recently the
    nephron nomenclature has been standardized by the Renal Commission of
    the International Union of Physical Sciences (Kriz & Bankir, 1988).
    This is summarized in Figures 3, 5, and 6, and Table 6.

    3.3.1  Cellular heterogeneity and cell-cell interaction

         There are well over 20 morphologically different cell types
    (based on light microscopy alone) in the kidney, and when
    histochemical and immunohistochemical methods are applied to renal
    tissue sections the diversity of cell types is even more apparent. The
    spectrum of biochemical (and structural and functional)
    characteristics in these cells demonstrates the very marked
    heterogeneity that is the hallmark of the kidney. It is well
    established that the expression of many of these biochemical
    characteristics is an integral of the functions of that particular
    region of the kidney, and there is the potential to change the
    expression of these characteristics in terms of the demands on the
    kidney. These include both water and electrolytes, dietary factors,
    and chemicals with pharmacological and toxic effects, or may be as a
    result of chemical and other types of injury. More importantly, the
    characteristics of a cell may make it either resistant or sensitive to
    the target selective toxicity of a chemical.

    FIGURE 4

    3.3.2  The glomerulus

         The glomerulus forms the initial part of the nephron and
    functions as a relatively poorly selective macro-molecular exclusion
    filter to the hydrostatic pressure of the blood. The number of
    glomeruli is, in general, related to the mass of the species, and the
    size of each glomerulus depends, among other factors, on the
    environmental water balance. Three anatomically distinct types of
    glomeruli can be identified: those in the superficial cortex, which
    are part of the superficial nephrons; those arising in the midcortical
    area; and those ofjuxtamedullary origin, which continue as nephrons
    that loop down into the medulla. The structure of the glomerulus is
    complex (Fig. 7) and has only been defined using scanning and
    transmission electron microscopy (Maunsbach et al., 1980; Moffat 1981,
    1982).

         The glomerular "tuft" is made up of a number of capillary
    branches that arise from the afferent arteriole, anastomose, and drain
    to the efferent arteriole. There are also communicating vessels
    between the branch capillaries. The fenestrated endothelium cannot
    prevent plasma molecules from leaving the lumen, but a negatively
    charged cell coat imparts some selective permeability. The capillaries
    are in direct contact with the glomerular basement membrane (or basal
    lamina), which, when viewed under the electron microscope, can be
    divided into three layers: the lamina rara interna on the endothelial
    side; the central lamina densa; and the lamina rara externa, which is
    in direct contact with the epithelial cells (the podocytes). The basal
    lamina contains collagen (mostly Type IV) and sialic acid and is rich
    in glycosaminoglycans, mainly heparan sulfate (Kanwar & Farquhar,
    1979), which provides a strongly anionic macromolecular filtration
    barrier.

         The capillary tuft (ensheathed in its basal lamina) is surrounded
    by a number of podocytes, each of which gives rise to several primary
    processes (trabeculae). These in turn give rise to secondary
    processes, and, finally, to numerous tertiary foot processes that are
    embedded in the lamina rara externa.

         The foot processes of one podocyte interdigitate with those of an
    adjacent epithelial cell for adjacent trabeculae. The surfaces of the
    podocytes are covered by a strongly anionic cell coat that extends to
    the spaces between the foot processes. It is through these spaces that
    the glomerular filtrate reaches the lumen of Bowman's space. Thus, the
    podocyte provides a structural support for the basal lamina and may
    also serve to provide additional anionic forces for the process of
    biological ultrafiltration. It has been suggested that podocytes may
    have phagocytic properties and undergo contraction (Moffat, 1981).

    FIGURE 5

    FIGURE 6

        Table 6. Summary of nomenclature of segments and cells of the renal tubule (From: Kriz & Bankir, 1988).  A continuous serpentine
    arrow ( ) means that the transition between the two structures is gradual. An interrupted serpentine arrow ( ) means that the
    transition is gradual in some species, abrupt in others. Abbreviations marked by a star were introduced by Morel and coworkers
    (Kidney Int 9: 264, 1976). They mean: DCTa = Distal convoluted tubule, initial portion; DCTb = Distal convoluted tubule, bright
    portion; DCTg = Distal convoluted tubule, granular portion; DCTI = Distal convoluted tubule, light portion; CCTg = Cortical
    collecting tubule, granular portion; CCTI = Cortical collecting tubule, light portion
                                                                                                                                    
    Micro-
    Anatomical   Main division   Subdivisions          Segmentation            Abbre-    Cell types      Other frequently used
    terms                                                                      viation                       denominations
                                                                                                                                    
    Proximal     PROXIMAL      pars convoluta    Proximal                             S 1  cells
    convolution  TUBLULE           or            Convoluted   S 1 - segment    PCT                    P 1    segment             PT
                               convulated part   Tubule

                                                              S 2 - segment           S 2  cells      P 2    segment

                               pars recta        Proximal
                                  or             Straight     S 3 - segment    PST                    P 3    segment             PR
                               straight part     Tubule                               S 3  cells
                                                                                                                                    
    Loop of      INTERMEDIATE  pars descendens   Descending                    DTL    DTL cells
    Henle        TUBULE           or             Thin         of short loops                  Type 1  Short Descending Thin Limb of
                               descending part   Limb         of long loops                           Henle's loop (SDL)
                                                              upper part                      Type 2  Long Descending Thin Limb,
                                                              lower part                              upper part (LDLu)
                                                                                              Type 3  Long Descending Thin Limb,
                                                                                                      lower part (LDLl)
                                                              pre-bend segment        ATL cells
                               pars of ascendens
                                   or            Ascending Thin Limb           ATL            Type 4  Tal    Thin Ascending Limb
                               ascending part                                                                (of long loops only)
                               ascending part
                                                                                                                                    

   Table 6 (contd.)
                                                                                                                                    
    Micro-
    Anatomical   Main division   Subdivisions          Segmentation            Abbre-    Cell types      Other frequently used
    terms                                                                      viation                       denominations
                                                                                                                                    
                 DISTAL        pars recta        Distal     Medullary          D MTAL DST or Tal      MAL    Thick Ascending Limb of
                 TUBULE           or             Straight   straight           S      cells                  Henle's Loop
                               straight part     Tubule     part               T                      mTALH    Medullary Thick Limb
                                 or                         Cortical           o                      CAL      Cortical Thick Limb
                               Thick                        straight           r CTAL                 cTALH     (incl. Macula Densa)
                               Ascending                    part  Macula Densa T MD            MD     MD
                               Limb                         postmacular        A               cells
                                                            segment            L                      DCTa*  early
    Distal                     pars convoluta                                         DCT cells              distal    Distal Tubule
    convolution                   or             Distal Convoluted Tubule      DCT    (+ IC cells)    DCTb*  tubule
                               convoluted part
                                                                                                                                    
                 COLLECTING    CONNECTING TUBULE                                      CNT cells
                 SYSTEM                                                        CNT       +            DCTg*  late      Connecting
                                                                                      IC cells        CCTg*  distal       Segment
    Collecting                 COLLECTING                                             CD cells               tubule
    duct                       DUCT              Cortical Collecting Duct      CCD    = principal
                                                                                      cells           DCTl*            Initial
                                                                                                                       Collecting
                                                                                                                       Tubule
                                                                                      = light cells   CCTl*            Cortical
                                                                                           +                           Collecting
                                                                                      IC cells                         Tubule (CCT)
                                                                                      = intercalated
                                                                                      cells
                                                                                      = mitochondria-
                                                                                      rich cells      Outer Medullary Collecting
                               Outer Medullary                                 OMCD   = carboanhyd-   Tubule (OMCT)
                               Collecting Duct                                        rase-rich cells
                                                                                      = dark cells
                                                                                      CD cells        Inner Medullary Collecting
                               Inner Medullary                                 IMCD   = prinicpal     Tubule (IMCT)
                               Collecting Duct                                        cells           Papillary Collecting Duct
                                                                                                      (PCD) or ducts of Bellini
    
         The axial regions of each glomerulus contain mesangial cells.
    Information on the structure and possible functions of mesangial cells
    has been reviewed by Moffat (1981). In brief, they undergo contraction
    and may thus control glomerular blood flow via biogenic amine or
    hormonal control. Of equal importance is the observation that these
    cells take up large molecules (such as colloids, immune complexes, and
    protein aggregates), which may eventually be disposed of via the renal
    lymphatic system.

         The driving force for filtration is provided by the glomerular
    capillary hydrostatic pressure (which is controlled mainly by the
    vascular tone of the afferent and efferent arterioles), minus both the
    plasma osmotic pressure and the hydrostatic pressure in the Bowman's
    space. The resulting "effective filtration pressure" across the basal
    lamina is about 1.2-2.0 kPa (10-15 mmHg). Selective filtration is
    achieved primarily on the basis of size restriction by the basement
    membrane, which impedes the passage of macromolecules with an
    effective radius greater than 1.8 nm and completely prevents the
    filtration of macromolecules with an effective radius greater than 4.5
    nm. In addition the presence of fixed negative charges on the
    endothelial, epithelial, and basement membranes hinders the filtration
    of anionic macromolecules while facilitating the passage of cationic
    macromolecules. The selectivity of filtration is, in part, a
    consequence of the anionic nature of the basement membrane, which
    blocks or slows the passage of negatively charged or neutral
    macromolecules and leaves those carrying a cationic charge and small
    molecules (irrespective of charge) to pass unimpeded.

    3.3.3  The proximal tubule

         The proximal tubule is found only in the cortex or subcortical
    zones of the kidney. Anatomically each proximal tubule can be divided
    into the convoluted portion (pars convoluta) and the shorter straight
    descending portion (the pars recta), which then continues to become
    the descending limb of the loop of Henle. It may be sub-divided, by a
    number of morphological and functional features, into three segments,
    S1, S2, and S3.

         The proximal tubule plays a decisive role in maintaining
    homeostasis. This is achieved when sodium and chloride ions flux from
    the tubule lumen to the peritubular capillaries under the control of
    a number of processes such as nonspecific electrophysiological
    gradients and selective active transport mechanisms. Water follows the
    ions by osmotic effects. In addition, hydrostatic pressure,
    attributable to the presence of both proteins and glycosamino-glycans
    (Wolgast et al., 1973), contributes to water movement from the
    epithelial cell to the interstitium and thence, by an osmotic
    gradient, into the capillaries (Valtin, 1973). The flux of ions within
    the proximal tubule, including the absorption and secretion of

    FIGURE 7


    HCO3- and H+ and the "lumen trapping" of ammonium ions, controls
    renal acid-base regulation (Valtin, 1973).

         Those proteins that have passed from Bowman's capsule (a
    significant amount of albumin in the case of normal rats) are
    reabsorbed in the proximal tubule by pinocytotic removal from the base
    of the microvillous brush border into the epithelial cells. The
    vesicles thus formed combine, form protein-filled vacuoles, and fuse
    with lysosomes, from which the digestion products of the protein
    diffuse, eventually, to the capillary system or are used in the
    metabolic processes of the cell.

         There are, in addition, other absorptive and secretory
    mechanisms. These include the co-transport process that reabsorbs
    glucose and the secretion of both acidic and basic organic compounds
    (Valtin, 1973; Orloff & Berliner, 1973; Brenner & Rector, 1986;
    Berndt, 1989).

    3.3.4  The medulla

         The medulla differs from the cortex (Fig. 5 and Fig. 8) both at
    the macroscopic and at the microscopic levels. This region can be
    divided into the outer medulla (which is made up of the thin
    descending and the thick ascending limbs of the loops of Henle,
    collecting ducts, the vasa recta, and a dense capillary network) and
    the inner medulla, the free part of which is referred to as the
    "papilla" (although some researchers apply that name only to the apex
    of this region). The inner medulla contains the thin limbs of the
    loops of Henle, collecting ducts, the vasa recta, and a diffuse
    network of capillaries.  Packed into the spaces between these
    structures are interstitial cells embedded in a matrix rich in
    glycosaminoglycans.

         The collecting ducts terminate as the ducts of Bellini around the
    tip of the papilla. Whereas the mouse, gerbil, rat, guinea-pig,
    rabbit, dog, cat, and primate kidneys have only a single papilla, the
    pig and man have multi-papillate kidneys. There are between 9 and 20
    papillae in each human kidney (Burry et al., 1977), of which there are
    two anatomically distinguishable types. The conical non-refluxing
    papillae, where the surface orifices of the ducts of Bellini are
    slit-like, close when there is an increase in the "back-pressure" of
    urine from the bladder and so prevent intrarenal reflux when reflux
    occurs from the bladder.  These papillae occur predominantly in the
    mid zone.  The refluxing papillae occur predominantly in the polar
    regions, and, as they have flattened tips, the collecting duct
    orifices are wide and prone to retrograde flow of urine into the
    tubules during vesico-ureteric reflux (Ransley & Risdon, 1979). The
    microscopic and ultrastructural features of the medulla have been
    described by several researchers (Moffat, 1979, 1981, 1982; Bohman,
    1980; Maunsbach et al., 1980).

    FIGURE 8

    3.3.4.1  The loops of Henle

         The loops of Henle may be divided into two populations on
    anatomical grounds. Short loops penetrate no further than the outer
    medulla. The proximal tubule and thick ascending limb are closely
    associated in the cortex, but in the medulla the descending limb is
    intimately related to the ascending vasa recta, and the ascending limb
    to the collecting duct.  The association of the ascending and
    descending limbs of the loop of Henle with the vascular system or with
    the collecting ducts provides a multi-dimensional network in which
    solutes or water may undergo countercurrent exchange.  These exchanges
    may  either provide a shunt that excludes selected solutes (and water)
    from the inner medulla or, alternatively, solutes (e.g., sodium
    chloride and urea) may be trapped in this zone. This exclusion of
    water and trapping of sodium chloride, urea, and osmolytes helps
    maintain the osmotic gradient along the inner medulla. In long loops
    (the length is proportional to the renal concentrating potential), the
    loop of Henle penetrates the inner medulla. Only about a third of the
    ascending and descending limbs of long loops lie together; in the
    other instances the ascending limbs are nearer to collecting ducts
    than to descending limbs.

    3.3.4.2  Collecting ducts

         Collecting ducts consists of three identifiable segments, which
    lie, respectively, in the cortex, the outer medulla, and the inner
    medulla. These segments demonstrate different permeabilities to water
    and osmolytes. The difference in permeability may be related to the
    presence of two cell types, the intercalated and collecting duct (or
    principal) cells.

    3.3.4.3  The distal tubule

         The distal tubule connects the thick ascending limb of the loop
    of Henle to that part of the collecting duct which originates in the
    cortex. The distal tubules are in-involved in both ion and water
    reabsorption, but play a much less significant role than the proximal
    tubules. The underlying mechanisms responsible for reabsorption
    appear, in essence, to be similar to those already outlined. The major
    differences include a stronger Na+ gradient against which to "pump",
    the ability to reabsorb sodium without reabsorbing water, the
    controlling effects of anti-diuretic hormone (ADH) and aldosterone
    (among other mediators), and the very limited (or lack of) protein
    reabsorption. The secretion of potassium ions appears to be under the
    control of an active transport mechanism, the regulating factors of
    which are many and complex (Valtin, 1973; Orloff & Berliner, 1973;
    Brenner & Rector, 1986).

    3.3.4.4  The countercurrent multiplier system and urine concentration

         Less than 1% of the glomerular filtrate leaves the kidney as
    urine (unless there is a state of diuresis), the remainder having been
    reabsorbed. The process of urine concentration is complex and depends
    (at least in part) on the countercurrent multiplier system, which
    establishes a steep osmotic gradient along the inner medulla. The high
    osmolality is a consequence of the differential permeability of the
    limbs of the loops of Henle and the collecting ducts to water and
    ions. The thick ascending limb is thought to have an active mechanism
    which transports chloride and sodium out of the lumen and into the
    interstitium, but the limb remains impermeable to water. As a
    consequence the osmolality decreases in this part of the tubule (the
    diluting segment). The descending limb, on the other hand, is freely
    permeable to water, but probably not to sodium ions. The high ion
    concentration in the interstitium would draw water out of the
    descending limb, increasing the osmolality towards the turn of the
    U-loop.  This is augmented by urea and other osmolytes that leave the
    collecting ducts and enter the ascending limb via the interstitium,
    thus being recirculated to the medulla.

         The collecting ducts regulate the final urine concentration by
    controlling the amount of water that is reabsorbed. The passage of
    water out of the ducts is thought to be mediated largely by cyclic
    adenosine-monophosphate (cAMP), the synthesis of which is stimulated
    by ADH, which increases the permeability of the luminal cell membrane
    to water.  Osmotic effects draw the water out of the cell (through the
    basement membrane) into the hyperosmotic interstitium. In the absence
    of ADH the collecting duct is thought to be impermeable and relatively
    little water is reabsorbed from it. The interstitial osmotic gradient
    is assumed to be maintained by the effective removal of water via the
    ascending vasa recta, which have both a greater radius than the
    descending vasa recta and are about twice as numerous. The
    countercurrent exchange associated with the loops of Henle arising
    from cortical nephrons offers an important "barrier" zone, which is
    thought to facilitate solute trapping in and solvent exclusion from
    the inner medulla, and thus helps to maintain the hyperosmolality in
    this "compartment".

         There are a number of other factors that control, alter, or
    contribute to urine concentration. Medullary blood flow is complex, as
    are the factors controlling it. Increased blood flow rates will
    decrease the efficiency of countercurrent exchange in the outer
    medulla, as a consequence of which the high osmotic gradient in the
    inner medullary compartment will be "washed out", and urine will not
    be concentrated. Diuresis is associated with increased blood flow
    rates (Earley & Friedler, 1964, 1965; Chuang et al., 1978).

         A unique feature of the vasa recta is their permeability to
    macromolecules, a consequence of which is that the medulla contains a
    large pool of albumin. The factors controlling the rapid turnover of
    this milieu are poorly understood.  It is generally assumed that
    (together with the glycosaminoglycans) these proteins provide an
    interstitial osmotic pressure that facilitates water reabsorption (see
    Brenner & Rector (1986) for a fuller discussion and list of
    references).

    3.3.4.5  The interstitial cells

         Interstitial cells occur in most organs. Three types of
    interstitial cells have been described in the medulla of the rat
    kidney (Bohman, 1980). Type I cells are the most abundant and
    represent the typical renal medullary cells.  Type 2 medullary
    interstitial cells are generally round and lack lipid droplets, while
    Type 3 cells correspond to the pericytes. Types 2 and 3 are sparsely
    distributed and are often overlooked between the tubules, ducts, and
    blood vessels. In the inner medulla, however, Type I cells are
    numerous and especially prominent because they are set in a dense
    matrix of glycosaminoglycans (previously referred to as
    mucopolysaccharides or acidic mucopolysaccharides).

        The medullary interstitial cells have been described by Moffat
    (1979, 1981, 1982), Bohman (1980), and Maunsbach et al. (1980). The
    number of cells and the amount of matrix substance occupies 10-20% of
    the tissue volume in the outer medulla, and 40% near the apex of the
    inner medulla (Bohman, 1980). The cells, which are arranged in a
    regular pattern perpendicular to the tubules and vessels, are
    irregular in shape and have many long slender processes. These come
    into close contact with adjacent interstitial cells, capillaries, and
    the limbs of the loop of Henle, but there is no such relationship with
    the collecting ducts.

         One of the most characteristic features associated with the Type
    I cells is the presence of lipid inclusion droplets, which occupy at
    least 2-4% of the total cell volume. The lipid content is largely
    triglycerides, with variable amounts of cholesterol esters and
    phospholipids. A number of conditions have been described where there
    are marked changes in the size and number of lipid droplets. The
    pathophysiological significance of these changes is difficult to
    interpret because of varied experimental approaches, species
    variation, and contradictory reports (Bohman, 1980).

    3.4  Species, strain, and sex differences in renal structure and function

         There are important differences between the renal structure of
    animals and man that may have a direct effect on the interpretation of
    toxicological data (Stolte & Alt 1980, 1982; Mudge 1985).  The kidney
    varies greatly in structure and function between different species and
    strains, and there are also more subtle differences between the sexes
    of several animals. In general, the kidneys can be classified into
    those that are multipapillate, such as those of man and the pig, those
    that have more than one papilla (e.g., spider monkey), and those of
    the vast majority of animal species, which are unipapillate. The
    papillae may either be present as a well defined pyramidal structure,
    as in rodents, man, pigs, and dogs, or represent only a ridge as in
    the non-human primates. Furthermore the kidney may be unilobar (and
    have a compact structure) or consist of a multilobar structure, as in
    bovines and elephants.

         Table 7 compares some of the structural and functional features
    of the most species most commonly used in toxicity studies. It
    illustrates clearly that there are a number of differences between man
    and the rodents, which are the most commonly used species to assess
    nephrotoxic potential and study mechanisms of injury. There are also
    major strain differences between Sprague-Dawley, Wistar, and Fischer
    344 rats, which probably account for the vast majority of animals
    studied. The Brown Norway rat is a most useful model for studying
    mercuric chloride immunomediated nephropathy (Druet et al., 1987), and
    the differences between the metabolism of methoxyflurane anaesthetic
    in Fischer 344, Buffalo, Wistar, Long Evans, and Sprague-Dawley rat
    was used as the basis for demonstrating the toxicity of the fluoride
    ion released by hepatic mixed-function oxidase activity (Mazze, 1976,
    1981). In addition there is evidence that there are consistent
    differences between the renal structure and function of male and
    female mice. Other sex differences have been reported in other
    species.

         Species or strains with unique anatomical and functional
    attributes can offer an important way of helping to understand toxic
    mechanisms, but at present there are few published data on non-human
    primates, marmosets, or the pig.

    3.5  Renal biochemistry

         There is little doubt that renal metabolism is coupled tightly to
    specific functions in the kidney.  In particular, reabsorption of
    sodium chloride can be correlated directly with oxygen consumption and
    is probably the most energy-demanding transport function of this
    organ. The regions of the kidney vary in their ability to metabolize
    and produce various substrates.

    3.5.1  Biochemistry and metabolism in the cortex

         The movement of the sodium ions from the tubular fluid to the
    blood is quantitatively one of the the most important functions that
    the kidney performs. This is accomplished by aerobic metabolism linked
    to adenosinetriphosphate (ATP) production and utilization. The exact
    mechanisms are not fully understood.

         Renal cortical nephrons are capable of utilizing a variety of
    substrates, and the substrate utilization varies from nephron segment
    to nephron segment.  For example, Klein et al. (1981) demonstrated
    that the convoluted portion of the proximal tubule utilized succinate,
    glutamate, glutamine and other substrates quite extensively. This same
    nephron segment, however, utilized glucose, lactate, and palmitate
    only minimally. The hexosemonophosphate shunt is present at highest
    activity in the distal segment of the nephron and in the thick
    ascending limb.  Although this pathway may account for relatively
    little glucose oxidation, it would appear important as a source of
    reduced nicotinamide adenine dinucleotide phosphate (NADPH).

         The kidney cortex is also capable of producing glucose from 
    non-carbohydrate precursors.  Although different substrates are
    utilized for gluconeogenesis in the kidney, compared with the liver,
    the gluconeogenic pathways are similar. Changes in hydrogen ion
    activity do not alter hepatic gluconeogenesis, but markedly effect
    that in the kidney.

         Additionally, renal gluconeogenesis is influenced by the
    concentration of substrate in the renal arterial blood.  Guder &
    Schmidt (1974) and Schmidt & Guder (1976) have demonstrated that the
    rate-limiting enzymes for gluconeogenesis are not uniformly
    distributed throughout the nephron. For example, the highest activity
    is found in the proximal convoluted tubule, there being relatively
    little activity in the thick ascending limb. The glycolytic enzymes,
    on the other hand, are present in the thick ascending limb, the distal
    tubule, and the collecting duct.  A specific role for glucose in
    supporting the various renal transport processes has not been
    adequately described. Renal phospholipid metabolism, however, may be
    important in support of transport and may play a direct role in sodium
    movement.

        Table 7.  Comparison between the renal structure and function in man
                     and in commonly investigated speciesa
                                                                                         
                                                          Man      Rat    Dog     Pig
                                                                                         
    Cortical structure

    Nephrons per g body weight                             16      128      45      26
    Glomerular radius (µm)                                100       61      90      83
    Proximal tubular length (mm)                           16       12      20      30
    Tubular radius (µm)                                    36       29      33      35

    Cortical function

    Glomerular filtration rate (ml/min per m2)             75       35     104      72
    Inulin clearance (ml/min per kg body weight)            2.0      6.0     4.3     2.1
    p-Aminohippurate transport maxima
    (mg/min per kg body weight)                             1.3      3.0     1.0      -

    Drug-metabolizing enzymesb

    Mixed-function oxidase                                  4        5        -       -
    NADPH-cytochrome c reductase                           15       48        -       -

    Medullary structure

    Number of papilla                                      15-20     1       1        6-10
    Percent long loops                                     14       28     100        3
    Relative medullary thickness                            3.0      5.8     4.3      1.6

    Medullary function

    Maximum urine osmolality
    (mOsmol/kg)                                           1400     2610    2610     1080
                                                                                         

    a     Data from Mudge, 1985; Stolte & Alt, 1980, 1982; Gyrd-Hansen, 1968
    b     Renal enzyme activity expressed as a percentage of liver activity

    -    No data published on this parameter
    
    3.5.2  Biochemistry and metabolism in the medulla

         Guder & Ross (1984) have highlighted the biochemical aspects of
    heterogeneity along the nephrons. Much of the published information
    has been derived from whole medulla or medullary slices and fails to
    differentiate between the metabolic contribution from the nephrons
    (loops of Henle), as opposed to the collecting duct epithelia, versus
    the interstitial cells.

    3.5.2.1  The biochemistry of renal prostaglandins (PG)

         The PGs and endoperoxides are a group of ubiquitously distributed
    hormones with a broad spectrum of potent biological activity that
    shows marked receptor specificity. They are synthesized from the C20:4
    fatty acid arachidonic acid by an enzyme system (which includes
    cyclo-oxygenases, peroxidases, isomerases, and reductases)
    collectively called PG synthetase.

         The PGs (Fig. 9) are structurally similar and are only present in
    minute concentrations. Several are labile and undergo spontaneous
    chemical changes. Thus, most of the methods (both qualitative and
    quantitative) needed for their biochemical investigations are fraught
    with subtle pitfalls (Frolich & Walker, 1980). The literature on renal
    PG biology is large, complex, contradictory, and difficult to
    interpret. The subject has been reviewed recently by Dunn & Hood
    (1977), Dunn & Zambraski (1980), Horrobin (1980), Morrison (1980),
    Zusman (1980), Dunn (1981), Frolich et al. (1981), and Levenson et al.
    (1982).

         PGs are not stored in renal tissue but are synthesized  de novo
    from arachidonic acid, which is released from stored phospholipid or
    triglyceride pools by the action of phospholipase A2.  The factors
    that regulate the release of arachidonic acid include both
    receptor-mediated responses (such as vasoactive peptides and biogenic
    amines) and non-specific stimuli (ischaemia). The prostaglandin
    precursor may be drawn from different lipid pools. Any arachidonate
    that is not channelled into prostaglandin synthesis may be re-acylated
    (as are the  de novo synthesized molecules) or disposed of via
    several other metabolic routes. Arachidonic acid (the availability of
    which is rate limiting) is converted to PGG2 and thence to other
    PG-related substances.

         The anatomically identifiable areas of the kidney each synthesize
    a different pattern of PGs  in vitro.  The  in vivo contributions of
    each area to PG synthesis and the function of each PG remains largely
    a matter of speculation at present. Total PG synthesis is several
    times higher in the medulla (where typically it is greatest in the
    papilla) than in the cortex (Dunn & Hood, 1977). However, the
    distribution of, for example, PGE2 synthesis reflects a more complex
    picture, its concentration being lower in the papilla than the rest of
    the inner medulla (van Dorp, 1971). A recent study using isolated
    nephrons showed that the highest PGE2 production is in the medullary

    FIGURE 9

    collecting tubules, followed by cortical collecting ducts and
    glomeruli. Furthermore, there are marked sex-related differences in
    the effects of cofactors on medullary PG synthetase activity (Hirafuji
    et al., 1980). Some PGs break down spontaneously (e.g., PGI2 to
    alpha-6-keto-PGF), but the majority are metabolically degraded
    (Morrison, 1980). The enzymic conversions are mediated by a number of
    enzymes, including dehydrogenases, reductases, and ß- and 
    omega-oxidases. The enzymes that degrade PGs are located mainly in the
    cortex, but there are species differences in the corticomedullary
    ratio of these enzymic activities (Powell, 1980).

         The factors regulating the biosynthesis of each type of PG are
    only poorly defined (Horrobin, 1980). A large number of endogenous and
    exogenous substances have been reported to alter renal PG synthesis,
    and several patho-physiological conditions have been described in
    which renal PG synthesis is increased. Most attention has been
    focussed on the inhibitory effects of the anti-inflammatory drugs. The
    steroidal compounds (e.g., corticosteroids) prevent the release of
    arachidonic acid from its lipid  pools, and the non-steroidal 
    products (e.g., indomethacin) inhibit cyclo-oxygenase. It is, however,
    essential to be aware that any factor that perturbs PG synthesis may
    act differently at different sites in the synthetic (or degradative)
    pathway.

         Indomethacin (one of the most extensively studied cyclo-oxygenase
    inhibitors) produces several alterations in renal PG dynamics.
    Uncertainties in defining PG-"related" pathophysiological changes are
    compounded by the observation (Attallah & Stahl, 1980) that PGE2
    synthesis in slices from each zone of the kidney has a different dose
    response to indomethacin inhibition: the cortex is most sensitive and
    the papilla least sensitive. The cyclo-oxygenase inhibitors are
    generally classified as either reversible or irreversible, but the
    multiplicity of effects and the possibility of enzymic polymorphism in
    different regions of the kidney suggest that such a classification may
    be an over-simplification.

         The exact physiological roles of the PGs in normal renal function
    and the way in which these are altered in the development of
    nephropathies are not clear. Firstly, indomethacin, for example, has
    been shown to cause biochemical changes that may be classified as
    either related or unrelated to altering PG dynamics. Secondly, many
    attempts to define renal PG function have been based on the hypothesis
    that urinary PG excretion reflects  de novo renal  synthesis (Dunn &
    Hood, 1977; Dunn & Zambraski, 1980; Dunn 1981), notwithstanding
    analytical difficulties of measuring very low levels of various PGs
    and apparently ignoring the fact that  de novo synthesized PGs may
    have undergone extensive degradation. The measurement of urinary PGs,
    as an estimate of their  de novo renal synthesis, remains equivocal
    because, firstly, seminal PGE2 is an unavoidable and variable
    contaminant in the urine of males (Suzuki et al., 1980) and, secondly,
    Brown et al. (1980) have demonstrated that both rabbit and rat urinary
    bladders can synthesize PGE from arachidonic acid. Finally, the

    physiology of renal function is controlled by several hormonal
    systems, the detailed functioning of which has not been clearly
    established. It is known that renal PGs may be altered by (or may
    alter) the renin/angiotensin II/aldosterone system (Hackenthal et
    al., 1980; Lee, 1980; Weber, 1980; Baer, 1981), the kallikrein-kinin
    system (Margolius, 1980; Rockel & Heidland, 1980), and the regulation
    of fluid balance and water reabsorption via ADH (Blair-West et al.,
    1980). Furthermore, each of these hormonal systems may interact with
    the others via direct or indirect mechanisms. It seems likely that a
    full understanding of the pathophysiology of the renal hormonal
    systems will take some time to crystallize.

         In spite of the rather abstruse biology, there is general
    consensus (Dunn & Hood 1977; Dunn & Zambraski, 1980; Morrison, 1980)
    that PGs have a central role in renal function. It seems, however,
    that renal PGs play little, if any, major regulatory role in basal
    renal blood flow in normal conscious animals. There is evidence that
    PGs are released in response to ischaemic and vasoconstrictive stress,
    where their role seems to be to provide a protective effect by
    maintaining glomerular dynamics. The role of PGs (especially PGE2) 
    in preventing experimentally induced acute renal failure is
    conflicting.  Arachidonic acid does stimulate renin release, a
    response that is blocked by cyclo-oxygenase inhibitors, but it remains
    uncertain which of the PGs mediate this effect  in vivo.  It is also
    unclear from which renal zone such mediators are synthesized and
    released. Renin release may, in turn, affect PG synthesis and the
    kallikrein-kinin system (which in turn may modulate PG synthesis and
    the renin system). ADH is assumed to stimulate PGE2,  but published
    data on the controlling effects of PGs on salt and water balance are
    very difficult to interpret. Similarly, the mass of literature on
    hypertension and PGs favours the concept that the two are related, but
    fails to propound a unifying hypothesis.

    3.5.2.2  Lipid metabolism

         The contents of the interstitial lipid droplets are too
    specialized to be used as a metabolic energy store (Bohman, 1980),
    although this droplet population undergoes marked and rapid change
    during the short periods that precede various pathophysiological
    conditions. The interstitial cells produce the ground substance matrix
    that surrounds them. Early evidence that the interstitial cells of the
    renal medulla are only a highly specialized PG-producing cell type has
    become equivocal. The interstitial lipid droplets do not, in fact,
    provide the sole source of arachidonic acid for PG synthesis. Only 50%
    of the medullary capacity for synthesizing PG is confined to the
    interstitial cells; the rest is in the collecting ducts (Bohman,
    1980). The significance of PG synthesis in the medullary cells cannot,
    however, be overlooked, as it may play an important role in regulating
    blood pressure and other renal functions. It has also been suggested
    to occupy a central position in the pathogenesis of renal papillary
    necrosis.

         In recent years the importance of the endocrine function of the
    medullary interstitial cells in regulating blood pressure has been
    highlighted by several workers (Mandal & Bohman, 1980). Three groups
    of vaso-active compounds have been isolated from the medulla or
    cultured interstitial cells. Experimentally induced, spontaneously
    occurring, and pathologically precipitated hypertensive states have
    been reversed by subcutaneous transplants of renal papillary fragments
    and by cultured interstitial cells.  In addition, the systemic
    administration of both the  polar and neutral reno-medullary lipids
    reduces arterial blood pressure (Muirhead & Pitcock, 1980).

         The numerous lipid droplets in the interstitial cells have been
    found to contain traces of cholesterol esters, a few percent of
    phospholipids, mainly phosphatidylcholine, and, rarely, trace amounts
    of phosphatidyl-ethanolamine. A few percent of free fatty acids and
    tri-acylglycerols make up the remaining 80-90%, the composition of
    which varies in different species. The most striking features are the
    varied types and large amounts of unsaturated fatty acids; most
    notably those of 20 or more carbon atoms such as arachidonic acid and
    especially adrenic acid. The large amount of arachidonic acid suggests
    that the interstitial lipid droplets may be an important pool for PG
    synthesis in the kidney (Bojesen, 1974).

    3.5.2.3  Carbohydrate metabolism in the medulla

         The metabolism of carbohydrate in the renal medulla has been
    reviewed by Cohen (1979). Some early observations suggested that the
    low oxygen tension in the inner medulla (a pO2 as low as 0.67-2.0
    kPa (5-l5 mmHg) compared with 10 kPa (75 mmHg) in the cortex) would
    necessitate anaerobic metabolism. However, aerobic metabolism is only
    limited at an O2 availability of less than 0.13 kPa (1 mmHg).

         Many metabolic investigations have used medullary slices or
    homogenates. Thus, the exact contribution of the different cell types
    in the inner medulla to functional energy dynamics and to the changes
    that underlie, for example, diuresis or anti-diuresis have yet to be
    related to the phosphorylation and redox states within these
    individual cell types.

         Carbohydrates are stored in the medulla as either glycogen or as
    glycosaminoglycan (GAG), the former in collecting ducts and epithelia
    and the latter as an important constituent of the interstitial ground
    substance. There is evidence to suggest that either can be mobilized
    to provide an energy source or the glucose units for the synthesis of
    the other macromolecular carbohydrates (Darnton, 1967, 1969).

    3.5.2.4  Medullary glycosaminoglycan (GAG)

         The biology of GAGs has been described by Kennedy (1979). GAGs
    are linear polysaccharides that are made up of repeating disaccharide
    units, one carbohydrate moiety of which is a hexuronic acid (or a
    neutral sugar in one case) and the other a hexosamine. The nature of
    the disaccharide units and the occurrence of  N-acetyl groups,
    together with the position of  O-sulfate groups, define the species
    of macromolecule. There are seven basic types of GAG. These molecules
    also show heterogeneity of relative molecular mass (when isolated from
    the same or different organs; Toledo & Dietrich, 1977), and the molar
    ratio of sulfate to hexosamine varies by up to 2-fold for the same
    type of GAG (Suzuki et al., 1976). Most of these substances probably
    occur  in vivo as proteoglycans (PoGs). These supramolecular
    structures are composed of a linear protein backbone that carries GAGs
    covalently bound at intervals along its length. In theory, any
    combination and ratio of GAGs may occur. It is only recently that the
    concept of PoGs has been accepted; before this the presence of protein
    was assumed to be a contamination and vigorous steps were taken to
    remove it.

         In spite of the ubiquity of PoGs and their composite GAGs,
    relatively little is known about their physiological functions, with
    the exception of their anti-coagulant and anti-lipaemic properties,
    which are best studied in heparin. These molecules are bound to cell
    surfaces (Kjellen et al., 1977), where they may control the access of
    endogenous  and exogenous molecules to cell membrane receptors.
    Similarly, the functions of this intercellular polyanionic matrix most
    probably extend beyond that of "immobilized anti-coagulants" or "space
    filling", and include controlling the micro-environment of cells (by
    binding either inorganic or organic cations and by their immense
    water-holding capacity) and regulating cell-cell communications. GAGs
    may also control cell recognition and adhesion, and contribute to the
    control of cell movement, growth,  differentiation,  and proliferation 
    (Long & Williamson, 1979). The association of GAGs with mitochondria
    and nuclear membranes suggests that these macromolecules may also play
    a direct role in controlling some intracellular functions.

         The distribution of GAGs has been assessed in tissue either by
    the autoradiographic distribution of precursor carbohydrates or
    35SO2 or  by histochemical staining. It is generally assumed that
    sulfate radiolabel distribution is relatively specific for GAGs.
    However, most of the staining procedures are nonspecific (e.g.,
    toluidine blue interacts with any polyanion to give a metachromatic
    colour shift) and depend either on  a priori knowledge of
    distribution or, for example, the use of control sections that have
    been exposed to selective enzymic digestion.

         The amount and types of GAGs in the kidneys of various species
    have been reported. The quantity of polyanionic macromolecule has been
    found to be greater in the medulla than in the cortex for the rat
    (Jacobsen et al., 1964; Kresse & Grossmann, 1970), pig (Kresse &
    Grossmann 1970), dog (Castor & Green, 1968; Kresse & Grossmann, 1970),
    and normal human kidney (Inoue et al., 1973; Constantopoulos et al.,
    1973). The medulla:cortex ratio in the human kidney was found to be
    age related, increasing rapidly to a maximum in the fourth decade and
    then declining slowly (Inone et al., 1973). The heterogeneous
    distribution of the types of GAG in the kidney is supported by the
    data of Constantopoulos et al. (1973) for the human kidney and Castor
    & Green (1968), who reported that hyaluronic acid had a high relative
    molecular mass in the medulla but a low one in the cortex of dogs.
    However, other data on the dog, pig, and sheep (Dicker & Franklin,
    1966) and the rat (Barry & Bowness, 1975) suggest that the types and
    quantities of GAG are the same in both the cortex and medulla.

         The processes underlying and controlling the biosynthesis of PoGs
    are complex and incompletely documented (Kennedy, 1979). Muirhead &
    Pitcock (1980) reported that medullary interstitial cells synthesize
    PoGs (both  in situ and in culture) and that these macromolecules are
    associated with the cellular cisternae (dilated rough endoplasmic
    reticulum). Darnton (1967, 1969) presented data to show that glycogen
    associated with the epithelial cells of the collecting duct in the
    rabbit is mobilized and incorporated into GAGs.

         The functions of the medullary GAGs have been the centre of
    controversy since Ginetzinsky (1958) suggested that the action of ADH
    was mediated by the release of hyaluronidase. This would depolymerize
    medullary GAG and (so it was argued) allow greater water reabsorbtion
    from the tubules into the interstitium and thence to the blood supply. 
    This hypothesis has been supported by some workers (Jacobson et al.,
    1964; Farber et al., 1971) but refuted by others (Sun et al., 1972;
    McAuliffe 1978, 1980; Sun, 1980) in animals with spontaneous diabetes
    insipidus. These conflicting data are difficult to resolve into a
    single unifying theory relating the physiological function of GAGs to
    the urine-concentrating process.

    3.6  The metabolism of xenobiotic molecules in the kidney

         Chemically induced lesions may depend to varying extents on the
    metabolic capacity of tissues to deal with "insults". The metabolism
    of xenobiotic molecules may either prevent lesions (by deactivation),
    or be directly responsible for damage (by bio-activation). The renal
    metabolism of chemicals (and its consequences) has been reviewed by
    Hook et al. (1979), Anders (1980), Connelly & Bridges (1980), Kluwe &
    Hook (1980), Davis et al. (1981), Rush et al., (1984) and Tarloff et
    al. (1987).

         It is likely that the liver meets the challenge of metabolizing
    a major proportion of exogenous compounds  in vivo before they reach
    the systemic circulation. Most fundamental types of bioconversion have
    been described for the perfused kidney of several species (Szefler &
    Acara, 1979; Elbers et al., 1980; Ross et al., 1980; Emslie et al.,
    1981) and for isolated renal cells and tubular fragments (Fry et al.,
    1978; Jones et al., 1979; Ormstad, 1982).  Similarly, kidney
    microsomes have been shown to have most of the enzymic and
    cytochrome-mediated metabolic activities that have been described in
    other tissues. The xenobiotic-transformating capacity of the kidney is
    about 3-50% (depending on the system, species, and source of data) of
    that found in the liver (Litterst et al., 1975a, Navran &
    Louis-Ferdinand 1975; Fry et al., 1978), but it may be much higher
    than that of the liver under certain circumstances (Anders, 1980).
    There are marked qualitative differences between hepatic and renal
    xenobiotic metabolism. Renal enzymes are stable during the Ca2+ 
    method of preparing microsomes (Litterst et al., 1975b). Enzymic
    kinetic constants vary between microsomes isolated from the two organs
    (Navran & Louis-Ferdinand, 1975). Whereas there are marked sex-related
    differences in hepatic metabolism, there are few in the kidney
    (Litterst et al., 1977). There is evidence to suggest that liver
    cytochrome P-450 is similar to that of the kidney, based on
    electro-phoretic and electron paramagnetic resonance  studies
    (Armbrecht et al., 1979), immunological criteria (Guengerich & Mason,
    1979), and on immunometabolic studies (Kaminsky et al., 1979). 
    However, these data are most difficult to interpret in "absolute"
    terms, because the samples of cytochrome P-450 were from organs
    exposed to different inducing agents. There is now substantial
    evidence that hepatic and renal tissue respond differently, both
    quantitatively and qualitatively, to the various inducers of
    cytochrome P-450 (Litterst et al., 1977; Zenser et al., 1978a;
    Kaminsky et al., 1979). Ascorbic acid deficiency (Sikic et al., 1977)
    and carbon tetrachloride pretreatment (Litterst et al., 1977) alter
    the metabolism of xenobiotics in a different way in the liver and
    kidney. In addition, the inhibitory effects of
    2-diethylaminoethyl-2,2-diphenylvalerate (SKF-525A) on renal and
    hepatic microsomes studied  in vitro are similar but not identical
    (Litterst et al., 1977).

         There are several enzymes involved in renal xenobiotic
    metabolism. It is not possible to comment on all of those that may be
    relevant to nephrotoxicity nor, indeed, is it clearly established what
    role each renal enzyme plays in the realization of the potential
    toxicity of a chemical. Many of the enzymes that metabolize
    xenobiotics are compartmentalized in specific regions of the kidney.
    The anatomical localization of these characteristics may play a key
    role in the toxicological consequence that follows the entry of a
    xenobiotic into the kidney. The distribution and regulation of these
    enzymes may predispose to the toxic effects of chemicals. Thus,
    although intrarenal metabolism may be a prerequisite for the target
    selective effects of some chemicals, the final outcome of the toxic
    response relates to the sum of a number of factors. These include the

    localization of those renal enzymes involved in xenobiotic metabolism
    (this may be metabolic activation or other processes, perhaps in an
    adjacent cell) and the processes controlling the intracellular
    concentration of toxic chemicals.  The intracellular concentration of
    a chemical can be influenced by xenobiotic metabolism  per se and by
    many of the inherent processes in the kidney, such as transport, pH,
    and solute gradients on either side of a membrane. The outcome of a
    chemical exposure may also be affected by the numbers and types of
    organelles in a specific cell type that have critical properties
    relevant to the functions of that cell and by the presence of a
    protective mechanism in a particular cell type (such as antioxidants,
    free radical scavengers).  Most of the processes that underlie
    nephrotoxicity are probably multi-step events that are affected by
    more than one metabolic pathway and occur via competing and sequential
    pathways. Little is known about the control of these pathways, so that
    it is difficult to predict from the structure of a chemical alone what
    effects it will have on the kidney. In addition, a major role is
    obviously played by the extra-renal metabolism (in the liver, lung,
    gut, etc.) of the parent chemical and by a variety of other organ
    functions. These include the lung (exhalation of volatile
    metabolites), liver (biliary excretion), and gastrointestinal
    microflora (enterohepatic circulation, serum protein binding), and
    they determine the types of chemicals, the concentration that reaches
    the kidney, and their renal pharmacokinetics.

    3.6.1  Oxidases

         Oxidases can convert chemicals into active intermediates or
    generate reactive species by redox cycling. This is potentially
    important for compounds that contain arylamine, quaternary bipyridyl
    (paraquat), quinone (adriamycin), or nitro (nitrofurantoin)
    structures. It is generally accepted that, in the liver, biologically
    reactive intermediates mediate their toxic effects by binding to
    cellular macromolecules and blocking normal functional processes
    (Jollow et al., 1976; Snyder et al., 1981). Similar mechanisms have
    been proposed to explain various types of chemically induced renal
    lesions, including papillary necrosis. The metabolically generated
    reactive intermediates have a relatively short life and are most
    likely formed in the organ or anatomical area in which they induce
    damage.

    3.6.1.1  Cytochrome-P-450-dependent mixed-function oxidases (monooxygenases)

         This enzyme system carries out a two-electron flow pathway, and
    the flavoprotein component can catalyse single electron reductions
    such as the reduction of quinones to semiquinone radicals (Bachur et
    al., 1979). Multiple forms of cytochrome P-450 have been identified in
    the kidney. These include phenobarbital-, 3-methylcholanthrene-, and
    ß-naphthoflavone-inducible cytochrome P-450. Renal cytochrome P-450
    induction varies in different species. Polycyclic aromatic
    hydrocarbons induce cytochrome P-450 in most species, whereas
    phenobarbital is effective in hamsters and rabbits but not in

    guinea-pigs, rats, and mice (Smith et al., 1986).  Similarly, renal
    P-450 responds differently to inhibitors of mixed-function oxidases.

         The effects of inhibitors such as SKF-525A are further
    complicated by multiple actions on renal transport, intracellular
    binding of chemicals at noncatalytic sites, and on cytochrome
    P-450-dependent metabolism. Other inhibitors do not appear to have
    been as fully studied, and the paucity of data in this area makes
    other studies on the effects of inhibitors most difficult to
    interpret. Some species also have sex-related differences, e.g., male
    mice have higher concentrations and activities of P-450 than female
    mice (Krijsheld & Gram, 1984; Smith et al., 1984; Hawke & Welch,
    1985), but this is not the case for rats (Litterst et al., 1977; Hook
    et al., 1982) or rabbits (Litterst et al., 1977). There is no clear
    data on other species, such as man, nor on how different types of
    renal disease affect the concentration of renal P-450 or its induction
    or inhibition.

         The specific activity of the renal mixed-function oxidases varies
    widely between species and is about 10% of the hepatic activity
    (Zenser et  al., 1978a,b; Endou, 1983). This suggests a role for renal
    P-450 that is quantitatively less important than that of the liver.
    However, this is not the case for all chemicals, since the renal
    metabolism of chloroform is about 2-fold higher than the hepatic
    activity (Smith & Hook 1984). More importantly, mixed-function 
    oxidase  activities are intra-renally localized to discrete areas
    where their significance in metabolism may be far greater than in the
    liver.  The S2 proximal segment has a cytochrome P-450 concentration
    that is 2-3 times higher than the S1 or S3 segments. The distal
    tubules, cortical collecting ducts, and the medulla contain no
    measurable cytochrome P-450 activity (Endou, 1983).  By contrast
    NADPH-cytochrome-P-450 reductase activity is highest in the S2 and
    S3 segments, but it also extends to the distal tubule and medullary
    structures.

    3.6.1.2  Prostaglandin peroxidase-mediated metabolic activation

         Recently, it has been shown that there are marked quantitative
    and qualitative differences in the regional distribution of microsomal
    mixed-function oxidase activity within  the rabbit kidney (Zenser  et
    al., 1978a,b; Armbrecht et al., 1979). Most mixed-function oxidase
    activity is located in the cortex and least in the inner medulla in
    control tissue and in that taken from animals induced with
    3-methylcholanthrene. Cytochrome P-450 is not detected in the medulla
    of controls or even those of induced animals. In addition, laurate
    hydroxylase activity (the only mixed-function oxidase activity found
    in the inner medulla) shows marked differences in the pattern of
    inhibition by carbon monoxide, alpha-naphthoflavone, and metyrapone in
    the cortex and the outer and inner medulla. This suggests differences
    in the genetic expression of the same type of enzymic activity in
    different zones of the kidney.

         Davis et al. (1981) showed that oxidative metabolism in the
    medulla is mediated in the absence of spectrophotometrically
    measurable cytochrome P-450. Zenser et al. (1979a) reported that
    cortex microsomes metabolized 1,3diphenylisobenzofuran to
     O-dibenzoylbenzene largely via a cytochrome P-450-like system (it
    was NADPH dependent and inhibited by carbon monoxide and metyrapone).
    The inner medulla microsomes had the same metabolic capacity in the
    presence of arachidonic acid (the system was independent of NADPH, and
    it was inhibited by non-steroidal antiinflammatory compounds such as
    indomethacin and not by carbon monoxide or metyrapone). The outer
    medulla microsomes  had both types of  activity. The antioxidant
    ethoxyquin inhibited the arachidonic acid and the NADPH-dependent
    metabolic processes.

         The specific arachidonic-acid-dependent PG
    cyclooxygenase-mediated metabolism of benzidine has been shown to be
    absent from hepatic and renal cortical microsomes but active in
    medullary microsomes, especially those from the inner medulla. The
    metabolism was inhibited by nonsteroidal anti-inflammatory drugs,
    ethoxyquin, and arachidonic acid analogues. Approximately 75% of
    metabolized benzidine was covalently bound to macromolecules,
    presumably via a reactive intermediate. Addition of sulfhydryl
    protectors, such as glutathione, reduced the amount of covalently
    bound metabolite to 25% (Zenser et al., 1979b,c). Using rabbit renal
    inner medullary slices, Rapp et al. (1980) have confirmed the
    arachidonic-acid-dependent co-oxidative activation of low
    concentrations of benzidine, and its covalent binding to tissue.
    Mohandas et al. (1981a,b) examined the activation of paracetamol
    (acetaminophen) and the covalent binding to protein in the inner
    medulla, outer medulla, and renal cortex, and compared this to the
    situation in the liver. The arachidonicacid-dependent pathway showed
    a ten times greater degree of activity in the inner medulla compared
    to the renal cortex, intermediate activity being found in the outer
    medulla. In the liver the arachadonic-acid-dependent pathway activity
    was approximately 50% of that of the renal cortex. The total
    activation of paracetamol (acetaminophen) by arachadonic-acid- and
    NADPH-dependent pathways in the renal cortex and the liver was
    essentially the same.

    3.6.2  Conjugation

         Conjugation takes place on existing groups or those produced by
    oxidation, and greatly increases the polarity of compounds. This
    facilitates their elimination and generally terminates any
    pharmacological activity. There are several examples, however, where
    conjugation may give rise to reactive compounds (e.g., the
    glucuronides of  N-hydroxy-2-acetylaminofluorene and
     N-hydroxyphenacetin are potently toxic). Similarly, glutathione
    conjugates may be toxic (section 6.3.2.1).

    3.6.2.1  Glucuronide conjugation

         Glucuronide conjugates are formed by the action of uridine
    diphosphate (UDP) glucuronyl transferase. This enzyme has at least
    three isozymes, each of which preferentially conjugates different
    types of molecules. Only UDP-GT1 occurs in the rat kidney, where the
    substrates include planar compounds such as 1-naphthol and
    4-nitrophenol. UDP-GT1 activity is increased by 3-methylcholanthrene.
    Human kidneys have UDP-GT1 and high GT2 activities, while rabbit
    kidneys have all three isoenzymes. UDPglucuronyl transferase activity
    is highest in the cortex, and the distal tubule activity is about 50%
    of that found in the proximal tubule (Cojocel et al., 1983). The
    enzyme is also measurable in rat kidney medulla, but here its activity
    may be limited by the availability of the cosubstrate, UDP-glucuronic
    acid.  Renal glucuronidation capacity may be comparable to or greater
    than that of the liver, depending on the substrate, and microsomes
    from female rats form considerably more glucuronide conjugates than
    those from male rat kidneys.

    3.6.2.2  Sulfate conjugation

         Sulfotransferases form highly polar and, therefore, rapidly
    excreted sulfate esters. The concentrations of both sulfotransferase
    and activated sulfate are higher in the renal cortex than in the
    medulla, and renal sulfotransferase activity is markedly lower than
    that of the liver. The capacity to synthesize sulfate conjugates is
    not increased by standard inducers.

    3.6.2.3  Glutathione conjugation

         Glutathione is the most abundant thiol-containing peptide in the
    kidney, where it is synthesized in the proximal tubule and provides a
    scavenger for detoxifying electrophilic radicals formed from alkyl and
    aryl halides, epoxides, and alkenes (Ormstad, 1987). These compounds
    are degraded to the cysteine conjugate and are generally excreted as
    the  N-acetyl-cysteine conjugate. Glutathione may also have ligand
    binding and transport properties, and it has been suggested that it is
    an important carrier in the transfer of amino acids from the extra- to
    the intracellular space. Glutathione  S-transferase plays an active
    role in metabolism, where it catalyses the initial step in glutathione
    conjugation of halogenated aromatics, epoxides, halogenated alkyls and
    aralkyls, and alpha,ß-unsaturated compounds (Reed & Beatty, 1980),
    drugs such as paracetamol (acetaminophen), and endogenous substrates
    such as estrogen and PGs (Moldeus et al., 1978; Jones et al., 1979;
    Kaplowitz, 1980).

         The total renal GSH  S-transferase activity per g wet tissue is
    considerably less than the corresponding hepatic activity (Hales et
    al., 1978). There are sex differences in the renal GSH  S-transferase 
    activities in rats, aralkyl, epoxide, and alkyl transferase activities
    being lower in males than in females. Rat kidney glutathione
     S-transferases consist of three distinct proteins. One is identical
    to hepatic transferase B (ligandin), a second conjugates
    alpha,ß-unsaturated substrates similarly to the hepatic enzyme, and
    the third, renal transferase, is unique to the kidney and active with
     p-nitrobenzyl chloride (Hales et al., 1978).

         Renal GSH  S-transferases are under complex hormonal control. 
    Hypophysectomy in male rats significantly increases GSH  S-aryl, 
    aralkyl, and epoxide transferase activities without altering GSH
     S-alkyl and alkene activities. GSH  S-transferase can be induced by
    a number of chemicals, but the profile of activities affected is
    complex.  Phenobarbital fails to induce cytochrome P-450 activity  in
    rat kidney, but increases GSH  S-aralkyl transferase activity without
    affecting GSH  S-alkyl, aryl, and epoxide transferases.
    3-Methylcholanthrene (3MC) induces renal GSH  S-aryl and aralkyl
    transferase activities but not GSH  S-alkyl or epoxide transferases
    (Clifton et al., 1975; Chasseaud, 1980).

    3.6.2.4  Mercapturic acid synthesis

         The formation of glutathione conjugates is the first in the
    pathway of renal metabolism to mercapturic acid. The enzyme
     gamma-glutamyl transpeptidase is localized on the brush border of
    the proximal tubule, where it cleaves the  gamma-glutamyl linkage of
    glutathione to produce the cysteinyl-glycine conjugate in the tubule
    lumen. This metabolite is a substrate for a number of peptidases that
    produce the cysteinyl conjugate, which in turn is converted to the
    mercapturic acid by microsomal  N-acetyltransferase (Green & Elce,
    1975). The cells of the proximal tubule in the outer medulla produce
    the  N-acetyl-cysteine conjugate of paracetamol (Jones et al., 1979).

         Glutathione conjugation is generally a detoxification pathway,
    but some compounds may undergo bioactivation by ß-lyase (localized in
    the outer mitochondrial membrane and the cytoplasm of the proximal
    tubule). This enzyme is now known to be capable of cleaving the C-S
    bond, leaving a reactive intermediate.

         Extrarenal biotransformations are now known to produce substrates
    for renal enzymes, which convert these metabolites into reactive
    intermediates that cause target selective toxicity.
    Hexachloro-1,3-butadiene (HCBD) is metabolized in the liver probably
    by GSH  S-transferasecatalysed halogen substitution rather than a
    cytochrome P-450-mediated reaction to the GSH conjugate. In the
    process, hepatic but not renal GSH is depleted in the male rat,
    whereas GSH decreases in the female rat kidney. The HCBD-GSH conjugate

    may be transported to the  bile, returned to the bloodstream via
    intestinal reabsorption, and excreted via the kidneys (Nash et al.,
    1984). The cysteine conjugate of HCBD,  S-pentachlorobuta-1,3-dienyl
    cysteine, causes the same lesion. The enzyme ß-lyase is present in
    both the liver and kidney. The unique renal susceptibility to the
    HCBD-GSH metabolite appears to be related to accumulation via the
    organic anion transport, as this is inhibited by probenecid both
     in vivo and  in vitro (Fig. 10).

         Some alkylhalides, such as 1,2-dibromoethane (Lock, 1987) and
    1,2-dibromo-3-chloropropane (DBCP) (Dybing et al., 1989), may form
    reactive, nephrotoxic intermediates (presumably episulfonium ions)
    following conjugation with glutathione without further metabolism by
    ß-lyase.  1,2-Dibromoethane and DBCP are metabolized in the liver by
    both cytochrome P-450 oxidative dehalogenation and by
    glutathione- S-transferase-mediated  substitution. The renal cortex
    of the rat contains substantial quantities of glutathione
     S-transferase  with a high activity towards 1,2-dibromoethane (Lock,
    1987). In the male rat kidney, studies with perdeutero-DBCP indicate
    that DBCP is not metabolized by cytochrome P-450, but presumably by
    glutathione  S-transferase.  Furthermore, inhibitors of
     gamma-glutamyl transpeptidase and ß-lyase do not affect DBCP-induced
    renal tubular necrosis (Omichinski et al., 1987; Dybing et al., 1989).

    3.6.2.5  Amino acid conjugation

         The amino acid glycine forms a glyco-conjugate via the activation
    of a carboxylic acid (e.g., salicylic acid) by coenzyme A. Salicyl-CoA
    and glycine are co-substrates for acyl-CoA-glycine-
     N-acetyltransferase, which  catalyses the condensation to
    salicyluric acid. Glycine  N-acetyltransferase activity is present in
    the kidneys of rabbits, monkeys, and humans (Bekersky et al., 1980).
    The kidney may be a major site for the metabolism of benzoic acid to
    hippuric acid,  p-aminobenzoic  acid to  p-aminohippuric acid, and
    salicylate to salicyluric acid (Wan & Riegelman, 1972a,b; Wan et al.,
    1972). The isolated rat kidney perfused with glycine and salicylic
    acid excretes 3-4% of the salicylic acid as the glycine conjugate (Wan
    & Riegelman, 1972a,b; Wan et al., 1972; Bekersky et al., 1980). These
    reactions are reversible in the kidney. About 20% of salicyluric acid
    is converted to salicylic acid by the isolated perfused rat kidney,
    whereas the liver does not deconjugate salicyluric acid to salicylate. 
    The freshly deconjugated salicylate is more rapidly excreted than the
    parent salicylate. Possibly the diffusion of salicylate into the renal
    cell is the rate-limiting step of elimination. This is not the case
    for salicyluric acid, which is converted to salicylate within tubular
    cells (Bekersky et al., 1980).

    FIGURE 10

         Other aromatic acids (such as benzoic acid or  p-aminobenzoic 
    acid) competitively inhibited glycine conjugation. Cellular glycine is
    limited, and so its use in conjugation would be saturated with
    increasing concentrations of substrate.

    3.6.3  Other enzymes involved in xenobiotic metabolism

         Highly electrophilic epoxides play a key role in tissue
    alkylation of macromolecules, especially nucleic acids, leading to
    mutagenic and carcinogenic effects. Epoxide hydrase (hydrolase)
    converts aliphatic and aromatic epoxides to their trans-hydrodiols, a
    process that may generate both toxic and non-toxic products (Anders,
    1980).

         The oxidation of aldehydes to ketones or carboxylic acids in the
    kidney is mediated by aldehyde oxidase and aldehyde dehydrogenase
    (Goldberg & Anderson, 1985). Renal aldehyde oxidase activity is
    localized in the proximal tubule but represents only 40% of the liver
    activity (Anders, 1980), but aldehyde oxidase represents only 10% of
    renal aldehyde dehydrogenase activity (Anders, 1980). At least two
    aldehyde dehydrogenase isozymes occur in proximal tubule mitochondria
    and cytosol.  Substrates include formaldehyde, acetaldehyde, acrolein,
    and malondialdehyde (Hjelle et al., 1983).

         DT-diaphorase (NADPH-quinone oxidoreductase) is a cytosolic
    enzyme that is present at highest concentration in the medulla and
    catalyses a two-electron reduction of quinones to less reactive
    hydroquinones.  It therefore blocks redox cycling and the generation
    of superoxide anion radicals (Lind et al., 1978).

    4.  THE MECHANISTIC BASIS OF CHEMICALLY INDUCED RENAL INJURY

         Over the past 20-30 years there has been a growing understanding
    of the molecular basis of disease and the biochemical mechanisms that
    are associated with chemically induced cellular degeneration and
    lesions in target organ systems. The application of this understanding
    provides a foundation upon which to study chemically induced renal
    injury and, in particular, a rational basis for the extrapolation of
    animal toxicity data to man and risk assessment.

    4.1  Immunologically induced glomerular disease

         Immunologically mediated glomerulonephritis can result from the
    deposition of free circulating antibodies interacting with a
    structural glomerular antigen or with a "planted", non-glomerular
    antigen (such as cationic proteins that fix on anionic sites of the
    glomerular basement membrane).  Alternatively, glomerular injury may
    be the consequence of the deposition of circulating immune complexes.
    The two main forms of antibody-mediated glomerulonephritis are the
    anti-GBM-mediated disease and membranous glomerulonephritis  (or 
    immune complex-type  mediated glomerulonephritis). The former disease
    is characterized by the presence of linear IgG deposits along the
    glomerular basement membrane. The latter form may result from either
    of the mechanisms mentioned, i.e. deposition of free circulating
    antibodies against an irregularly distributed antigen, which may be of
    glomerular origin, or deposition of circulating immune complexes. A
    membranous glomerulonephritis is characterized by the presence of
    granular IgG deposits along the glomerular basement membrane.

         Besides antibody-mediated glomerulonephritis, it is more and more
    apparent that there are cell-mediated glomerulonephritides without Ig
    deposition. An example of such a disease is probably the nephrotic
    syndrome with minimal glomerular changes at the light microscope
    level.

         The role of genetic factors, which has been well demonstrated in
    the mercury model in the rat, is also clear in the human situation.
    Membranous glomerulonephritis induced by gold and D-penicillamine is
    much more frequent in DR3-positive rheumatoid arthritis patients and
    in poor sulfoxydators. The lupus-like disease observed in
    hydralazine-treated patients is more frequent in those with the DR4
    antigen and "slow-acetylators".

         Other drugs such as non-steroidal anti-inflammatory agents or
    lithium salts may induce the nephrotic syndrome with minimal
    glomerular changes. Immunofluorescence is negative in these cases, and
    there is indirect evidence that  such disease could be due to T
    cell-mediated immunity.

         There are many potential environmental agents, drugs, and toxic
    chemicals that have been related to this form of glomerulonephritis. 

    Drugs and toxic chemicals that may induce glomerulonephritis in humans
    include gold and mercury, d-penicillamine, non-steroidal
    anti-inflammatory agents, and heroin. Other drugs and chemicals, in
    which association is suspected but not certain, include silica
    exposure, toxic-oil syndrome, hydrocarbon exposure, and interferon.
    Other drugs may induce an immune complex type of glomerulonephritis in
    the context of a lupus-like syndrome (e.g., hydralazine, procanamide,
    and diphenylhydantoin.  Furthermore, there are many other substances
    that have been implicated by case reports or unconfirmed experiments.
    Anti-GBM-mediated glomerulonephritis is not the usual mechanism
    responsible for toxic glomerular nephropathy.  However, exposure to
    organic solvents is thought to be a factor in some cases of
    Goodpasture's syndrome, which consists of the simultaneous occurrence
    of necrotizing  haemorrhagic interstitial pneumonitis and
    proliferative usually rapidly progressive glomerulonephritis.  This is
    an auto-immune disease resulting from antiGBM antibodies that
    cross-react with basement membranes in lung alveoli.

         The mechanisms by which drugs and chemicals induce
    immunologically related renal disease are both complex and not
    entirely understood. It is unlikely that drugs or toxins induce renal
    damage by modifying self antigens or by acting as haptens. On the
    other hand, many agents such as  gold, d-penicillamine, and mercury 
    may have an immunomodulatory effect. This mechanism of action is also
    not understood, but Druet et al. (1987) and Druet (1989) have
    summarized the evidence that this effect is dependent on genetic
    factors and may be related in the rat to the appearance of anti-class
    II T cells.

    4.2  Direct glomerular toxicity

         Glomerular lesions may be caused by the direct toxicity of a drug
    or chemical. Direct toxicity to components of the glomerular
    apparatus, apart from immunologically induced injury, is relatively
    uncommon.  However, it has been described following exposure to
    puromycin or to materials that may be deposited in the basement
    membrane (Caulfield et al., 1976). Particulate substances such as gold
    and silica may become deposited in mesangial cells (Burkholder, 1982).
    Whether material within mesangial cells is actually phagocytosed is
    unclear, but the reaction to such deposits may be a proliferation of
    glomerular cells and inflammatory cell response.  Injury to the
    mesangium may alter glomerular permeability. Solutes and water move
    across glomerular capillary walls through an extracellular pathway
    that consists sequentially of endothelial fenestrae, the glomerular
    basement membrane, the pores of slit diaphragms, and the filtration
    slits. Water permeability is determined by the total area of
    epithelial slit pores.  Contraction of the glomerular mesangium
    shortens and narrows glomerular capillaries, which in turn narrows
    epithelial slit pores and reduces glomerular filtration.

         Damage to the glomeruli may also occur as a result of fibrin
    deposition due to local or systemic activation of the coagulation
    system. This may be induced by a variety of renal diseases including
    toxic or immunotoxic disorders. Fibrin deposits  per se may damage
    glomeruli in several ways, including occlusion of glomerular
    capillaries, involvement in an inflammatory reaction, or direct
    toxicity to glomerular mesangial cells (Kanfer, 1989).

    4.3  Tubulointerstitial disease

         Tubulointerstitial disease may result from hypersensitivity to
    specific drugs or chemicals or from direct toxicity to tubular
    epithelial cells.  Most forms of tubular injury also involve the
    interstitium, and so these forms of renal injury are considered
    together. However, they may be subdivided, on the basis of
    clinico-pathological features, into three groups of disorders.

    4.3.1  Acute interstitial nephritis

         Acute interstitial nephritis (AIN) occurs as an immunoallergic or
    cell-mediated immune response to a variety of drugs, particularly
    penicillin and its derivatives (e.g., methicillin), but is also
    reported after therapy with thiazdes, non-steroidal anti-inflammatory
    drugs, gold salts, and occupational exposure to mercury (Kleinknecht
    et al., 1978; Clive & Stoff, 1984).

         Both humoral and cellular immunity are involved. Anti-tubular
    basement membrane antibodies are probably involved in some cases of
    methicillin- or diphenyhydantoin-induced immunological nephritis. In
    the majority of cases of acute interstitial nephritis no immune
    reactants are found. The most striking feature is the presence of
    cells infiltrating  the interstitium with mononuclear cells and
    eosinophiles.  Most lymphocytes have been shown to be T cells; most of
    these are T4 cells (helper/inducer cells) and a lesser fraction
    composed of T8 cells (suppressor/ cytotoxic cells). It has been
    suggested that the T cells may be activated by drug exposure (Druet et
    al., 1987; Druet, 1989).

         When extrarenal signs and clinical symptoms are present, they may
    reflect a systemic hypersensitivity reaction that includes fever, skin
    rash, and eosinophilia. Renal involvement is manifested by mild
    proteinuria and haematuria. AIN tends to be more severe, with a high
    incidence of renal failure, in adult patients, but is usually milder
    in patients under 15 years of age. It has also been pointed out that
    absence of prior allergic reaction to a drug (e.g., penicillin) does
    not alter the risk of AIN. The kidneys are usually swollen (as
    visualized radio-graphically) because of oedema fluid and cellular
    infiltration, composed most commonly of lymphocytes and plasma cells 
    as well as eosinophilic and polymorphonuclear neutrophils. In some
    instances the histological appearances may suggest chronic
    inflammation, and macrophages and giant cells may be present. Renal
    tubular cell damage is always present, but there is no fibrosis in the

    acute stages. Glomerular and vascular lesions are uncommon, and the
    lesions are usually reversible.  However, persistent loss of renal
    function indicates progression to fibrosis and chronic interstitial
    disease in an undetermined number of cases. Anti-tubular basement
    membrane antibodies are probably involved in some cases of
    methicillin- or diphenylhydantoin-induced nephritis. Deposits of IgG
    and C3 may be detectable along tubules in biopsies during the acute
    phase. IgE may be elevated in the serum, confirming the allergic or
    hypersensitivity process. The reaction can be further indicated by
    tests of cell-mediated hypersensitivity (lymphocyte transformation
    test) or antibodymediated hypersensitivity (circulating antibodies
    reacting with the drug).

    4.3.2  Acute tubular toxicity

         Acute tubular effects of drugs and toxins are the result of
    direct cellular toxicity. They may vary from necrosis of tubular
    cells, leading to acute renal failure, to subtle subcellular lesions
    and functional effects. The major groups of agents causing acute
    tubular toxicity are antibiotics, particularly aminoglycosides,
    contrast agents, non-steroidal anti-inflammatory drugs, and
    chemotherapeutic agents including cyclosporin A and  cis-platinum.

         Injury to proximal tubular lining cells is manifest by increased
    excretion of substances normally resorbed by these cells, such as
    glucose, amino acids, phosphate, and sodium. Extension of the lesion
    to distal portions of the tubule is accompanied by loss of the ability
    to acidify the urine and to maintain water and electrolyte balance.
    Tubular toxicity may be accompanied by glomerular effects, and, if the
    process is persistent, may lead to a chronic interstitial nephropathy,
    as in lead and cadmium toxicity.

    4.3.3  Chronic interstitial nephritis

         Chronic interstitial nephritis (CIN) generally has fewer
    distinguishing morphological features than most forms of acute renal
    disease. It is characterized morphologically by infiltration with
    mononuclear cells, prominent interstitial fibrosis, and tubular
    atrophy. The best documented cause of CIN is analgesic nephropathy
    which is often accompanied by acute papillary necrosis. CIN may also
    occur as a sequela to severe acute tubular disease or acute
    interstitial nephritis, or as an expression of chronic low-dose
    exposure to specific nephrotoxins (lead or cadmium nephropathy). The
    term "tubulointerstitial disease" may be preferable, because it better
    identifies the primary sites of the pathological process. Progressive
    fibrosis of the interstitial tissue results in a decreasing number of
    functional nephrons with eventual reduction in glomerular filtration
    rate and azotaemia. There may be few symptoms preceding the onset of
    renal failure.

    4.4  Mechanisms of cellular toxicity

         There are several mechanisms that are thought to be central to
    toxicological injury, including impaired lysosmal function, membrane
    changes, and oxidative stress. It is now widely accepted that
    Ca2+-homeostasis in the cell and Ca2+-mediated cell functions are
    critical targets for numerous pathophysiological processes including
    toxicant-induced cell death (Recknagel, 1983; Pounds & Rosen, 1988).
    Many classes of pharmaceuticals and other chemicals (e.g., metals,
    pesticides, and solvents) impair the calcium messenger system (Pounds,
    1984; Olorunsogo et al., 1985; Moore et al., 1986). Disturbances in
    intracellular Ca2+ homeostatis and sustained increase in cytosolic
    Ca2+ cause cell death by the disruption of the plasma membrane,
    cytoskeleton, endoplasmic reticulum, and mitochondria.  In addition,
    chemicals (alkylating or arylating agents) can be toxic and may induce
    cell death through an initial DNA damage or by apoptosis
    (receptor-mediated programmed cell death). In cell injury caused by
    chemical toxicants, cellular accumulation of Ca2+ and the generation 
    of oxygen free radicals damage cellular components, particularly
    mitochondrial membranes. Indeed Ca2+ potentiates oxygen free radical
    injury to renal mitochondria (Malis & Bonventre, 1986), and the result
    of this detrimental interaction could be due, in part, to the
    activation of phospholipase A2.

         Lipid peroxidation has been suggested as one of the possible
    mechanisms whereby chemicals may produce membrane damage and cell
    death. Free radicals, generated either directly by the metabolism of
    a chemical or from the reduction of oxygen (forming O2-, H2O2
    and  OH*),  can initiate lipid peroxidation via hydrogen abstraction
    from polyunsaturated fatty acids. This interaction will form lipid
    peroxyradicals and lipid hydroxyperoxides, propagating the chain
    reaction. Such a chain reaction may destroy cellular membranes and
    thereby result in increased plasma membrane permeability or altered
    fluidity and cell death. Lipid peroxidation may also cause cell death
    through the formation of potent toxic lipid metabolites (such as
    hydroxyalkenals). However, several lines of evidence indicate that
    lipid peroxidation is most often independent (or is a consequence
    rather than the cause) of cell death (Witschi et al., 1987). One or
    more of these mechanisms of cellular injury could closely interact.

         Proximal renal tubular cells are particularly vulnerable to the
    toxic action of chemicals, owing to their high energy demand (such as
    reabsorptive and secretory functions). Redox-active agents may cause
    extensive oxidation of GSH to oxidized glutathione (GSSG). Under such
    conditions,  often referred to as "oxidative stress", reduction of
    GSSG back to GSH by the NADPH-dependent GSSG reductase is lower than
    the rate of GSH oxidation. This may lead to gluthathione depletion and
    cause oxidation of cellular enzymes, depletion of cellular ATP, and
    loss of mitochondrial function (Trump et al., 1989).

         Reactive electrophilic metabolites are known to bind covalently
    to tissue proteins, and it has been suggested that cell injury and
    death are a consequence of the interaction of such reactive
    intermediates with critical cellular molecules. Free sulfhydryl groups
    are involved in the catatylic activity of many proteins. Modification
    of such sulfhydryl groups by covalent binding or by oxidation may
    inactivate critical enzymes and lead to cell death. For some chemicals
    the loss of protein sulfhydryl groups results mainly from a reversible
    oxidative process, which leads to the formation of disulfide
    cross-links or mixed disulfides with another protein or GSH. Enzymes
    involved in Ca2+ homeostasis may be one example of such critical
    cellular target molecules for alkylating/arylating or oxidizing
    metabolites.

         Studies on isolated cells have shown that exposure to lethal
    concentrations of some cytotoxic chemicals leads to a rapid and
    sustained rise in cytosolic Ca2+.  Ca2+-mediated cytotoxicity may
    at least in part be related to effects on cytoskeletal organization.
    High Ca2+ concentrations affect the regulation of the formation of
    actin bundles and tubulin polymerization. Activation, by high
    intracellular Ca2+,  of Ca2+-dependent  proteases that cleave
    cytoskeletal proteins has been proposed as one mechanism of cell death
    (Fig. 11). Activation of other enzymes, such as phospholipase A2 
    (causing disruption of the plasma membrane and formation of toxic
    membrane breakdown products) and endonucleases (extensive DNA
    fragmentation), has also been associated with cell death (Trump et
    al., 1989).

         Certain alkyl halides such as DBCP (Omichinski et al., 1987;
    Dybing et al., 1989) and reactive oxygen species cause single-strand
    breaks in DNA. Extensive DNA damage activates poly
    ADP-ribosyltransferase, which leads to a critical reduction in
    cellular NAD+ levels,  followed by depletion of ATP and eventual
    cell death.

         The process by which glucocorticoids induce killing of immature
    thymocytes has been termed programmed cell death (apoptosis). Recent
    studies indicates that the environmental contaminant
    2,3,7,8-tetrachlorinated dibenzo- p-dioxin (TCDD) causes a
    receptor-mediated influx of Ca2+ in thymic cells, which in turn
    activates endonucleases and thereby causes programmed cell death. The
    role of such a process in the chemically mediated killing of kidney
    cells has yet to be determined (McConkey et al., 1988).

    4.5  Factors that modify cellular injury by toxins

    4.5.1  Cellular transport and accumulation

         Drugs and other chemicals including metals may be transported
    across proximal tubular cells, i.e., from renal capillaries across
    tubular cells to be excreted in the tubular lumen or vice versa
    (absorption). Many organic anions are excreted against concentration

    gradients at rates that exceed glomerular filtration. This implies an
    active carrier-mediated transport process. Such a process requires
    energy obtained from oxidative metabolism located in mitochondria. An
    active process for transporting solutes in renal tubular cells has
    certain implications concerning the susceptibility of tubular cells to
    effects of toxins (Berndt, 1989). If cationic drugs or chemicals are
    actively transported, there is the immediate problem of competition
    with the transport of essential cations. Active transport with the
    capability of concentrating absorbed material may concentrate
    potential nephrotoxins as well as essential solutes in the renal
    cortex.  The same toxins that impair energy metabolism will impede the
    cellular transport of essential solutes (Rennick, 1978). Other toxic
    substances may be concentrated in the medulla or the papillae,
    probably as a consequence of the physiological mechanism that
    concentrates urine.  The renal accumulation of chemicals such as
    gentamicin, cephaloridine, or cadmium is well documented.

    4.5.2  Metabolic degradation

         Metabolic degradation or transformation most often occurs in the
    liver, but many of the same enzyme systems are present in the kidney
    as well. The metabolism of drugs and chemicals within the kidney may
    result in substances that are either more or less toxic. Those drugs
    and chemicals that are metabolized by the mixed-function oxidase
    system have received the most attention. For example, several
    chlorinated alkyl hydrocarbons of low relative molecular mass, such as
    carbon tetrachloride and trichloromethane, may be transformed into
    reactive, toxic products that bind covalently to renal tissue,
    producing membrane injury. In addition, low-level exposure to other
    substances, such as polychlorinated biphenyls (PCBs), that activate
    the enzyme systems may enhance the production of toxic products (Kluwe
    et al., 1979). Similarly, pretreatment with phenobarbital enhances the
    activity of mixedfunction oxidase enzymes and, hence, the toxicity of
    compounds like methoxyflurane whose metabolic products are fluoride
    and oxalate, two substances potentially toxic to the kidney. The
    fluoride ion is toxic to cell membranes, whereas oxalate may
    precipitate within the lumen of nephrons (Mazze, 1976). On the other
    hand, phenobarbital reduces the renal toxicity of DBCP due to an
    increase in its detoxication in the liver (Kluwe, 1983).

    4.5.3  Intracellular protein binding

         The intracellular concentration of toxins may be influenced by
    protein binding. The soluble cytoplasmic protein, metallothionein, and
    insoluble acidic protein complexes forming nuclear inclusion bodies
    are examples of a phenomenon that concentrates two different groups of
    metals.

         Metallothioneins are proteins of low relative molecular mass
    (6000-7000 Daltons) characterized by a high cysteine content (23-33%),
    a complete lack of aromatic acids, and a high content of heavy metals
    (7-12 metal atoms/molecule of protein).  Metallothioneins can bind

    FIGURE 11

    several essential or non-essential heavy metal ions including zinc,
    copper, cadmium, mercury, silver, gold, and cobalt (Goyer & Cherian,
    1977). The metal ions are bound exclusively through thiolate
    coordination complexes, which involve all the cysteine residues (20 in
    rat liver metallothionein) located in two domains (alpha and ß
    domains). Metal ions that can bind to metallothioneins can also, to
    variable extents, promote the transcriptional activity of
    metallothionein genes. In the kidney, Cd2+ and Hg2+ are the best
    inducers of metallothionein synthesis. Metallothionein synthesis can
    also be induced by various stresses (e.g., tissue injury, food
    restriction, infections). Not all the biological functions of
    metallothionein have been fully elucidated. They probably include
    protection against and detoxification of heavy metals, regulation of
    the metabolism and possibly the function of essential elements such as
    copper and zinc, and a protective response to various stresses by
    altering zinc distributions between tissues and within cells (e.g.,
    macrophages) and by acting as a free radical scavenger (Dunn et al.,
    1987).

         Lead and bismuth accumulate in renal tubular cells bound to a
    complex of acidic proteins that form morphologically discernible
    inclusion bodies (Goyer & Cherian, 1977). As with metallothionein, the
    sequestering of toxic metals by the protein complex is thought to
    reduce the intracellular toxicity of these metals.

    4.5.4  Membrane reactions and pinocytosis

         Macromolecular substances are transported by pinocytosis and
    included in intracellular vacuoles. Proteins that are normally in the
    glomerular filtrate are taken up by the cell membrane by pinocytosis. 
    Such pinocytotic vesicles fuse with primary lysosomes, which contain
    lytic enzymes. Secondary lysosomes are formed, and the macromolecular
    material is degraded or broken down. The products of low relative
    molecular mass then leave the lysosomes in order to prevent an
    increase in osmolality and lysosome swelling (Jacques, 1975).

         Potential nephrotoxins that may be taken into renal tubular cells
    in this manner include chelating agents such as nitrilotriacetic acid,
    ethylenediaminetetraacetic acid (EDTA), and metallothionein.  Membrane
    binding of EDTA administered as the calcium-EDTA chelate persists, the
    calcium but not the EDTA being dislocated to other cellular
    components. This suggests the manner in which EDTA may sequester
    cellular lead or other metals for excretion (Schwartz et al., 1970).

    5.  THERAPEUTIC AGENTS AND CHEMICALS THAT HAVE THE POTENTIAL TO CAUSE
        NEPHROTOXICITY

    5.1  Therapeutic agents

         Many therapeutic agents have been linked to clinically
    significant nephrotoxicity. At present much is known and understood
    about some of these agents, as there is a substantial amount of
    relevant animal toxicity and human data for comparison.

    5.1.1  Analgesics and non-steroidal anti-inflammatory drugs (NSAIDs)

         Analgesic nephropathy may be a consequence of the excessive
    consumption of mixed analgesics. Originally, phenacetin was common to
    all of these mixtures, which led to the conclusion that this drug was
    the only cause of "phenacetin kidney". Subsequently, a variety of
    analgesics, NSAIDs, and a number of industrial and environmental
    chemicals have been shown to have the potential to cause RPN and
    interstitial nephritis (Burry et al., 1977; Schwarz, 1987).

         The prevalence of analgesic-associated nephropathy varies
    worldwide and is probably related to patterns of analgesic use. It has
    been found most often in women aged 45-55 as a result of a high
    incidence of analgesic abuse, and has been reported more frequently in
    Australia and Switzerland and less frequently in the USA, Canada, and
    Germany. It is estimated that more than 37 million people in the USA
    have arthritis and use these drugs.  This is indeed a very large
    population at risk. Analgesic effects are said to be a factor in as
    many as 20-30% of cases of interstitial nephritis in the south-east of
    the USA (Murray & Goldberg, 1975).

         Diagnosis is often made by coupling the history of analgesic
    abuse with morphological evidence of renal papillary necrosis and
    chronic interstitial nephritis. The resultant lesions can be
    recognized by radiological examinations or ultrasonography and consist
    of calcifications along the line of Hodson, shrinking of the kidneys
    resulting in irregular contours, and decreased length of both kidneys.
    Necrotic papillae may be voided in the urine and this is occasionally
    observed.

         At least two countries have legislated to prevent phenacetin from
    being sold over the counter. This legislation has had the effect of a
    change in analgesic formulation, usually towards that of a single drug
    such as aspirin or paracetamol (acetaminophen). In Sweden, the removal
    of phenacetin resulted in the progressive decline in the incidence of
    analgesic nephropathy. This began approximately 6 years after the
    removal of phenacetin and the major impact was seen after 12 years. In
    Australia, phenacetin was removed from combination analgesics by 1976
    and replaced with acetaminophen (the principle metabolite of
    phenacetin); by 1979 legislation prohibited the overthe-counter sale
    of combination analgesics. According to the Australia-New-Zealand

    Dialysis and Transplant Registry, this has resulted in a progressive
    decline in the incidence of analgesic nephropathy in patients
    presenting with end-stage chronic renal failure in dialysis and
    transplant programmes. In 1982, 22% of patients in dialyses and
    transplant programmes had analgesic nephropathy; by 1988, the
    incidence had declined to approximately 13%. By contrast, the sale of
    phenacetin-containing analgesics declined to 21% (in 1976) and 9% (in
    1983) of the total volume of analgesics sold in Belgium, but the
    prevalence of analgesic nephropathy remained unchanged over this
    period.  Retrospective and prospective studies in Belgium have shown
    decreased renal function in analgesic abusers who never took analgesic
    mixtures containing phenacetin (Elseviers & De Broe, 1988). 
    Paracetamol, which largely replaced phenacetin in analgesic mixtures),
    is assumed to be involved in the pathogenesis of RPN.  Recently, an
    increased risk of renal disease was found in daily users of
    paracetamol in North Carolina, USA (Sandler et al., 1989).  Other
    therapeutic agents (mostly analgesics and NSAIDs)  have been
    implicated in RPN, and a number of industrial and environmental
    chemicals also have the potential to cause this lesion (Bach &
    Bridges, 1985a).

         Most patients deny analgesic abuse, thus making epidemiological
    studies that attempt to identify the causative agent difficult. Most
    of the epidemiological data that has been reported over-reflected the
    intake of phenacetin at the expense of other agents that could equally
    be implicated. The estimate of the total lifetime dose of analgesic
    that produces papillary necrosis varies from < 1 to 35 kg (but this
    was based only on phenacetin). No data was given in the
    epidemiological studies on what quantities of other analgesics were
    also taken or on exposure to papillotoxic chemicals. Thus the etiology
    of RPN has been complicated by poly-pharmacy, lack of documentation on
    the intake of other analgesics and NSAIDs, exposure to other
    chemicals, and the difficulty in diagnosing RPN.

         Analgesic abuse (Nanra et al., 1978; Nanra, 1980; Prescott, 1982;
    Bach & Bridges, 1985a; Schwarz, 1987) and addiction have been linked
    to co-formulation with caffeine, but there is no firm supporting
    evidence. Few analgesic abusers take the drugs for appropriate
    indications. The origins of abuse are usually psycho-social  and
    represent  neurotic, dependent, immature, introverted, anxious, or
    depressed individuals, up to 20% of whom also smoke and abuse alcohol,
    psychotropic drugs, and sleeping tablets. Most analgesic abusers are
    women (over 30 years of age) from lower socioeconomic/education
    groups, who have taken these mixtures for 5-30 years. Several factors
    such as dehydration, secondary to high ambient temperatures and
    bacterial infection, have been implicated in the development of renal
    failure (Kincaid-Smith, 1979). Renal function may be preserved and the
    progression to ESRD may be prevented by stopping analgesic exposure,
    but patients who continue to abuse analgesics have a poor prognosis
    (Schwarz, 1987). Analgesic abusers also have an increased incidence of
    anaemia, gastric ulcers, and cardiovascular heart disease (Dubach et
    al., 1978).

         Clinical features of analgesic nephropathy include loss of
    urine-concentrating capacity (Bengtsson, 1962), electrolyte
    disturbances, (sodium wastage and hypocalcaemia),  and  defective
    urinary  acidification after ammonium chloride loading (Bengtsson,
    1962; Nanra et al., 1978). An increase in BUN or creatinine 
    identifies incipient renal failure; at this stage papillary necrosis
    is well advanced and includes secondary degenerative changes. 
    Radiology and ultrasound may identify irregular shrinking of the renal
    tissue and medullary calcifications, but these are advanced changes.
    Very early in the course of the injury histological changes are
    confined to the medulla (a region of the kidney that is not always
    assessed at autopsy). This situation progresses to include other
    changes, especially in the cortex, that are easily biopsied to show
    interstitial nephritis, but they do not define the underlying cause.

         The earliest degenerative changes begin at the papilla tip and
    affect interstitial cells, loops of Henle, capillaries, and the
    proteoglycan ground substance, and result in lipidosis. More advanced
    or intermediate RPN affects the outer medulla, as seen by atrophy,
    sclerosis, and inflammatory response and calcification of the necrosed
    papilla tip (Burry et al., 1977; Gloor, 1978). Total RPN affects the
    corticomedullary and cortical regions and is characterized by chronic
    interstitial nephritis, tubular dilatation, atrophy, basement membrane
    thickening, fibrosis, sclerosis, and inflammatory cell infiltration.
    Vascular degeneration such as suburothelial capillary sclerosis is
    pathognomonic for RPN (Mihatsch et al., 1984). Pelvic, ureteric, and
    bladder urothelia show thickening of capillary walls, sclerosis of
    lamina propria, altered fat and collagen deposition, and epithelial
    hyperplasia advancing to malignancy and tumours (Burry et al., 1977;
    Mihatsch et al., 1984).

         The pathological changes in humans with RPN have been obtained
    from autopsy or postmortem tissues, where autolytic degradation may
    alter the appearance and make interpretation of the stage of the
    lesion difficult. Animal models of RPN in a laboratory situation have
    provided more detail on the focus of primary injury and the course of
    degenerative changes and also allow mechanisms to be studied, but they
    do not necessarily reflect the situation in humans.

         Analgesics (e.g., aspirin, phenacetin, and paracetamol) do not
    always cause RPN in rats. Thus inappropriately high doses have
    sometimes been given for prolonged periods (Prescott, 1982; Bach &
    Bridges, 1985a). Biological variation within such groups is very high
    and experimental use of NSAIDs has caused fatalities due to
    extra-renal toxicity (gastric ulceration and perforation; see Kaump,
    1966).  This has produced experimental models that are of limited
    value for studying the course of the lesion or mechanism of RPN. The
    renal functional changes and the pathomorphological progression of the
    lesion in several acute model systems show marked similarities with
    those reported for the analgesic-associated lesion in both
    experimental animals and man (Bach & Hardy, 1985; Bach & Bridges,
    1985a).

         The histological changes induced by experimental RPN follow a
    similar pattern of early, intermediate and total RPN to that described
    in man (Fig. 12), and are dose and time dependent. The earliest
    morphological changes induced by papillotoxins occur in the renal
    medullary interstitial cells. The medullary glycosaminoglycan matrix
    also undergoes changes, showing an increase and then a decrease in
    staining intensity. It is only subsequently that there are platelet
    adhesions, blocking of blood vessels, degenerative changes in the
    collecting ducts and proximal tubules, and the accumulation of lipid
    material in capillaries and epithelial cells. At the same time as
    repair and reepithelization are taking place, there is an increase in
    the presence of tubular casts, proximal and distal tubular dilatation,
    and hyperplasia of the collecting ducts, and pelvic  urothelia, and
    the  suburothelial capillaries undergo sclerotic changes.  When the
    repair phase is advanced or complete, there are also degenerative
    changes in the cortex, including fibrosis, glomerular sclerosis, and
    cystic dilatation. The histological and functional changes produced by
    model RPN are remarkably similar to those observed in human analgesic
    abusers. The use of high-resolution  light microscopy and 
    ultrastructural studies (in conjunction with histochemistry and
    immuno-histochemistry) can help establish the changes in adjacent
    cells and link the "cause-and-effect" relationships in the sequence of
    degenerative events.

         The mechanism of analgesic-induced renal papillary necrosis is
    still not fully understood. Progress in our understanding of the
    pathogenesis of the model lesions (Bach & Bridges, 1984, 1985a) has
    enabled some factors to be identified that may be involved in the
    molecular changes. There is no evidence to suggest that the model
    lesion has an early immunological basis, nor that it is a consequence
    of renal hypoxia or vasoconstriction, and there is no experimental
    basis to suggest that the altered intermediary metabolism is a
    critical factor (Bach et al., 1983).  The concept that altered PG
    metabolism gives rise to vascular (or other) changes is an attractive
    one, but the exceptionally low levels of these hormones, combined with
    their instability, have made it very difficult to test this
    hypothesis. The countercurrent concentration mechanism is an important
    normal renal function and is thought to play an important role in
    concentrating chemicals to toxic levels within the medulla.  One of
    the earliest changes in the development of RPN is the loss of
    concentrating processes, which detracts from this hypothesis.
    Furthermore, the concentrating of a compound in the medulla does not
    explain the molecular mechanism by which it causes RPN (Bach &
    Bridges, 1985a).

         At present the most attractive explanation for the development of
    RPN relates to a metabolic activation within the kidney. There are two
    major oxidative systems for xenobiotic metabolism in the kidney. The
    cytochrome P-450 system is localized to the cortex, whereas the PG
    hydroperoxidase system (PGH) is located almost exclusively in the
    medulla. The reasons for the selective targeting of particular
    chemicals for the renal medulla interstitial cells are uncertain, but

    FIGURE 12

    may relate to the absence of free radical scavengers or nucleophiles
    and/or to the presence of extensive numbers of lipid droplets
    containing polyunsaturated fatty acids within these cells. These would
    form an ideal substrate for extensive lipid peroxidation (Porter et
    al., 1980) within the renal medullary interstitial cells once a
    reactive species had been generated within these cells. The role of
    co-oxidation of substrates as a consequence of PG synthesis has become
    an attractive mechanistic basis on which to explain papillary damage
    (Fig. 13) and the activation of bladder carcinogens, and it may also
    be pertinent to other types of renal toxicity that are not associated
    with mixed-function oxidase activity (Bach & Bridges, 1984).
    Prostaglandin endoperoxide synthetase consists of two inseparable
    activities. Fatty acid cyclo-oxygenase catalyses the oxidation of
    arachidonic acid to PG hydroperoxy-endoperoxide (PGG2),  while the
    other activity, PG hydroperoxidase, reduces PGG2 to PGH2 and
    co-oxidizes another molecule. PGH2 is the precursor for both PGs and
    thromboxanes (Davis et al., 1981). PG hydroperoxidase is inhibited by
    antioxidants and will reduce many different fatty acid peroxides,
    other organic peroxides (cumene hydroperoxide and  tert-butyl 
    hydroperoxide), and inorganic peroxides (hydrogen peroxide) and, in
    the process, co-oxidize a range of substrates including several
    bladder carcinogens that produce free radicals.

         Paracetamol forms a reactive intermediate that co-valently binds
    to trichloroacetic-acid-precipitable macro-molecules but is inhibited
    by aspirin and other inhibitors of fatty acid cyclo-oxygenase.
    Ethoxyquin, ascorbic acid, and glutathione also inhibit covalent
    binding of paracetamol during peroxidative activation by reacting with
    electrophilic intermediates generated by co-oxidation (Zenser et al.,
    1983).

         Other non-steroidal anti-inflammatory agents may produce
    hypersensitivity reactions, lipoid nephrosis, and interstitial
    nephritis (Finkelstein et al., 1982).

    5.1.2  Paracetamol (acetaminophen) and para-aminophenol

         Large doses of paracetamol can produce acute proximal tubular
    necrosis, especially in male Fischer-344 rats (Mitchell et al., 1977;
    McMurtry et al., 1978; Hennis et al., 1981; Newton et al., 1983a,b).
    Microsomal cytochrome P-450 activation to a reactive arylating
    intermediate is thought to be an obligatory biochemical event in
    paracetamol-induced hepatic necrosis (Mitchell et al., 1973; Nelson,
    1982). Nephrotoxic dosages of paracetamol bind covalently to renal
    protein (Mitchell et al., 1977; McMurtry et al., 1978; Nelson, 1982)
    by an NADPH-dependent, cytochrome-P-450-mediated process (McMurtry et
    al., 1978; Newton et al., 1983a,b).  Alternatively, paracetamol is
    enzymically deacetylated to  para-aminophenol, a potent selective
    nephrotoxin that damages the proximal tubule (Calder et al., 1979).
     Para-aminophenol produces acute necrosis of the proximal convoluted
    tubules in rats after a single injection (Green et al., 1969), and has

    FIGURE 13

    been demonstrated to be a minor metabolite of paracetamol in the
    Fischer-344 rat and its isolated perfused kidney (Newton et al.,
    1982). Paracetamol ( N-acetyl- p-aminophenol) is structurally
    closely related to  para-aminophenol, and metabolites have been shown
    to be excreted by the biliary route in rats (Siegers & Klaassen, 1984)
    and mice (Fischer et al., 1985). These metabolites are the glucuronic
    acid and sulfate conjugates (Siegers & Klaassen, 1984) and the
    glutathione conjugate (Hinson et al., 1982). Toxicity arising from
     para-aminophenol has been previously suggested to result from a
    dose-related depletion of kidney reduced glutathione and covalent
    binding to essential renal macromolecules (Crowe et al., 1977; 1979).

         Both paracetamol and  para-aminophenol deplete renal cortical
    reduced glutathione concentrations and arylate renal macromolecules
    (McMurtry, et al., 1978; Crowe et al., 1979). The changes produced by
     para-aminophenol are indistinguishable from those caused by
    paracetamol (Newton et al., 1983a,b). Mouse renal cortical slices and
    homogenates are capable of deacetylating paracetamol to
     para-aminophenol (Carpenter & Mudge, 1981), which has also been
    identified as a urinary metabolite of paracetamol in both hamsters
    (Gemborys & Mudge, 1981) and Fischer-344 rats (Newton et al.,
    1983a,b). Thus the rat is capable of deacetylating paracetamol to
     para-aminophenol.  In the renal cortex, paracetamol deacetylation
    occurs primarily in the cytosolic fraction (Newton et al., 1983a).
    Similarly, metabolic activation of paracetamol to an arylating
    intermediate is dependent on the presence of a cytosolic deacetylase
    (Newton et al., 1983b).

         Both  para-aminophenol and bis-( p-nitro-phenyl)-phosphate (a
    carboxylesterase-amidase inhibitor) inhibit the covalent binding of
    paracetamol to renal macromolecules (Newton et al., 1983b). Conclusive
    evidence that  para-cetamol binds to renal macromolecules after
    deacetylation and metabolic activation to  para-aminophenol has been
    provided by the demonstration of covalent binding of [ring-
    14C]-paracetamol, but not [acetyl-14C]-paracetamol, to renal
    protein (Newton et al., 1983b).

         Thus, paracetamol activation by renal cortical tissue can occur
    in two different ways, i.e. either a microsomal
    cytochrome-P-450-dependent pathway or deacetylation to
     para-aminophenol and subsequent metabolic activation. The reactive
    intermediates formed by each pathway suggest that both mechanisms may
    play a role in paracetamolinduced renal cortical necrosis.

    5.1.3  Antibiotics

         Nephrotoxicity related to antibiotics is most often due to
    effects on transport, concentration, and excretory functions.  All
    parts of the nephron or kidney may be affected.  However, there is
    usually some specificity in the site of action, particular toxins
    affecting specific portions of the nephron (Curtis, 1979).

         Mechanisms of injury span a broad spectrum of potential lesions.
    The most common effect is direct toxicity to renal tubular cells
    manifested by cell injury and necrosis. Direct toxicity to glomeruli
    is not as conspicuous but does occur. Immunologically induced lesions
    in glomeruli and interstitial tissue may also occur.

    5.1.3.1  Aminoglycosides

         Nephrotoxicity is a common complication of aminoglycoside
    antibiotic therapy in man (Bennett, 1983; Matzke et al., 1983;
    Kahlmeter & Dahlager, 1984). Early signs of nephrotoxicity include
    increased urinary excretion of proximal tubular cell brush-border
    membrane enzymes such as  alanine aminopeptidase, proteins of relative
    low molecular mass such as lysozyme and ß2-microglobulin, and granular
    casts (Schentag, 1983). A urine-concentrating defect is usually
    evident and may explain the non-oliguric acute renal failure typically
    observed in these patients. Less common manifestations of tubular
    dysfunction include potassium, magnesium, calcium, and glucose loss in
    the urine.  Azotaemia and elevation of the serum creatinine
    concentration  are relatively late  manifestations of nephrotoxicity 
    and reflect depression  of glomerular filtration rate consequent to
    extensive proximal tubular cell necrosis. Patients receiving standard
    doses of amino-glycoside antibiotics usually do not manifest
    depression of glomerular filtration until after seven or more days of
    drug therapy. However, pathological changes confined to the proximal
    tubule can be seen in renal biopsy material obtained before this time
    (DeBroe et al., 1984).  At the light microscope level, these changes
    range from loss of brush-border membrane, apical blebbing, and
    prominent vacuoles to cloudy swelling, patchy cell necrosis, and
    sloughing of necrotic cells with cast formation in the lumen. Electron
    microscopy reveals the presence of multicentric multilamellar membrane
    structures known as myeloid bodies within distended lysosomes (Kosek
    et al., 1974). These lysosomal lesions can be seen within 1-2 days of
    drug treatment and they increase in size and number as therapy is
    prolonged.

         The nephrotoxicity potential of aminoglycosides has been ranked
    as neomycin > gentamicin > sisomicin = kanamycin > tobramycin >
    netilmicin > streptomycin (Parker et al., 1982). The situation with
    amikacin has been somewhat controversial, but recent studies have
    suggested that it is less nephrotoxic, even in experimental animals,
    than the other clinically available aminoglycosides, except for
    streptomycin.  Clear-cut therapeutic advantages of any particular
    aminoglycoside are not readily apparent in patients because of the
    serious nature of their underlying illness and concurrent therapy with
    multiple drugs. Furthermore, the relative nephrotoxicity is usually
    assessed by insensitive techniques, such as blood urea nitrogen, serum
    creatinine, and enzymuria, that do not give a quantitative
    representation of the extent of renal injury. In humans, few would
    argue that neomycin and gentamicin are much more nephrotoxic in
    therapeutic use than streptomycin, but there are also other important
    risk factors that relate to the clinical condition of the patient:

    *    dehydration, volume depletion, diuretic-induced volume depletion;

    *    advanced age;

    *    pre-existing renal disease;

    *    electrolyte imbalance (acidosis, hypomagnesaemia, hypokalaemia,
         hypocalcaemia);

    *    hypotension/renal ischaemia;

    *    extrarenal target organ disease such as cirrhosis of the liver;

    *    exposure to multiple nephrotoxins;

    *    frequent dose regimens as opposed to larger doses given less
         frequently;

    *    elevated aminoglycoside trough concentrations.

         Current understanding of the pathogenesis of amino-glycoside
    nephrotoxicity has been derived primarily from studies in rats, which
    exhibit a pattern of renal injury indistinguishable from that observed
    in man (Kaloyanides & Pastoriza-Munoz, 1980; Humes et al., 1982;
    Bennett, 1983; Tulkens, 1989). The drug dose, in relation to body
    weight, required to induce injury in the rat is considerably larger
    than that required in man, whereas the dose is approximately the same
    when expressed in relation to body surface area. From such studies has
    emerged unequivocal evidence that aminoglycoside nephrotoxicity is
    causally linked to the transport and accumulation of drugs by renal
    proximal tubular cells. Following parenteral administration,
    aminoglycosides are eliminated unchanged in the urine by glomerular
    filtration. A small fraction of the filtered drug is taken up by the
    renal proximal tubular cells via a low affinity, high capacity
    transport mechanism that exhibits saturation kinetics (Kaloyanides,
    1984a; Giuliano et al., 1986). The first step in this transport
    process involves binding of the cationic aminoglycoside to apical
    membrane receptors, thought to be anionic phospholipids such as
    phosphatidylinositol (Sastrasinh et al., 1982). This is followed by
    uptake into the cell by adsorptive endocytosis (Silverblatt & Kuehn,
    1979) with subsequent translocation and sequestration of the drug in
    high concentration within lysosomes (Morin et al., 1980; Josepovitz et
    al., 1985). In addition a small quantity of drug appears to gain
    access into the cell across the basolateral membrane (Collier et al.,
    1979). Following uptake into proximal tubular cells, aminoglycosides
    express their nephrotoxicity potential by disrupting one or more
    critical intracellular metabolic pathways.

         Although these drugs have been shown to effect a variety of
    biochemical processes at several sites within proximal tubular cells
    (Kaloyanides, 1984b), it remains to be established which if any of
    these actions are causally linked to the cascade that eventuates in

    cell injury and necrosis. Prominent among the biochemical derangements
    is a disturbance of phospholipid metabolism reflected by an increase
    in renal cortical phospholipid enriched in phosphatidylinositol
    (Kaloyanides, 1984b). The phospholipidosis has been shown to be due
    primarily to the accumulation of lysosomal myeloid bodies (Josepovitz
    et al., 1985), which form as a consequence of the inhibition of
    lysosomal phospholipases (Laurent et al., 1982; Carlier et al., 1983)
    by the high concentration of drug within the lysosomal compartment
    (Ramsammy et al., 1989a). The mechanism of inhibition is thought to be
    related to an electrostatic interaction  between the cationic 
    aminoglycoside and anionic phospholipid. Another example of an adverse
    interaction between aminoglycosides and phospholipid is the
    observation that gentamicin inhibits agonist activation of the
    phosphatidylinositol cascade (Ramsammy et al., 1988a), an effect that
    localizes the site of interaction at the cytoplasmic surface of the
    plasma membrane and most likely reflects binding of the polycationic
    gentamicin to the polyanionic phospholipid,
    phosphatidylinositol-4,5-bis-phosphate.  This effect may also explain
    the observation that aminoglycosides inhibit phosphatidylinositol-
    specific phospholipase C in renal brush-border membranes (Schwertz et
    al., 1984). Alterations of other biochemical processes associated with
    plasma membranes have been described, including  depressions of
    Na+-K+-ATPase, adenylate cyclase, alkaline phosphatase, and
    calcium binding (Morin et al., 1980; Williams et al., 1981). Impaired
    mitochondrial respiration (Weinberg & Humes, 1980) and decreased
    incorporation of leucine into microsomal protein (Bennett et al.,
    1988) have also been observed prior to the onset of obvious
    irreversible cell injury.  These findings emphasize that multiple
    sites serve as targets for drug-cell interaction. However, it remains
    uncertain which of these biochemical abnormalities are  proximal
    events causally linked to toxicity.

         One theory that attempts to integrate these diverse observations
    focuses on the lysosomal accumulation of aminoglycosides, with
    induction of a lysosomal phospholipidosis as the critical first step
    (Tulkens, 1989). If the injury threshold concentration of
    aminoglycoside is not reached, the lysosomal phospholipidosis
    regresses without any biochemical or morphological evidence of
    cellular necrosis and regeneration (Giuliano et al., 1984).  However,
    if the injury threshold concentration is exceeded, the lysosomal
    phospholipidosis progresses and the overloaded lysosomes swell,
    resulting in the loss of integrity of the lysosomal membrane and the
    release of lysosomal enzymes, toxins, and large quantities of
    aminoglycosides into the cytosol. The extralysosomal aminoglycoside
    interacts with and disrupts the functional integrity of other
    subcellular membranes, thereby initiating the injury cascade that
    eventuates in cell death.

         It should be emphasized that aminoglycoside-induced proximal
    tubular cell necrosis is accompanied by a conspicuous regenerative
    response (Parker et al., 1982; Toubeau et al., 1986). Thus, the
    clinical threshold for nephrotoxicity is determined by the balance

    between the rate of necrosis and the rate of regeneration of proximal
    tubular cells (Soberon et al., 1979). If necrosis dominates, overt
    renal failure ensues.

         Aminoglycoside nephrotoxicity is accompanied by increased
    generation of free radicals. Furthermore, nephrotoxicity is blocked
    with free radical scavengers/antioxidants such as dimethylthiourea,
    dimethyl sufoxide, sodium benzoate, or deferoxamine (Walker & Shah,
    1987). However other studies have demonstrated that antioxidants such
    as vitamin E do not protect against aminoglycoside-induced injury
    (Ramsammy et al., 1986, 1987, 1988b). The reasons for this apparent
    discrepancy are not known, and the exact role of lipid peroxidation in
    gentamicin nephrotoxicity therefore remains unclear.

         Polyaspartic acid has recently been shown to protect rats
    completely from developing aminoglycoside nephrotoxicity without
    inhibiting proximal tubular cell drug uptake (Williams et al., 1986;
    Gilbert et al., 1989; Ramsammy et al., 1989b).  In vitro studies
    suggest that the protective effect of polyaspartic acid is due to the
    ability of this polyanionic peptide to bind the cationic
    aminoglycosides, thereby preventing these drugs from interacting
    electrostatically  with various targets, presumably anionic
    phospholipids, within the cell.

    5.1.3.2  Cephalosporins

         The nephrotoxicity of cephalosporins was first noted when the
    drugs were used in combination with aminoglycosides, but it is now
    recognized that cephalosporins, particularly cephaloridine, may
    produce degeneration and necrosis of proximal tubular lining cells and
    acute renal failure. It has been suggested that the cellular toxicity
    is the result of metabolic activation of the five-member thiophene
    ring present in cephalothin and cephaloridine, the only two
    cephalosporins that seem capable of producing dose-dependent direct
    nephrotoxicity (Mitchell et al., 1977).  More recently, it has been
    suggested that lipid peroxidation (Goldstein et al., 1989) and direct
    mitochondrial toxicity may be involved in the mechanisms of
    cephaloridine nephrotoxicity.  Necrosis occurs when the concentration
    exceeds 1000 mg/kg wet tissue. The correlation between dose and
    response, as well as the localization of the lesion in the proximal
    portion of the nephron, may be explained by a striking
    corticomedullary gradient in tissue concentration. The cellular uptake
    of cephaloridine and nephrotoxicity have been modified or eliminated
    in experimental animals by pretreatment with either probenecid or
     p-aminohippuric acid. The mechanisms of toxicity are complex (Wold
    et al., 1979; Tune & Fravert, 1980; Tune, 1986; Goldstein et al.,
    1987; Tune et al., 1988).

    5.1.3.3  Amphotericin B

         The increasing use of immunosuppressive therapy and the attendant
    systemic mycotic infections have resulted in an increase in the

    administration of amphotericin B. This drug is almost always
    associated with some degree of toxicity to the distal renal tubule and
    accompanying acidosis, hypokalaemia, and polyuria (Butler, 1966;
    Douglas & Healy, 1969). Reduced renal blood flow and glomerular
    filtration rate may also occur. Pretreatment of experimental animals
    with furosemide or sodium protects against decreases in renal plasma
    flow and glomerular filtration rate immediately following amphotericin
    B treatment. Salt loading  also  protects  against 
    amphotericin-induced decreases in renal plasma flow and glomerular
    filtration rate upon chronic drug administration in rats. However, it
    is important to note that tubular toxicity in these studies was not
    ameliorated by salt loading (Tolins & Raij, 1988). These data suggest
    that the tubular toxicity of amphotericin B is not secondary to renal
    vasoconstriction and ischaemia.

    5.1.3.4  Tetracyclines

         The nephrotoxicity of tetracycline incited considerable interest
    in the early 1960s, shortly after its introduction.  People,
    particularly children, developed a reversible proximal tubular
    dysfunction after receiving outdated drugs. The nephrotoxicity was
    found to be due to a degradation product, anhydro-4-epitetracycline.
    The problem has disappeared with the substitution of citric acid for
    lactose as a vehicle (Curtis, 1979).

         Other rare effects of tetracycline that have been reported are
    impairment of renal-concentrating ability by
    demethylchlorotetracycline  and  occurrences of  acute interstitial
    nephritis after minocycline treatment. More important to current usage
    is the awareness that the serum half-life of the two most commonly
    used drugs, tetracycline and oxytetracycline, is greatly prolonged in
    renal failure, and that the anti-anabolic effect of the tetracyclines,
    which inhibit the incorporation of amino acids into protein, may
    further contribute to negative nitrogen balance and uraemia by raising
    blood urea  nitrogen (Curtis, 1979).

    5.1.4  Penicillamine

         Penicillamine (3,3-dimethylcysteine) was first used clinically as
    a copper-chelating agent to treat Wilson's disease. Because of the
    drug's potential for decreasing collagen formation, its use has been
    extended to a number of clinical disorders in which fibrosis is a
    major component,  such  as rheumatoid  arthritis, pulmonary fibrosis,
    and liver disorders.

         It has been suggested that the drug acts by reducing disulfide
    linkages.  This inhibits polymerization  of macromolecules and leads
    to impairment of collagen formation. Use of the drug has been tempered
    by the occurrence of side effects in as many as 30% of patients.  The
    most important side effect (20% of cases) is proteinuria. The
    morphological appearance of the glomerular lesions is typically that
    of perimembranous glomerulonephritis with segmental subepithelial

    immune-complex deposits. These changes are best demonstrated as
    granular immunofluorescent deposits of IgG and C3. Withdrawal of
    penicillamine therapy results in disappearance of the proteinuria and
    repair of the basement membrane changes in 60% of cases (Gartner,
    1980). Immune-complex glomerulonephritis with granular deposits along
    basement membrane and in the mesangium can be produced experimentally,
    confirming the role of an immunological mechanism in the pathogenesis
    of the nephropathy. In addition the drug is associated with the
    development of Goodpasture's syndrome with linear glomerular basement
    membrane deposits.

    5.1.5  Lithium

         Lithium salts, mainly lithium carbonate, have been used for 40
    years to prevent relapses of maniac-depressive illness. Impaired renal
    ability to acidify and concentrate urine is a common finding among
    patients given lithium (Batelle et al., 1982). It is usually regarded
    as a minor side-effect of the drug, i.e. a pharmacologically induced
    physiological impairment of distal tubules and collecting ducts,  such
    a target-selective effect usually being reversible after
    discontinuation of the therapy. Since the polyurea is resistant to
    ADH, the effect has been characterized as resembling nephrogenic
    diabetes insipidus (Bendz, 1983).

         Although several case reports of lithium-induced chronic renal
    insufficiency have been published (Hestbech et al., 1977; Hansen et
    al., 1979, 1981; Kincaid-Smith et al., 1979; Cohen et al., 1981;
    Walker et al., 1982,1986; Ottosen et al., 1984), the overall evidence
    suggesting progressive renal damage in patients taking lithium is
    rather limited because of methodological weaknesses in human studies
    (Lippmann, 1982). However, animal studies support the view that
    long-term treatment with lithium salts may lead to tubulo-interstitial 
    nephropathies. Experimentally induced focal fibrosis, tubular atrophy,
    and cystic dilatation of distal tubules were obtained by exposing
    animals to toxic doses (Ottosen et al., 1984; Walker et al., 1986).

         In addition to tubular effects, the occurrence of nephrotic 
    syndrome in psychiatric patients has been attributed to long-term
    treatment with lithium salts (Richman et al., 1980; Depner, 1982).
    Thus, although case reports do not constitute evidence, there is some
    indication that lithium may adversely affect other segments along the
    nephron. A proportion of cases, ranging from 0 to 50% (median 8%) of
    patients on lithium, may eventually develop chronic renal
    insufficiency, evolving towards end-stage renal disease (Cohen et al.,
    1981). Such an increased risk may still be regarded as acceptable,
    especially when compared to the benefits of such a therapeutic
    approach to serious psychiatric problems. Thus, fear of renal disease
    may not require the therapy to be stopped. However, a close monitoring
    of renal function is strongly recommended. Once-daily dosing to
    maintain serum lithium levels at the lower therapeutic range, i.e.
    0.4-0.7 mEq/litre, is advisable. Furthermore, it is critical to avoid

    salt depletion, which can disturb the equilibrium of serum lithium and
    induce acute intoxication.

    5.1.6  Urographic contrast media (UCM)

         Radiographic procedures are normally safe, but a small proportion
    of patients subsequently experience a transient or, very rarely, a
    permanent decline in renal function (Cedgard et al., 1986). There are
    various predisposition factors such as dose, age, multiple utilization
    of UCM, dehydration, diabetes (Taliercio et al., 1986), multiple
    myeloma (Harkonen & Kjellstrand, 1981), hypertension, atherosclerosis,
    prior kidney or liver diseases, the co-administration of nephrotoxic
    drugs, and kidney transplantation. Prospective studies suggest that
    diabetes  per se is  not a risk factor when matched for pre-existing
    renal disease (Teruel et al., 1981; D'Elia et al., 1982). This
    highlights pre-existing renal disease as a major risk factor. In view
    of the fact that there are several million radiological procedures
    each year, the number of patients at risk of developing adverse
    effects from the administration of contrast media is significant. Up
    to 10% of cases of acute renal failure in hospitalized patients may be 
    due to intravascular urographic  contrast medium administration (Hou
    et al., 1983).

         The cause of such renal injury is not well understood, but
    hyperosmolality (e.g., with meglumine diatrizoate) has been claimed to
    be an important factor in renal damage (Forrest et al., 1981). New
    low-osmolar urographic contrast media (such as iopamidol) are being
    introduced, some of which are isotonic with plasma, but a progressive
    increase in the incidence of ARF from 0-12% up to 100% in high-risk
    patients has been reported (Eisenberg et al., 1980). About 65% of ARF
    follow intravenous urography, and 30% follow arteriography. The rest
    are associated with computerized tomography (Hanaway & Black, 1977;
    Harkonen & Kjellstrand, 1979; Fang et al., 1980).  The reported
    increase in UCM-induced ARF could be due to better monitoring and
    awareness, higher health standards, or a prolonged survival of
    patients with critical illnesses (who would then be more prone to
    multiple X-ray contrast media examinations), or could represent other
    types of nephrotoxicity.

         The pathophysiology of UCM-induced ARF is unclear, but may
    involve renal ischaemia and haemodynamic effects on glomerular
    function and/or intrarenal flow distribution. Several hormonal systems
    may be activated prior to and/or during ARF (Caldicott et al., 1970;
    Chou et al., 1974; Katzberg et al., 1977). Thus, the effects of
    contrast media on kidney function continue to be conflicting and
    represent both glomerular and tubular dysfunctions (Milman & Gottlieb,
    1977; Rahimi et al., 1981; Teruel et al., 1981; Khoury et al., 1983). 
    It has been suggested, but not confirmed, that non-ionic low-osmolal
    contrast media have reduced nephrotoxicity (Gale et al., 1984;
    Spataro, 1984; Smith et al., 1985; Cedgard et al., 1986; Cavaliere et
    al., 1987).

         Attempts to induce radiocontrast nephrotoxicity in animals have
    led to inconclusive or contradictory results. Intact  hydrated rats
    with or without experimentally induced acute renal failure do not
    develop radiocontrast nephrotoxicity (McIntosh, et al., 1975; Moreau
    et  al., 1980). Transient reductions in glomerular filtration rate and
    renal blood flow have been reported immediately following
    radiocontrast injection in rats and dogs (Norby & DiBona, 1975;
    Cunningham et al., 1986; Katzberg et al., 1986), but rarely has acute
    renal failure been studied or documented in the intact animal
    following these acute measurements. Nephrotoxicity may occur, however,
    when the radiocontrast agent is given in association with experimental
    manoeuvres designed to reduce renal function. These include repeated
    dehydration with furosemide injections, renal ischaemia (Schultz et
    al., 1982), and acute renal failure induced by mercuric chloride or
    glycerol (McLachlan et al., 1972).

    5.1.7  Anticancer drugs

    5.1.7.1  Cisplatin

         Cisplatin ( cis-diamminedichloroplatinum II) has become the
    chemotherapeutic agent of choice in the treatment of several solid
    tumours, particularly testicular and ovarian cancers (Einhorn &
    Donohue, 1977). Unfortunately cisplatin is also one of the most toxic
    anticancer drugs, its dose-limiting toxicity being nephrotoxicity 
    (Madias  & Harrington, 1978; Goldstein & Mayor, 1983; Safirstein et
    al., 1986). Despite the use of optimal methods for administering
    cisplatin, such as the use of active hydration (Cvitkovic et al.,
    1977) or sodium chloride as the vehicle (Ozols et al., 1984),
    approximately 30% of patients will manifest nephrotoxicity.

         Early clinical trials of cisplatin in cancer patients showed a
    striking incidence of persistent azotaemia and acute renal failure
    (Rossof et al., 1972; Lippman et al., 1973). In later studies serum
    creatinine levels increased within 6-7 days of treatment, and then
    apparently returned to pre-treatment levels by approximately 3 weeks
    (Hayes et al., 1977). Similar results were seen following the
    injection of cisplatin into rats (Ward & Fauvie, 1976; Chopra et al.,
    1982). Thus cisplatin-induced  nephrotoxicity initially appeared to be
    an acute reversible condition. However, more recent findings suggest
    that cisplatin causes a permanent reduction in GFR (Dentino et al.,
    1978; Meijer et al., 1983; Fjeldborg et al., 1986), which may indeed
    be progressive in nature (Groth et al., 1986; Jaffe et al., 1987).

         Hypomagnesaemia is frequently noted in patients receiving
    cisplatin (Buckley et al., 1984; Vogelzang et al., 1985), and is
    associated with inappropriately high levels of urinary excretion of
    magnesium. This deficiency in magnesium  leads to hypokalaemia and
    hypocalcaemia. This selective renal loss of magnesium is not unusual
    and may be even more common than other renal abnormalities as an
    expression of cisplatin nephrotoxicity.

         Light microscope studies of human kidneys have revealed focal
    acute tubular necrosis, affecting primarily the distal and collecting
    tubules, with dilatation of convoluted tubules and cast formation
    (Gonzalez-Vitale et al., 1977). More recently, Tanaka et al. (1986)
    reported sporadic degenerative lesions, necrosis, and regenerative
    changes in the S2 and S3 regions of the proximal tubule and also in
    the distal tubule and collecting duct. The glomeruli and vasculature
    appeared uninvolved.  These observations are somewhat different to
    those seen in the rat, where cisplatin-induced damage is largely
    confined to the S3 segment of the proximal tubule, located in the
    outer stripe of the outer medulla (Chopra et al., 1982). With
    increasing time cystic tubules develop in this region (Dobyan, 1985). 
    However, these cysts have not been reported clinically.

         The activity of a number of urinary enzymes, including alanine
    aminopeptidase,  N-acetyl-ß-D-glucosaminidase, leucine aminopeptidase
    and ß-glucurinidase,  has been shown to be elevated as early as 36-48
    h after cisplatin treatment (Kuhn et al., 1978; Jones et al., 1980).
    ß2-micro-globulin excretion has also been shown to be transiently
    increased after cisplatin treatment (Daugaard et al., 1988a,b). It is
    of interest to note that this proteinuria (involving proteins of low
    relative molecular mass), predominantly tubular in origin, was
    transient, whereas a persistent proteinuria consisting of proteins of
    high relative molecular mass, such as albumin and IGg, and glomerular
    in origin was seen after the completion of cisplatin treatment.

         The pathophysiology of the GFR reduction remains ill defined. It
    is clear that cisplatin produces an acute, mainly proximal, tubular
    functional impairment within hours of administration (Daugaard et al.,
    1988a,b).  It has been suggested that the former is a consequence of
    the latter. Thus, Groth et al. (1986) attributed the chronic reduction
    in GFR to increased intratubular pressure within damaged tubules. The
    glomerular proteinuria reported by Daugaard et al. (1988a) suggests
    that cisplatin may directly damage glomeruli. Cisplatin-induced
    glomerular lesions have been reported in the pig (Robbins et al.,
    1990).

    5.1.7.2  Adriamycin

         The anthracycline antibiotic Adriamycin is widely used in
    clinical oncology to treat several cancers, including breast
    carcinoma, malignant lymphomas, and sarcomas (Blum & Carter, 1974).
    Its clinical use is limited by its cardiotoxicity. Experimentally
    adriamycin has been shown to produce a nephrotic syndrome in rats
    (Young, 1975), rabbits (Fajardo et al., 1980), and pigs (van Fleet et
    al., 1979).

         Rats treated with a single dose of Adriamycin exhibited a marked
    proteinuria evident within several days of treatment (Bertani et al.,
    1982). Maximal levels were seen after approximately 2 weeks, after
    which levels declined but remained significantly above control levels
    10 weeks after treatment. Serum albumin levels were also significantly

    reduced, whereas there was a concomitant hyperlipidaemia (Bertani et
    al., 1986). Adriamycin-induced changes in renal functional parameters
    are less well defined. Litterst & Weiss (1987) reported that BUN and
    serum creatinine values were either unaffected or only minimally
    increased. However, more recent studies indicate significant and
    progressive reductions in GFR (Hall et al., 1986). Single nephron
    glomerular filtration rate is reduced due to a decreased
    ultrafiltration coefficient (Michels et al., 1983).

         Morphological damage is first seen in the glomerulus;
    ultrastructural examination reveals extensive damage to the glomerular
    epithelial cells occurring within 36-48 h of injection (Bertani et
    al., 1982). This leads to the eventual loss of the foot processes.
    Light microscope studies reveal the characteristic presence of
    vacuoles in the glomeruli; with time progressive glomerulosclerosis is
    seen. Associated with these glomerular changes are tubular changes;
    these consist of dilated tubules filled with casts, predominantly in
    the outer stripe of the outer medulla, and atrophic tubules associated
    with areas of interstitial fibrosis. It appears that Adriamycin
    primarily damages the glomeruli and that the tubulo-interstitial
    damage results from the proteinuria, which induces cast formation and
    interstitial inflammatory reaction. Fajardo et al. (1980) reported
    that juxtamedullary glomeruli were more sensitive than cortical
    glomeruli; micropuncture studies in the Munich Wistar Fromter rat
    confirm this observation (Soose et al., 1988).

         Renal toxicity in patients appears rare, although there has been
    a report of acute renal failure following Adriamycin-treatment (Burke
    et al., 1977). This may reflect species differences in sensitivity or
    may reflect the use of inappropriate test protocols for detecting
    renal damage.

    5.1.8  Immunosuppressive agents

    5.1.8.1  Cyclosporin A

         Cyclosporin A has been widely used for preventing organ rejection
    after transplantation and in auto-immune diseases, but it is highly
    nephrotoxic in the clinical situation (Kostakis et al., 1977; Calne et
    al., 1978; Powles et al., 1978). Clinically, cyclosporin
    nephrotoxicity has been reported as an acute reversible renal
    dysfunction, an acute vasculopathy (thrombotic microangiopathy),
    and/or a chronic nephropathy with interstitial fibrosis (Mihatsch et
    al., 1985; Palestine et al., 1986). After cyclosporin A treatment
    there is an inverse linear relation between GFR and the severity of
    the lesions. During the first 6 months of treatment, renal fibrosis in
    patients given high doses of cyclosporin A shows a dosedependent
    progression of increased severity (Klintmalm et al., 1981). The rat
    model of cyclosporin A nephrotoxicity may adequately represent the
    acute condition in man (Mihatsch et al., 1985, 1986) but does not
    represent that seen in the clinical situation. Although rats dosed
    with cyclosporin also develop renal surface changes that correspond to

    focal areas of collapsed proximal tubular regions with subcapsular
    fibrosis, degenerating tubular epithelium and thickening of the
    basement membrane, the chronic striped fibrosis and arteriolar lesions
    have not been reproduced experimentally.

         Some animal data have shown increased blood urea nitrogen and
    creatinine, brush-border and lysosomal enzyme leakage, and
    vacuolation, necrosis, and regeneration of P3 cells in rats (Dieperink
    et al., 1983, 1985; Ryffel et al., 1983, 1986; Murray et al., 1985;
    Dieperink, 1989). However, most studies find no necrosis or enzymuria
    despite profound reductions in GFR. Functional changes in animals and
    man given cyclosporin A are similar. They represent haemodynamic
    changes such as an increased renal vascular resistance (Murray et al.,
    1985), proximal fractional  reabsorption (Dieperink et al., 1983,
    1985; Dieperink, 1989), and renal blood flow, plasma flow, and GFR
    decrease. The reduction in renal perfusion and filtration has been
    prevented experimentally by vasoactive alpha-adrenergic  antagonists
    and renal denervation (Murray et al., 1985), but this has not been
    established in humans. Cyclosporin A appears to have a direct
    preglomerular vasoconstriction effect, decreasing the ultrafiltration
    pressure and increasing proximal fractional reabsorption. Presumably,
    tubular flow rates and end proximal tubular delivery decrease, and,
    due to varying tubular hypoperfusion, there is a focal tubular
    collapse, and degeneration and peritubular interstitial fibrosis
    develop. The precise relationship between renal vasoconstriction and
    chronic tubulo-interstitial pathology is poorly understood.

    5.1.9  Heroin

         About 1% of heroin addicts  develop haematuria, proteinuria, or
    the nephrotic syndrome (Cunningham et al., 1980).  Morphologically, 
    the renal lesion has been described as focal sclerosing
    glomerulonephritis. In the absence of proliferative lesions or immune
    deposits, a direct toxic effect of heroin or even a contaminant or
    solvent employed in the administration of heroin has been suggested as
    the major pathogenetic mechanism. However, heroin-related increases in
    IgM titres have been regarded as evidence that an immunological
    mechanism may play some role in this disorder.

    5.1.10  Puromycin aminonucleoside

         Direct toxicity of the puromycin aminonucleoside to glomerular
    components may also occur and may be a factor in renal failure.
    Experimental studies of the effects of puromycin have provided
    considerable basic information regarding the pathogenesis of direct
    chemical injury to glomerular structures. The effects are limited to
    rats and monkeys. Epithelial cells become swollen, and there is an
    increase in lysosome and pinocytotic activity, fusion and loss of foot
    processes, and a reduction in the number of filtration slits. As the
    lesion progresses, epithelial cells become detached, leaving "naked"
    basement membrane in direct contact with Bowman's capsule, which may
    account for the severe proteinuria (Caulfield et al., 1976). It is

    suggested that as the lesion progresses there is an increase in
    basement membrane synthesis, mesangial cell proliferation and fusion,
    and crescent formation leading to the light microscope appearance of
    focal glomerular sclerosis (Gartner, 1980). The glomerular lesions
    produced by puromycin do not appear to invoke an immunological
    response, so that the resulting alterations are entirely related to
    the direct toxicity of the drug.

    5.2  Chemicals

         The diversity of organic molecules is such that there are
    chemicals that are now known to adversely affect each part of the
    kidney. This section will examine those chemicals that do not fit into
    any conventional section on therapeutically used agents.

    5.2.1  Ethylene glycol

         Ethylene glycol, a constituent of antifreeze, is occasionally
    ingested and causes severe acute toxicity to the brain and kidney.
    Acute tubular necrosis is followed by renal failure. Exposure often
    leads to permanent renal damage. A morphological feature of mild
    ethylene glycol toxicity is cytoplasmic vacuolation, which may suggest
    hypokalaemic nephropathy or osmotic nephrosis due to mannitol. Most
    ethylene glycol is excreted unmetabolized, but a small percentage is
    metabolized to oxalic acid. This is accompanied by deposition of
    calcium oxalate crystals in the kidney, which may contribute to a
    persistent inflammatory reaction and interstitial fibrosis. Excessive
    urinary excretion of oxalate and crystal formation may also be seen
    following administration of halogen-containing anaesthetic agents,
    particularly methoxyflurane and halothane (Roxe, 1980). Acute ethylene
    glycol toxicity is treated with ethanol, which competes as a substrate
    for alcohol dehydrogenase (Peterson et al., 1981).

    5.2.2  Organic chemicals and solvents

    5.2.2.1  Volatile hydrocarbons

         Volatile hydrocarbons, particularly chlorinated compounds such as
    carbon tetrachloride and trichloroethylene, may produce glomerular
    lesions leading to nephrotic syndrome and renal failure. The
    relationship of volatile hydrocarbon exposure to the development of
    glomerulo-nephritis in populations is not clear. It has been found
    that among patients with glomerulonephritis there are more with a
    history of exposure to hydrocarbon solvents than would be expected.
    Attempts to reproduce in rats the glomerular lesions observed in
    patients have only been partially successful. Solvent-exposed rats had
    increased proteinuria and glomerular sclerosis, but proliferative
    lesions and significant immune deposits were not observed (Zimmerman
    & Norbach, 1980). Of 15 patients studied in Sweden with
    post-streptococcal glomerulonephritis, 6 had a history of brief
    exposure to organic solvents before the development of their disease.
    This suggested to these investigators that solvent exposure may

    influence the outcome of an infection with streptococci. Prior
    exposure to hydrocarbon-containing solvents has been identified in a
    number of patients with Goodpasture's syndrome (Gartner, 1980).  Apart
    from the pulmonary manifestation of cough, shortness of breath, and
    haemoptysis, there may be haematuria and proteinuria. Renal morphology
    consists of a proliferative glomerulonephritis with IgG and C3 in the
    glomerular basement membrane.

         Studies conducted in Sweden (Askergren 1981; Askergren et al.,
    1981) and in Belgium (Viau et al., 1987) have reported a slight
    increase in the urinary excretion of albumin in groups of workers
    exposed to industrial solvents, particularly styrene. This effect
    probably reflects an enhanced glomerular permeability since the
    urinary output of markers of proximal tubular function
    (ß2-micro-globulin, retinol-binding protein) was not affected. In
    Italy, Franchini et al. (1983) also reported slight renal disturbances 
    in  workers  occupationally exposed  to solvents. These effects
    consisted of enhanced urinary excretion of total proteins, lysozyme,
    and ß-glucuronidase and pointed to a tubular lesion, since they were
    not accompanied by a rise in albuminuria. It is at present impossible
    to relate nephrotoxic effects reported in these studies to exposure to
    one solvent or one class of solvents, although styrene has been
    incriminated by Askergren et al. (1981). One must also recognize that
    the renal effects reported in these studies are mild and do not appear
    to correlate with indices of exposure to solvents.

         An auto-immune mechanism following chronic exposure is probably 
    responsible for the glomerular lesions. In patients who have
    Goodpasture's syndrome, the primary site of damage may be the alveolar
    basement membrane of the lung, which is damaged by inhalation of the
    solvents, and antibodies to altered alveolar basement membrane may
    cross-react with glomerular basement membrane. Alternatively,
    auto-immunity may follow direct toxic injury to renal tubular or
    glomerular structures. Acute exposure to these solvents does produce
    acute tubular necrosis, and it is likely that prolonged exposure to
    low levels that do not result in cell necrosis produces cell injury
    sufficient to damage renal cell membranes and provide the antigen for
    the immune reaction (Gartner, 1980).

    5.2.2.2  Chloroform

         In vitro exposure to chloroform has been shown to produce
    toxicity in kidney slices from male but not from female mice (Smith &
    Hook, 1984).  Furthermore, 14C-labelled chloroform was metabolized
    to 14CO2, and the radioactivity was covalently bound by cortical
    microsomes from male but not female mice. The  in vitro metabolism of
    chloroform by male, but not female, renal slices is consistent with
    reduced susceptibility of female mice to  in vivo chloroform
    nephrotoxicity (Smith et al., 1983; Smith et al., 1984). Metabolism is
    dependent on oxygen, a NADPH-regenerating system, incubation time,
    microsomal protein concentration, and substrate concentration, and is
    inhibited by carbon monoxide (Smith & Hook, 1984). The negligible

    degree of chloroform metabolism and toxicity in female mice is
    consistent with a lower renal cytochrome P-450 concentration and
    activity in female mice than in males (Smith & Hook, 1984).
    Pretreatment of rabbits with phenobarbital, a renal cytochrome P-450
    inducer in this species, enhances the toxic response of renal cortical
    slices to chloroform  in vitro (Bailie et al., 1984). The rate at
    which deuterated chloroform is metabolized by the liver to phosgene is
    approximately half that of chloroform. Deuterated chloroform is also
    less hepatotoxic that chloroform since the C-D bond is stronger than
    the C-H bond. These data suggest that cleavage of the C-H bond is the
    rate-limiting step in the activation of chloroform. Deuterated
    chloroform is also less toxic to the kidney than chloroform
    (Ahmadizadeh et al., 1981; Branchflower et al., 1984). This deuterium
    isotope effect on chloroforminduced nephrotoxicity suggests that the
    kidney metabolizes chloroform in the same manner as the liver, e.g.,
    by oxidation to phosgene. Indeed, rabbit renal cortical microsomes
    incubated in media supplemented with L-cysteine metabolize
    14C-labelled  chloroform to radioactive phosgene-cysteine
    2-oxothiazolidine-4-carboxylic acid (Bailie et al., 1984). These
     in vitro data collectively support the hypothesis that mouse and
    rabbit kidneys biotransform chloroform  to a metabolite (phosgene)
    that mediates nephrotoxicity.

    5.2.2.3  Halogenated alkenes

         The nephrotoxin  S-(1,2-dichlorovinyl)-L-cysteine (DCVC) is
    formed by trichloroethylene extraction of proteinaceous substances and
    was first identified in extracted animal food (McKinney et al., 1957,
    1959). It has been widely used as a model compound in nephrotoxicity
    studies. DCVC is accumulated in the proximal tubules by an active
    carrier system for organic anions (Elfarra et al., 1986a,b). It is
    then activated by cysteine conjugate ß-lyase to a reactive thiol
    (Bhattacharya & Schultze, 1967) and causes tubular damage (Terracini
    & Parker, 1965). DCVC is a potent specific nephrotoxin, which produces
    proximal tubular damage  in vivo and  in vitro (Elfarra et al.,
    1986a,b; Lash & Anders, 1986; Lash et al., 1986). In vivo DCVC causes
    its primary lesion in the straight segment (S-3) of the proximal
    tubule, and the molecule is also cytotoxic both for primary cultures
    of proximal tubular cells and for cell lines derived from this region
    of the nephron. There is a close correlation between the  in vivo and
     in vitro effects of this compound with regards to its metabolism and
    effects on cells (Hassall et al., 1983).

         Cysteine conjugates such as DCVC are metabolized by cysteine
    conjugate ß-lyase to their ultimate toxic species i.e. pyruvate,
    ammonia, and a reactive thiol (Anderson & Schultze, 1965). This
    reaction plays a role in the nephrotoxicity of DCVC (Lash et al.,
    1986). ß-lyase has been found to predominate in cytosolic and
    mitochondrial fractions (Lash et al., 1986) and has a requirement for
    pyridoxal phosphate. The enzyme activity can be inhibited by pyridoxal
    phosphate inhibitors such as amino-oxyacetic acid and propargylglycine
    (Elfarra et al., 1986a). In addition to monitoring enzyme activity,

    renal cortical slices can be utilized to assess the regulation of
    enzyme activity and the resultant effects on toxicity.

         There is a greater sensitivity to DCVC-induced kidney damages in
    the adult mouse than there is in the newborn. Similar findings have
    been reported using cephaloridine, where the newborn animal is more
    resistant to nephrotoxicity than the adult rabbit (Tune, 1975).  These
    finding for DCVC differ from those for hexachlorobutadiene (Kuo &
    Hook, 1983; Lock et al., 1984), where nephrotoxicity is greater in the
    young rat and mouse than in the adults.

         Chlorotrifluoroethylene is a potent nephrotoxin (Potter et al.,
    1981) and is metabolized by hepatic cytosolic and microsomal
    glutathione  S-transferases to
    S-(2-chloro-1,1,2-trifluoroethyl)glutathione (Dohn et al., 1985a),
    which is nephrotoxic in rats and cytotoxic in isolated rat kidney
    proximal tubular cells (Dohn et al., 1985b). The corresponding
    cysteine  S-conjugate,  S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine
    (CTFC) is also nephrotoxic in rats and cytotoxic in isolated kidney
    cells, and its bioactivation is dependent on metabolism by renal
    ß-lyase (Dohn et al., 1985b). Pyruvate and hydrogen sulfide have been
    identified as metabolites of CTFC (Banki et al., 1986; Lash et al.,
    1986).

    5.2.2.4  Hydrocarbon-induced nephrotoxicity

         Inhalation of unleaded gasoline for 2 years produced renal
    tumours (adenomas and adenocarcinomas) in male Fischer-344 rats but
    not in female rats or mice of either sex (Kitchen, 1984; MacFarland,
    1984; Mehlmann et al., 1984).  Subchronic inhalation exposure
    increased protein (hyaline) droplets in proximal convoluted tubules
    (Halder et al., 1984), accumulated casts at the cortico-medullary
    junction and single cell necrosis and regeneration of the nephron
    (Short et al., 1986) in male rats. When different fractions of
    unleaded gasoline were screened for their specific effect on the male
    rat kidney, it was found that the branched-chain saturated hydrocarbon
    components (used as anti-knocking agents) caused hyaline droplet
    formation. A number of chemicals, such as 2,2,4-trimethyl-pentane
    (Phillips & Egan, 1984a,b; Halder et al., 1985; Viau et al., 1986a),
    decalin (Alden et al., 1984), 1,4-dichlorobenzene (NTP, 1987), and
     p-dichlorobenzene  (NTP, 1986; Bomhard et al., 1988), have now been
    shown to cause such hyaline droplets in male rats. Although these
    chemicals cause minimal renal functional impairment, a protein droplet
    nephrosis develops, progressing to mild tubular degeneration,
    necrosis, and regeneration after several weeks of treatment (Phillips
    & Cockrell, 1984a,b).

         Male mice excrete a sex-related protein, which results in a
    urinary protein level 2.5 to 3 times that of female mice. However, the
    male mouse sex-associated urinary protein hydrolyses readily and thus
    does not accumulate in the proximal tubule (Alden et al., 1984; Alden,
    1989). In humans, it has recently been reported that protein 1 (an

    alpha2-microglobulin  of about 20 000 Daltons) has a sex-linked
    behaviour just like the androgen-dependent alpha2u-globulin. Protein
    1 is excreted in greater amounts in the urine of males after puberty.
    In the age group 15 to 20 years, its concentration in the urine of
    males is on average fifty times higher than that in the urine of
    females (Bernard et al., 1989). The relevance of this observation in
    humans is unknown.

         The basis for the marked sex dependence and species difference in
    the development of hyaline droplet deposition in male rats relates to
    the fact that they excrete the sex-hormone-related and therefore
    male-specific protein alpha2u-globulin  (Stonard et al., 1986; Loury
    et al., 1987; Olson et al., 1987). This is also species specific to
    the rat and has not been reported in any other commonly used animals
    or man. alpha2u-Globulin  is synthesized by the male rat liver and is
    an important constituent of the physiological proteinuria in adult
    male rats. At maturity the total urinary protein is 20-30%
    alpha2u-globulin  and 10% albumin. At 160 days of age the excretion of
    albumin and total urinary proteins is markedly increased.  By one
    year, albumin represents nearly 60% of the total protein while
    alpha2u-globulin  is less than 10%.  This reversal in relative content
    may be the consequence of a progressive glomerulonephrosis, associated
    with an apparent spontaneous accumulation of hyaline droplets. The
    nephrotic condition  may be the consequence of the burden of excreting
    alpha2u-globulin. Its early onset is sex dependent; female rats do not
    exhibit the proteinuria until a later age.  In both aging and the
    response to hydrocarbons, the common  pathological factor may be the
    accumulation of alpha2u-globulin, via a susceptible pathway not shared
    with other proteins (Neuhaus, 1986; Stonard et al., 1986).

         The chronic regenerative response subsequent to moderate proximal
    tubular damage in the kidney of male rats exposed to petroleum
    hydrocarbons may be an important stimulus in renal tumour formation
    caused by this group of chemicals (Short et al., 1986). Loury et al.
    (1987) have shown a 5- to 8-fold increase in S phase of renal cells
    induced by unleaded gasoline, and the renal proliferative effects of
    2,2,4-trimethylpentane are localized to the S2 segment  of the
    proximal tubule of the male rat (Short et al., 1986).

         Administration of 2,2,4-trimethylpentane to male rats produces a
    dose-related increase in the concentration of
    2,2,4-trimethylpentane-derived radiolabel in the kidney, which appears
    to parallel the dose-related accumulation of 2u-globulin (Stonard et
    al., 1986; Charbonneau et al., 1987). The reversible binding of a
    metabolite of 2,2,4trimethylpentane to alpha2u-globulin in the male
    rat kidney (Lock et al., 1987) is thought to alter endocytosis or
    lysosomal handling of the alpha2u-globulin-2,2,4-trimethylpentane
    metabolite complex. This may increase cell turnover of the S2 cells
    via lysosomal enlargement and/or instability, leading to cell death.
    It appears that lysosomal catabolism of the
    2,4,4-trimethyl-2-pentanol-alpha2u-globulin complex compound to

    alpha2u-globulin causes lysosomal  protein overload, resulting in 
    cell necrosis (Swenberg et al., 1989).

         There are other sex-related differences in the handling  of
    2,2,4-trimethylpentane by rats. Female rats rapidly metabolize
    2,2,4-trimethylpentane and excrete it in urine, while in male rats the
    compound is eliminated more slowly and is retained in the kidneys
    (Kloss et al., 1985). Recent studies have also shown that
    2,2,4-trimethylpentane is metabolized in male and female rats to
    trimethyl-pentanols, pentanoic acids, and hydroxypentanoic acids
    (Olson et al., 1986; Charbonneau et al., 1987).

    5.2.2.5  Bipyridyl herbicides

         Paraquat is a potent bipyridyl herbicide that has multiple organ
    effects. The kidney is frequently involved in serious cases of
    paraquat poisoning (WHO, 1984). The compound is actively secreted by
    the organic cation transport in the proximal tubule. Renal
    histological examinations in a variety of animals exposed to paraquat
    show vacuolation of the proximal convoluted tubules and proximal
    tubular cell necroses (Lock, 1979; Lock & Ishmael, 1979). Acute
    oliguric renal failure is common in severely poisoned patients.  Less
    severe manifestations include impaired glomerular filtration, which
    often recovers after several days and before the paraquat induces
    severe pulmonary fibrosis. Other renal functional abnormalities
    include proteinuria and haematuria. Tubular damage may be shown by the
    presence of glucosuria or all of the features of the Fanconi syndrome.
    The severity of the acute renal failure is a major determinant of the
    outcome of the poisoning (WHO, 1984).

         The biochemical mechanism of nephrotoxicity has not been fully
    elucidated, but it is assumed to be identical to that seen in other
    tissues. Paraquat undergoes redox cycling in the presence of NADPH and
    oxygen with the generation of superoxide and subsequent development of
    lipid peroxidation and membrane damage. The development of hydroxyl
    radicals results in oxidative damage to nucleic acids, proteins, and
    polysaccharides (Autor, 1977).

         Diquat is another bipyridyl herbicide that produces multiple
    organ toxicity.  It undergoes active tubular secretion by the organic
    cation system in the proximal tubule (Lock, 1979; Lock & Ishmael,
    1979). The histological lesion produced by diquat is necrosis of the
    proximal tubular cells and some distal tubular cells (Lock & Ishmael,
    1979). Human cases of diquat poisoning result in acute renal failure.
    The mechanism of toxicity of diquat appears to be identical to that of
    paraquat (Autor, 1977; WHO, 1984).

    5.3  Mycotoxins

         A high frequency of endemic chronic nephropathy has been
    recognized in localized areas of Bulgaria, Rumania, and Yugoslavia
    since the 1920s. The affected people live in villages in valleys near

    the Danube (Hall & Dammin, 1978; WHO, 1979; Hall, 1982). The
    condition, known as Balkan endemic nephropathy (Fig. 14), is an
    interesting case study of an environmentally related chronic renal
    disease.  The etiology is unknown at present. Mycotoxins, particularly
    ochratoxin A, have been implicated because of similarities with
    disease in animals and identification of the mycotoxin in food (Krogh
    et al., 1977; Pepeljnjak & Cvetnic, 1985; Petkova-Bocharova &
    Castegnaro, 1985) and in human tissues (Hult & Fuchs, 1986) where
    nephrotoxicity is most frequent. Silicates have been suggested because
    of the proximity of villages with affected families to streams and
    rivers containing silicon.

         Although not clearly implicated in BEN, the fungal toxin citrinin
    has been suggested as a causative agent in porcine citrinin
    nephropathy and clearly has nephrotoxic effects in a number of species
    (Berndt & Hayes, 1977; Phillips et al., 1979; Phillips et al.,
    1980a,b; Lockard et al., 1980). Citrinin produces acute tubular
    necrosis primarily of the S1 section of the proximal tubule. It is
    eliminated rapidly by the kidney and only metabolized to the extent of
    10-15%, which suggests that effects are due to the parent compound.
    Little information is available concerning the cascade leading from
    the initial insult to the production of acute tubular necrosis 2-4
    days after its administration. Citrinin has been shown to exert a
    synergistic effect on ochratoxin A toxicity in animal models. This is
    important because the same fungal species that synthesize ochratoxin
    A also produce citrinin. This was clearly demonstrated by the presence
    of citrinin in 19 out of 21 food samples contaminated with ochratoxin
    A. Both genetic and environmental factors, such as exposure to
    ochratoxin A, appear to be involved in BEN and the associated renal
    tract tumours (Castegnaro & Chernozemsky, 1987). In one endemic area
    in Bulgaria the relative risk of patients with BEN developing urinary
    tract tumours is 90-fold greater than in people from non-endemic areas
    (Castegnaro & Chernozemsky, 1987). Inhabitants of the 15 villages in
    the Vratza region of northern Bulgaria have a 30-40% mortality rate
    from chronic nephropathy, while urinary tract tumours comprise 25-30%
    of all neoplasms in males and females in these geographic areas
    (Markovic, 1972). In recent studies ochratoxin A has been found to
    induce  renal adenomas and carcinomas  both in mice (Kanisawa &
    Suzuki, 1978; Bendele et al., 1985) and in rats (NTP, 1988).  Frequent
    metastases, mainly to the lung, were found in the rat study. The
    target of ochratoxin nephrotoxicity has been reported to be the S2 
    and S3 nephron segments (Jung et al., 1989).

         Thus the animal and human data indicate that ochratoxin A is a
    risk factor for toxic nephropathies and in the etiology of human
    nephropathy and associated renal tumours.

    5.4  Silicon

         An association between occupational exposure to free silica
    (SiO2) and chronic nephropathy has been suspected for several years,
    but the number of reported cases is few. Clinically, lung fibrosis is

    FIGURE 14a

    FIGURE 14b

    the primary problem, but in an early study from Italy chronic renal
    failure was found in 40% and proteinuria in 20% of 20 patients with
    chronic silicosis. The renal silicon content of patients with
    proteinuria and chronic silicon exposure has been shown to be much
    higher than the normal level, and there appears to be a direct
    relationship between level of exposure and probability of renal
    disease. Animal studies have demonstrated that silicon is excreted by
    glomerular filtration, and a morphological study of experimental
    animals and human biopsy material has demonstrated silicon deposits in
    subepithelial and subendothelial areas of the basement membrane and in
    epithelial cells. Human biopsy materials show a mild focal or
    segmental proliferative glomerulonephritis and the absence of
    significant immune-complex deposits. These findings suggest a direct
    toxic effect on the glomerulus. These cases also have varying degrees
    of tubular cell degeneration. Animal studies demonstrate a
    dose-related nephropathy that is primarily tubular, with an
    interstitial inflammatory reaction and fibrosis. The proliferative
    glomerular lesions observed in humans are not seen in animals, but
    this difference in response may be related to dose or species
    (Hauglustaine et al., 1980).

    5.5  Metals

         Metals constitute some of the earliest recognized and the best
    investigated nephrotoxins.  X-ray fluorescence gives a clear
    indication of the metal burden that an individual carries.

    5.5.1  Lead

         Lead has been a very common cause of acute or chronic renal
    failure in the past. Acute tubular necrosis has been described
    following accidental or intentional absorption of high doses of lead.
    Cases of chronic renal failure have been reported in adults who
    ingested large amounts of leaded paint during childhood (Queensland,
    Australia), in people who consumed alcohol illicitly distilled in
    lead-containing stills, and in workers with a long history of
    occupational lead exposure (Emmerson, 1973;  Bennett, 1985).

         Several epidemiological studies have consistently reported that
    workers with a heavy industrial exposure to lead experience an
    increased risk of death through chronic renal failure (Cooper &
    Gaffey, 1975; Malcolm & Barnett, 1982; McMichael & Johnson, 1982;
    Davies 1984; Selevan et al., 1985; Cooper et al., 1985). There is also
    some evidence that occult lead poisoning may contribute to renal
    insufficiency in patients with gout and essential hypertension
    (Batuman et al., 1981, 1983; Colleoni & D'Amico, 1986).

         In adults, lead nephropathy occurs as an insidious progressive
    disease characterized by the absence of proteinuria, albuminuria, and
    urinary concentration deficit in its early phases (Wedeen et al.,
    1979). This renal disease can be diagnosed only by functional tests
    (e.g., estimation of GFR on the basis of blood urea nitrogen or

    creatinine clearance). Several cross-sectional studies have attempted
    to detect early renal effects in workers exposed to lead (Hammond et
    al., 1980; Buchet et al., 1980; Verschoor et al., 1987). These studies
    confirm that lead nephropathy in adults, even at an advanced stage
    (i.e. with decreased GFR), cannot be detected by the determination of
    urinary proteins of low or high relative molecular mass (e.g.,
    ß2-microglobulin, albumin). The only marker that seems to respond at
    an early stage of lead nephropathy is the urinary excretion of the
    lysosomal enzyme,  N-acetyl-ß-D-glucosaminidase (NAG) (Verschoor et
    al., 1987). However, the underlying mechanism of this renal effect
    remains to be elucidated. Increased urinary leakage of NAG might
    result from cell damage and exfoliation, but also from a stimulation
    by lead of exocytosis or of the renal activity of the enzyme.

         The renal effects of lead are primarily tubular or
    tubulo-interstitial and they may be both acute  and chronic. However,
    the acute effects of lead differ from those of most of the other
    metals in that cell injury is for the most part reversible and
    necrosis is uncommon. Cells of the proximal tubule are most severely
    affected, and this effect is characterized by a reduction in
    resorptive function leading to a generalized amino-aciduria,
    glycosuria, and hyperphosphaturia. These components of the Fanconi
    syndrome have been observed in children with acute lead toxicity and
    who also have overt symptoms of central nervous system toxicity, and
    in rats exposed to lead. Proximal tubular dysfunction has been more
    difficult to demonstrate in workers with chronic lead nephropathy
    (Goyer & Rhyne, 1973).

         The effects of lead on renal tubular cells and sodium
    reabsorption are less clear. Increase in plasma renin and aldosterone
    while a low-sodium diet is consumed has been observed in a group of
    men with a history of "moonshine" ingestion and occult lead toxicity
    (Sandstead et al., 1970). In contrast, studies on the effects of
    minimally toxic levels of lead exposure in rats showed a reduction in
    plasma renin activity in spite of a significant increase in blood
    pressure (Victery et al., 1982). These differences  may  reflect a 
    difference in time-dose relationship.

         The renal effects of lead may also be influenced by interactions
    with calcium. Decreasing dietary calcium increases lead retention,
    possibly because of a decrease in lead excretion. Increased blood lead
    in children is associated with decreased 1,2,5-dihydroxyvitamin D
    (synthesized in the kidney) and may reflect impaired synthesis
    (Mahaffey, 1980).

         The renal proximal tubular cells of people and experimental
    animals with lead poisoning are characterized morphologically by the
    presence of intranuclear inclusion bodies. In conventional
    paraffin-embedded haematoxylin and eosin-stained sections of renal
    tissue, the inclusions appear as dense, homogeneous, and eosinophilic
    bodies, and at the electon microscope level they have a characteristic
    fibrillary margin around a dense central core. Morphologically they

    are always separate and distinct from the nucleoli and several may be
    found in the same nucleus (Fig. 15).  The inclusion bodies contain a
    protein-lead complex, and they may be isolated by differential
    centrifugation.  The protein is a non-histone protein rich in glutamic
    and aspartic acids and glycine, and may be a mixture of acidic
    proteins with similar physicochemical properties (Moore et al., 1973).
    The origin and nature of the protein has not yet been determined, but
    recent studies of formation of inclusion bodies in renal cell cultures
    suggest that they form initially in the cytoplasm and then migrate
    into the nucleus (McLachlan et al., 1980).  The major fraction of lead
    in the kidney during the acute phase of lead toxicity is bound in the
    inclusion bodies. For this reason, the inclusion bodies have been
    interpreted as serving as an intracellular depot for lead.
    Nevertheless, proximal renal tubular cells during the acute phase of
    lead toxicity are usually swollen, and the mitochondria show a
    decrease in matrical granules and altered cristae. Functional studies
    of mitochondria show reduced respiration and oxidative
    phosphorylation. Lysosomes do not seem to have a role in sequestering
    intracellular lead.

         Chelation therapy following lead toxicity produces a marked
    increase in lead excretion. This is accompanied by reversal of the
    acute morphological effects of lead on proximal renal tubular cells,
    loss of inclusion bodies from nuclei, and restoration of normal renal
    cell morphology and function (Goyer & Wilson, 1975).

         Both experimental animals and people with chronic exposure to
    lead may develop a progressive interstitial nephropathy. In laboratory
    animals, progression from acute tubular to chronic tubulo-interstitial
    disease may be followed as a continuum.  An increase in chronic
    interstitial renal disease has been reported in workers with long
    histories of occupational exposure, but the nonspecific nature of the
    morphological changes makes it difficult to identify lead as the
    etiological agent except by association. There is a progressive
    increase  in fibrosis, beginning in peritubular areas extending into
    the interstitium (Cramer et al., 1974). Inflammatory cells are
    uncommon and are probably not a primary component of the process.
    There is eventual tubule atrophy and hyperplasia of surviving tubules.
    There is little evidence that the glomerulus is directly affected by
    excessive exposure to lead, except for some nonspecific swelling of
    mesangial and epithelial cells. In the terminal stage, glomeruli
    become sclerotic.  An immunological basis for the progression  of 
    lead-induced nephropathy,  as suggested following gold and mercury
    exposures, might be suspected. However, there is at present no
    published documentation of antirenal antibodies or immune-complex
    formation in the pathogenesis of lead nephropathy. One study suggested
    that lowered glomerular filtration rate occurs in occupational
    exposure to lead that does not produce clinical toxicity (Wedeen et
    al., 1979). The pathophysiological basis for this observation has not
    yet been determined but may be a consequence of direct toxicity to
    epithelial cells of the glomerular apparatus.

    FIGURE 15

         Intranuclear inclusion bodies are uncommon in the late stages of
    lead nephropathy, although they may be seen in renal biopsy or autopsy
    as a manifestation of a super-imposed severe acute exposure.  It has
    been shown that inclusion bodies may be found in the urine of workers
    with occupational exposure to lead, but their presence or absence in
    urine has not been related to the severity of lead nephropathy
    (Schumann et al., 1980).

    5.5.2  Cadmium

         Cadmium is an occupational and environmental contaminant that has
    received a great deal of attention.  An important toxicological
    feature of cadmium is its exceptionally long biological half-life in
    the human organism (10-30 years). Once absorbed, cadmium is
    efficiently retained in the organism and accumulates throughout life.
    In the newborn baby, cadmium is present only at very low levels, but
    by the age of 50 the cadmium body burden may have reached up to 20-30
    mg and, in people occupationally exposed, it may reach values as high
    as 200-300 mg. Furthermore, cadmium concentrates in vital organs,
    particularly in the kidneys. At low levels of exposure, such as those
    prevailing in the general environment, 30-50% of the cadmium body
    burden is found in the kidneys alone (Nomiyama, 1980; Bernard &
    Lauwerys 1986; Friberg et al., 1986).

         The accumulation of cadmium in the kidney, may give rise to a
    progressive form of tubulo-interstitial nephritis. In contrast to the
    situation with many nephrotoxins, including other heavy metals such as
    lead and mercury, there are virtually no acute effects of inorganic
    cadmium salts on the kidney, except perhaps for some nonspecific
    effects that have been seen in animals given near-lethal doses. One of
    the most challenging questions regarding the metabolism of cadmium has
    been the role of metallothionein in cellular metabolism and its
    potential toxicity. Metallothionein synthesized within the kidney
    protects from cadmium toxicity, but intravenously injected
    cadmiummetallothionein is more nephrotoxic than inorganic cadmium (see
    section 4.5.3.).

         Cadmium nephropathy was first described by Friberg (1948, 1950)
    who studied a group of alkaline battery workers in Sweden during the
    late 1940s. Since these reports, a number of epidemiological studies
    have shown the occurrence of nephrotoxic effects in populations
    exposed to cadmium at work and in the general environment. These
    studies have demonstrated that the most prominent feature and probably
    the earliest sign of cadmium nephropathy is increased proteinuria.

         Studies performed in the 1950s and 1960s (reviewed by Friberg et
    al., 1986 and Bernard & Lauwerys, 1986) showed that cadmium
    proteinuria is similar to the tubular-type proteinuria described by
    Butler & Flynn (1958) in patients with tubular disorders, and consists
    of unidentified proteins of low relative molecular mass derived from
    plasma.  Characterization of these proteins led to the discovery of
    ß2-microglobulin,  retinol-binding protein, and alpha1-microglobulin. 
    Subsequent studies demonstrated that the increased urinary excretion
    of proteins of low relative molecular mass observed in cadmium
    nephropathy and other renal disease was due to the failure of the
    proximal tubules to reabsorb proteins filtered through the glomeruli.

         The effects of cadmium on the excretion of ß2-microglobulin have
    been extensively documented (Bernard et al., 1976, 1979a,b, 1982,
    1987). Measurement of retinol-binding protein is much more reliable in
    acidic urine and detects tubular proteinuria with equal sensitivity
    (Bernard et al., 1982).

         As cadmium nephropathy progresses, it increasingly presents the
    signs of a complete Fanconi's syndrome, i.e. aminoaciduria,
    glucosuria, increased urinary excretion of calcium, phosphorus, and
    uric acid, and decreased concentrating ability of the kidneys. In the
    most severe cases, the GRF decreases. The disturbances in calcium and
    phosphorus metabolism may lead to a demineralization of the bones and
    the formation of kidney stones (Friberg et al., 1974).

         In cadmium-polluted areas of Japan, signs of renal dysfunction
    very similar to those observed in cadmium workers have been frequently
    found. A higher incidence of proteinuria, glucosuria, and
    aminoaciduria, and increased excretion of ß2-microglobulin  have been
    observed in the Zinzu river basin in Toyama where Itai-Itai disease
    was first seen (Fukushima et al., 1974; Kjellström et al., 1977;
    Shiroishi et al., 1977; Kjellström & Nordberg, 1978). In the endemic
    area of Toyama, the increased urinary excretion of ß2-microglobulin 
    was strongly related to the residence time in that area as well as to
    the purposes for which contaminated river water was used (Kjellström
    et al., 1977). In urine, ß2-microglobulin concentration correlated
    with the cadmium level (Nogawa et al., 1979a,b).

         Further investigations of the renal function of the inhabitants
    in this area revealed a significant decrease in both creatinine
    clearance and renal phosphorus reabsorption.  Renal dysfunction due to 
    chronic cadmium poisoning was also found in other areas of Japan where
    the rice was contaminated by cadmium (Saito et al., 1977; Kojima et
    al., 1977). Studies carried out in Belgium suggested that
    environmental exposure to cadmium in an industrialized area polluted
    by this metal may exacerbate the age-related decline of renal function
    in elderly residents (Lauwerys et al., 1980; Roels et al., 1981a,b).
    Since cadmium-induced nephropathy may occur within the general
    population, it is of major public health importance to know what level
    of cadmium exposure carries a risk of renal tubular dysfunction and
    cadmium nephropathy.

         The concept of a critical concentration of cadmium has very
    important implications with regard to establishing maximum levels of
    cadmium that human populations may be exposed to with some margin of
    safety. From a comparison of the cadmium concentrations in the renal
    cortex of cadmium-exposed people with and without signs of kidney
    damage, Friberg et al. (1974) suggested that the critical level of
    cadmium in the renal cortex for the appearance of tubular proteinuria
    is around 200 mg/kg. With the development of neutron activation
    techniques allowing the  in vivo determination  of cadmium in
    tissues, the critical level of cadmium in the human  kidney has been
    more precisely assessed.

         Investigations conducted by Roels et al. (1981a) in Belgium and
    Ellis et al. (1981) in the USA have shown that when the concentration
    of cadmium in the kidney cortex reaches about 200 mg/kg, signs of
    renal dysfunction (e.g., increased urinary excretion of albumin and
    ß2-microglobulin) develop in about 10% of male workers exposed to this
    metal. On the basis of the relationship between the concentrations of
    cadmium in the urine and renal cortex and the prevalence of renal
    anomalies, the critical concentration of cadmium in urine has been
    estimated to be 10 µg/g creatinine (Bernard et al., 1979a,b; Buchet et
    al., 1980; Roels et al., 1981a,b). Epidemiological studies of people
    living in cadmium-polluted areas of Japan have shown that
    ß2-microglobulinuria occurs after a lifetime accumulation of 2000 mg
    cadmium or more (Nogawa et al., 1989).

         Since several studies have shown that, in most cases, once
    cadmium proteinuria has developed, it is irreversible, the progression
    of renal dysfunction after cessation of exposure is very slow (Roels
    et al., 1982; Elinder et al., 1985a,b). Persistent proteinuria is
    found frequently among retired cadmium workers with no evidence of
    renal insufficiency. In a group of workers removed from exposure after
    the finding of microproteinuria (low or high relative molecular mass),
    the reduction in GFR during a 5-year follow-up was about five times
    greater than that accounted for by aging (Roels et al., 1989).

    5.5.3  Mercury

         It has been known for a long time that patients treated with
    mercurial compounds can develop a glomerulonephritis that is usually
    of the immune complex type (Becker et al., 1962; Druet et al., 1982).
    Cases of mercury glomerulonephritis have also been reported as a
    result of chronic exposure to high levels of mercury in industry
    (Tubbs et al., 1982). Patients with mercurial nephropathy usually
    present with a  proteinuria and occasionally a nephrotic syndrome, but
    no renal insufficiency (Druet et al., 1982).

         Mercury may produce different effects on the kidney depending on
    the biochemical form of the metal and nature of exposure.  Inorganic
    mercury compounds are classic examples of agents that cause acute
    tubular necrosis. Mercuric chloride was used as a suicidal agent
    during the nineteenth and early part of the twentieth centuries but
    was unpopular for this purpose because of the painful accompanying
    corrosive injuries it produced.

         Regardless of the route of administration, mercuric chloride
    produces acute tubular necrosis within hours of administration,
    resulting in anuria and death. If the patient can be maintained by
    dialysis, regeneration of tubular lining cells is possible. These may
    be followed by ultrastructural changes consistent with irreversible
    cell injury, including actual disruption of  mitochondria, release of
    lysosomal enzymes, and rupture of cell membranes.

         The necrosis of the epithelium of the pars recta following 
    injection of mercuric chloride has been described in detail in the
    rat. Cellular changes include fragmentation and disruption of the
    plasma membrane and its appendages, vesiculation and disruption of the
    endoplasmic reticulum and other cytoplasmic membranes, dissociation of
    polysomes and loss of ribosomes, mitochondrial swelling with
    appearance of amorphous intramatrical deposits, and condensation of
    nuclear chromatin. These changes are common to renal cell necrosis
    resulting from a variety of causes (Gritzka & Trump, 1968).

         Mercury and its compounds are used widely, not only in various
    industrial processes but also in a number of other applications such
    as fungicides, contraceptive spermicides, and disinfectants.  Several
    studies have been carried out to determine the extent to which current
    exposure of human populations to mercury can cause adverse renal
    effects. Foa' et al., (1976) reported an increased prevalence of
    glomerular proteinuria in workers exposed to mercury vapour in a
    chloralkali plant.

         Studies carried out between 1979 and 1984 (Buchet et al., 1980;
    Roels et al., 1985) provided further evidence that occupational
    exposure to mercury vapour can lead to subclinical renal 
    disturbances. These consisted of increased urinary excretion of
    proteins of high relative molecular mass (albumin, transferrin, and
    immunoglobulin G), lysosomal enzymes, and retinol-binding protein,
    which occurred at a higher prevalence in subjects who excreted more
    than 50 µg mercury/g creatinine. These observations were not confirmed
    by Stonard et al. (1983), who found only a slight increase in the
    prevalence of NAG and gamma-glutamyltranspeptidase (an enzyme of the
    brush-border) in workers with urinary mercury levels higher than 100
    µg/g creatinine.

         Increased urinary excretion of NAG has also been found in workers
    involved in the production of various mercuric salts (Rosenman et al.,
    1986). Studies on patients with Minamata  disease have provided 
    inconsistent results regarding the induction of proximal tubular
    injury by methylmercury (Iesato et al., 1977; Ohi et al., 1982). By
    contrast, in a study of 509 infants exposed to phenylmercury fungicide
    on cloth diapers, Gotelli et al. (1985) clearly demonstrated that the
    kidney is a target organ during prolonged exposure to this compound.
    They showed that the urinary excretion of
    gamma-glutamyl-transpeptidase increased in a dose-dependent manner
    when urinary mercury exceeded approximately 220 µg/litre.  This effect
    was, however, completely reversible and had disappeared when the
    infants were re-examined two years later.

         Although exposure to a high dose of mercuric chloride is directly
    toxic to renal tubular lining cells, chronic low-dose exposure to
    mercuric salts or even elemental mercury vapour may induce an
    immunological glomerular disease.  This form of mercury injury to the
    kidney is clinically the most common form of mercury-induced
    nephropathy.  Exposed workers may develop a proteinuria that is
    reversible after they are removed from exposure.  It has been stated
    that mercury-induced nephropathy seldom occurs without sufficient
    exposure to produce detectable mercury neuropathy as well.

         Experimental studies have shown that the pathogenesis of mercury
    nephropathy has two phases: an early phase characterized by an
    anti-basement-membrane glomerulonephritis  followed by a superimposed
    immune-complex glomerulonephritis (Roman-Franco et al., 1978). The
    pathogenesis of the nephropathy in humans appears similar, although
    antigens have not been characterized. Also, the early
    glomerulonephritis may progress in humans to an interstitial
    immune-complex nephritis (Tubbs et  al., 1982).

    5.5.4  Gold

         The use of gold in the form of organic salts to treat rheumatoid
    arthritis may be complicated by development of proteinuria and the
    nephrotic syndrome (Hall et al., 1987). Morphologically, the kidney
    shows an immune-complex glomerulonephritis  with granular deposits 
    along the glomerular basement membrane and in the mesangium. The
    pathogenesis of the immune-complex disease is not known for certain,
    but gold may behave as a hapten and generate the production of
    antibodies with subsequent deposition of gold protein-antibody
    complexes in the glomerular subepithelium.  Another hypothesis is that
    antibodies are formed against damaged tubular structures, particularly
    mitochondria, providing immune complexes for the glomerular deposits
    (Viol et al., 1977).

         The pathogenesis of the tubular cell lesions induced by gold
    therapy is probably initiated by the direct toxicity of gold to
    tubular cell components. From experimental studies it appears that
    gold salts have an affinity for the mitochondria of proximal tubular
    lining cells. This is followed by autophagocytosis and accumulation of
    gold in amorphous phagolysosomes (Stuve & Galle, 1970). Gold particles
    can be identified in degenerating mitochondria, in tubular lining
    cells, and in glomerular epithelial cells by X-ray microanalysis
    (Ainsworth et al., 1981).

    5.5.5  Bismuth

         The effects of bismuth on the kidney are similar to those of
    lead, but it is a less frequent cause of renal disease. This is
    because bismuth is not present in such large amounts in the ambient
    environment, nor is it as important industrially. However, bismuth has
    been used therapeutically to treat a variety of ailments, most
    particularly syphilis. Bismuth administration results in the formation
    in proximal renal tubular lining cells of characteristic nuclear
    inclusion bodies that are similar to the lead-induced bodies and are
    composed of a bismuth-protein complex. The protein is acidic and has
    an amino acid composition similar to that forming the lead inclusion
    bodies. However, there is a slight difference in morphology between
    the lead- and bismuth-induced inclusion bodies. The bismuth-protein
    complexes are also observed in the mitochondria of proximal tubular
    lining cells (Fowler & Goyer, 1975). The bismuth content of the
    inclusion bodies has been confirmed by X-ray microanalysis of tissue
    sections. Whether bismuth produces a chronic interstitial nephropathy
    like lead has not yet been documented.  However, bismuth inclusions
    have been found at autopsy more than 30 years after a course of
    bismuth therapy.

    5.5.6  Uranium

         Exposure of humans or experimental animals to compounds of
    uranium results in injury and necrosis of proximal renal tubules. The
    most sensitive site is the pars recta (as in the case of mercury),
    but, depending on the dose, injury and necrosis may extend to other
    parts of the proximal tubule. Acute injury is followed by regeneration
    of tubular epithelial cells. Chronic effects have not been reported. 
    An increase in the urinary excretion of ß2-microglobulin and of
    specific amino acids has been reported by Thun et al. (1985) in
    uranium mill workers.

    5.5.7  Chromium

         The acute and chronic effects of chromium (mainly on the
    respiratory tract and skin) are due largely to hexavalent compounds.
    The acute tubular toxicity of chromate and dichromates salts in

    animals is well documented, and renal tubular necrosis has also been
    described in humans following acute poisoning (Langard & Norseth,
    1986). Epidemiological studies have shown that chromium(VI) can
    produce slight tubular dysfunction in chronically exposed workers.
    Mutti et al. (1979) reported an increased prevalence of elevated
    ß-glucuronidase and total protein levels in the urine of welders
    exposed to chromium. These observations have been confirmed by recent
    studies using more sensitive, reliable markers of tubular injury, such
    as ß2-microglobulin (Lindberg & Vesterberg, 1983), retinol-binding
    protein, and the BB-50 renal antigen (Mutti et al., 1985).

         Franchini & Mutti (1988) have studied dose-effect/response
    relationships between the urinary excretion of chromium and that of
    retinol-binding protein or the renal antigen BB-50. Most of the
    abnormal values were observed in subjects with urinary excretion of
    chromium greater than 15 µg/g creatinine; however, above this
    threshold the degree of tubular impairment was not related to urinary
    excretion of chromium. Franchini & Mutti (1988) explained this
    phenomenon by postulating that the tubular damage observed in
    chromium(VI)-exposed workers is transient and due mainly to acute
    exposure, and that workers become progressively resistant to the
    effects of more severe or prolonged exposure.

    5.5.8  Arsenic

         Acute arsenic poisoning may cause tubular necrosis. Acute or
    severe chronic poisoning is usually treated with the chelating agent
    BAL (2,3-dimercaptopropanol). Inhalation of arsine may also produce an
    acute tubular necrosis as a result of intravascular haemolysis.

         In a cross-sectional study, Foa' et al. (1987) failed to show
    significant differences between occupationally exposed workers and
    matched controls, with the exception of a slight increase in the
    urinary excretion of retinolbinding protein. However, owing to the
    small sample size and the low power of the study, no definite
    conclusion could be drawn from the slight increases in albuminuria,
    ß2-microglobulin,  and the brush-border antigen BB50. The authors
    concluded that extended population surveys would be desirable for a
    complete definition of such subtle effects.

    5.5.9  Germanium

         Germanium is naturally present in the diet, normal intake being
    about 1 mg per day. It is being used increasingly in the semiconductor
    industry. A recent report from Japan documented renal failure in ten
    individuals (including two deaths) among previously healthy
    individuals taking large doses of germanium (of the order of 50-250 mg
    per day) over periods of 4-18 months (Matsusaka et al., 1989). Renal
    biopsy or autopsy in seven cases showed degeneration of the renal
    tubular epithelium in all cases with or without interstitial fibroses
    or oedema. The glomeruli were only minimally effected in two cases.

    6.  RENAL CANCER

         Tumours of the renal parenchyma, pelvis, and ureters are
    uncommon, accounting for less than 2-3% of all human cancers
    (DeKernion & Berry, 1980; Dayal & Kinman, 1983). The role of drugs,
    chemicals, and other environmental factors in the etiology of
    parenchymal and urinary tract tumours is unclear, but cancer of these
    sites is most common in certain industrialized nations (Sweden) and in
    people in the higher socio-economic groups (Rimpela & Pukkala, 1987).
    Other risk factors have been stratified (Selli et al., 1983). The
    ratio between tumours of the renal parenchyma and pelvis is fairly
    constant (about 5:1), and the parallel trends in increasing incidence
    argue for some commonality in the etiology of tumours at both sites,
    although some factors may be site specific. Tumours are nearly twice
    as common in males as in females. A compilation of trends in cancer
    rates in the USA indicates that the incidence of both kidney and
    bladder cancer is increasing (Pollack & Horm, 1980). A similar
    observation has been made with regard to renal parenchymal tumours in
    males in Scotland (Ritchie et al., 1984).

    6.1  Renal tumour classification

         The International Classification of Diseases for Oncology (code
    189) divides tumours of the urinary system into five groups according
    to their size, i.e. parenchyma of the kidney (189.0), renal pelvis
    (189.1), ureter (189.2), urethra (189.3), and paraurethral gland
    (189.4) (Mostofi et al., 1981; WHO, 1990). These distinctions have
    only been made in recent years, so that many mortality studies of
    renal tumours have included this whole category. Of the five types of
    urinary tract tumours, about 90-95% of renal tumours in adults are
    adenocarcinomas arising from the renal parenchyma. Nephroblastoma
    (Wilm's tumour) is the second most common histological type of renal
    tumour and accounts for 2-4% of kidney cancer in Sweden and the USA.
    It is easily distinguished morphologically from renal adenocarcinoma
    and usually appears in the first five years of life; 95% of cases
    occur before the age of 15 years. Although nephroblastomas are the
    fourth most common tumour in childhood, they are relatively rare in
    adults.

    6.2  Renal adenocarcinoma

         Renal adenocarcinoma has been known under several synonyms (clear
    cell carcinoma, hypernephroma, Grawitz tumour) reflecting uncertainty
    about its origin, but immunological studies have established that
    renal adenocarcinomas arise from the proximal convoluted tubule
    (Wallace & Nairn, 1972). They tend to be circumscribed, ranging in
    size from microscopic lesions to large neoplasms (Hamilton, 1975). The
    spectrum from small benign lesions to clearly malignant lesions
    suggests a continuous pathological process, so that it is often
    difficult to label smaller tumours as benign or malignant. Hellsten et
    al. (l983) have defined all tumours of 2 cm or more in diameter as
    adenocarcinomas.  Postmortem studies have shown that adenomas are

    present in approximately 25% of all males over 50 years of age, and it
    was found that 34% of 235 clinically unrecognized tumours present at
    autopsy were less than 3 cm in diameter. In the absence of invasion of
    surrounding tissue, features such as frequent mitotic figures,
    cellular pleomorphism, and haemorrhage and necrosis generally indicate
    a malignant potential regardless of size.  Calcification may be
    detected by X-ray examination in about 15% of cases.  Although 2-3%
    may be cystic, the commonest form is a solid tumour that is usually
    composed of clear cells rich in lipid, glycogen, or both, but may
    contain granular cells or even tightly packed eosinophilic cells
    referred to as oncocytes. The cell pattern may be trabecular, solid,
    or mixed, but it is doubtful that cell type or structural pattern has
    any clinical significance. Grading is difficult and has not been shown
    to be clinically useful. These tumours usually grow slowly, and
    overall survival is 20-25%  after nephrectomy. The presence  of
    multiple tumours, renal vein invasion, or regional lymph node
    metastases indicates a poorer prognosis. Hellsten et al. (1983)
    recorded metastasizing renal carcinoma as cause of death in 21% of a
    postmortem series, and in 33% a second malignant tumour was observed
    causing the death of 20%.

         Immunological mechanisms are thought to determine the natural
    history of the disease.  The development of monoclonal antibodies and
    flow cytometry have provided new methods for investigating
    immunological responses. Total T lymphocyte counts were found in a
    study of 32 patients to be lower than in controls (Ritchie et al.,
    1984), due largely to a deficit of T helper cells but not T
    suppressor-cytotoxic cells. This effect of the tumour is reversed by
    removal of the primary tumour, and recurs with return of the tumour.
    These findings are believed to suggest that there is a systemic effect
    of the tumour acting at the level of the bone marrow or thymus to
    affect the production or maturation of T helper cells.

         The role of specific environmental factors in the etiology of
    urinary tract tumours has been difficult to define (Newson & Vugrin,
    1987). Apart from increases in renal tumours in asbestos workers and
    the identification of some occupationally related bladder tumours,
    there does not appear to be a clearly defined association with
    specific chemicals or environmental factors.  However, there is an
    association with a combination of exposure to substances in the
    environment and life-style practices such as tobacco use. This
    suggests that there may be interactions between substances or that the
    urinary tract, like the lung, has to deal with a number of substances
    with promoter activity. Cigarette smokers have a 2-fold increase in
    risk of urinary tract tumours (Goodman et al., 1986), and an increase
    associated with alcohol and coffee usage has been suggested (Jacobsen
    et al., 1986). In addition, chronic interstitial nephritis may
    predispose to urinary tract tumours. People with endemic (Balkan)
    nephropathy have an increase in renal tumours and a possible
    relationship between chronic interstitial nephritis and renal
    neoplasia (see section 5.3). It is also noteworthy that human

    populations with excessive exposure to some known carcinogens (e.g.,
    cycasin in Guam and aflatoxins in Africa and Asia) have not yet been
    shown to have an increase in kidney cancer (Sufrin & Beckley, 1980).
    Renal adenocarcinoma has been diagnosed with increasing frequency in
    patients with chronic renal failure, particularly in those patients
    treated with long-term dialysis (Dunhill et al., 1977).

         An association between renal cancer and excess exposure to lead
    has not been clearly established, but a study of lead smelter and
    battery workers found a significant excess of malignancies at all
    sites, these being mostly lung tumours (Cooper & Gaffey, 1975). Case
    reports of renal tumours in workers with lead nephropathy have
    appeared (Baker et al., 1980; Lilis, 1981).

    6.3  Upper urothelial carcinoma (transitional cell carcinoma)

         Tumours of the renal pelvis form a spectrum from benign
    papillomas to frank papillary carcinomas and, like bladder tumours,
    are generally low-grade cancers. However, they tend to recur
    regardless of their morphology. Upper urothelial carcinoma has been
    associated with RPN and analgesic abuse, but a cause-and-effect 
    relationship between the two has not been proven (Bach & Bridges,
    1985a). The incidence of upper urothelial carcinoma among analgesic
    abusers is very high, and females predominate in analgesic-associated 
    upper urothelial carcinoma.  The female:male ratio is 2.5 to 1
    (Bengtsson et al., 1978), which is in keeping with the ratio for
    analgesic abusers. Analgesic abusers also develop upper urothelial
    carcinoma at a younger age than non-analgesic abusers (Mihatsch et
    al., 1980a,b,c). The distribution of urothelial carcinomas in
    analgesic abusers has a distinct pattern; tumours of the renal pelvis,
    ureter, and bladder are found 80 times, 90 times, and 7 times,
    respectively, more frequently than in non-analgesic abusers. The
    tumours are typically multiple, diffuse, and poorly differentiated,
    and spread rapidly (Mihatsch et al., 1980c).

         Patients who discontinue abuse of the drugs are at a greater risk
    of developing upper urothelial carcinoma, often  after a latent 
    period of 10-20 years  after initiating analgesic abuse.  Greatly
    improved dialysis techniques may result in the survival of
    analgesic-abusing patients who would otherwise have developed
    end-stage renal disease and subsequently died (Mihatsch et al.,
    1980a). It has therefore been suggested that the incidence of upper
    urothelial carcinoma will increase.

         The diagnosis of upper urothelial carcinoma is difficult in the
    clinical situation because of few specific clinical  symptoms to
    indicate the malignant changes (Johansson et al., 1976; Bengtsson et
    al., 1978; Mihatsch et al., 1980a,b,c; Mihatsch & Knusli, 1982; Bach
    & Bridges, 1985a; Pommer et al., 1986). The prognosis is poor, and
    patients with upper urothelial carcinoma only have a mean survival
    time of 22 months (Mihatsch et al., 1980a) owing to the difficulty of

    diagnosis, compromised renal function of patients with RPN, and
    multifocal sites of rapidly developing and widespread invasion and
    metastases (Johansson et al., 1976; Mihatsch & Knusli, 1982).

    6.4  Experimentally induced renal adenomas and adenocarcinomas

         Renal adenomas and adenocarcinomas may be induced in laboratory
    animals by various natural products and biological and chemical
    agents. However, linkage of exposure of these substances to renal
    cancer in humans is lacking in most instances, or, at best, is only
    suspected.

    6.4.1  Background incidence of spontaneous tumours in experimental animals

         The incidence of spontaneous renal parenchymal tumours in most
    commonly used strains of male rats and mice is in the region of 0.2%,
    whereas it is < 0.1% for females (Crain, 1958; Goodman et al., 1979,
    1980; Ward et al., 1979; Maekawa et al., 1983). The incidence may be
    up to 2.7% (Pour et al., 1979) in hamsters, which is generally low
    enough for investigative studies.

         Renal tumours have been induced experimentally by a large number
    of compounds including lead salts (Kilham et al., 1962; van Esch &
    Kroes, 1969; Goyer & Moore, 1974), nickel sulfides (Jasmin & Riopelle,
    1976; Sunderman et al., 1984), methylmercury chloride (Mitsumori et
    al., 1981),  N-(4 -fluoro-4-biphenylyl)  acetamide (Hinton et al.,
    1980), trisodium nitrilotriacetic acid (Goyer et al., 1981), potassium
    bromate (Kurokawa et al., 1983), halogenated alkenes (Kociba et al.,
    1977; Reichert et al., 1984), and tris(2,3-dibromopropyl)phosphate
    (Reznik et al., 1979). Natural products include cycasin (Laqueur &
    Spatz, 1968), aflatoxins B1 (Butler et al., 1969; Epstein et al.,
    1969), ochratoxin A (Kanisawa & Suzuki, 1978), citrinin (Arai &
    Hibino, 1983), the fermentation-derived anti-neoplastic agent
    daunomycin (Sternberg et al., 1972), and streptozotocin (Rakieten et
    al., 1968; Hard, 1985). Diethylstilbestrol (Horning & Whittick, 1954)
    and related estrogens (Li et al., 1983) produce a high incidence of
    parenchymal tumours in hamsters.

         The classical two-stage model of carcinogenesis also applies to
    several of the nitrosamines, where promoters include sodium arsenite
    (Shirachi et al., 1983), DL-serine (Hiasa et al., 1984a), folic acid
    (Shirai et al., 1984), lead acetate (Hiasa et al., 1983), nicotinamide
    (Rosenberg et al., 1985), trisodium nitrilotriacetate (Hiasa et al.,
    1984b), and citrinin (Shinohara et al., 1976). There are several
    factors that may affect the development of carcinomas, including diet
    (Hard & Butler, 1970; McLean & Magee, 1970; Hard, 1980, 1984; Swann et
    al., 1980), partial hepatectomy (Evarts et al., 1982), unilateral
    nephrectomy (Ito et al., 1969), and unilateral hydronephrosis (Ohmori
    & Tabei, 1983).

    6.4.2  Inorganic compounds

         Various inorganic compounds of lead have been investigated over
    the last 30 years (Van Esch & Kroes, 1969). The tumours arise from
    kidney tubular epithelial cells in kidneys and are similar to renal
    cortical tumours found in humans. Production of tumours requires
    continuous exposure to relatively high concentrations of lead in the
    diet or drinking-water for 1 to 2 years. The tumours occur in a
    background of severe interstitial nephritis characterized by tubular
    atrophy as well as focal areas of hyperplasia. They are usually
    multifocal and vary from microscopic adenomas to large renal
    adenocarcinomas that may invade contiguous structures or metastasize
    to the lungs. Intra-nuclear inclusions, which are usually present in
    proximal tubular epithelial cells in lead toxicity, are absent in
    neoplastic cells, and the tumors contain much less lead than adjacent
    renal parenchyma (Mao & Molnar, 1967). Tumor cells are pleomorphic,
    and ultrastructural studies have shown marked morphological
    alterations in mitochondria.

         Renal cancer occurs after injection of crystalline nickel
    subsulfide (Ni3S2) into the kidney of rats, but not after
    treatment with amorphous nickel sulfide (NiS). No evidence indicates
    that nickel compounds are carcinogenic in experimental animals when
    administered by oral or subcutaneous routes (Sunderman, 1981).

    6.4.3  Organic molecules

         Nitrilotriacetic acid, a polyamino polycarboxylic acid with
    chelating properties similar to EDTA (used to treat lead poisoning),
    produces chronic interstitial nephropathy in rodents. A spectrum of
    tubular cell histological changes occurs from hyperplasia to small
    adenomas to adenocarcinomas (Goyer et al., 1981). Renal adenocarcinoma
    has been induced in male rats by the chronic inhalation of unleaded
    gasoline vapour (MacFarland et al., 1984), but this relationship has
    not been supported by epidemiological  studies on workers in the
    petroleum  industry (Enterline & Viren, 1985).

    6.4.3.1  Nitrosamines and related compounds

         The nitrosamines represent one of the most widely investigated
    groups of model compounds.  They include dimethylnitrosamine (Murphy
    et al., 1966; Mohr et al., 1974; Hard, 1984), which produces
    mesenchymal (connective tissue) tumours in young animals but adenomas
    and adenocarcinomas in mature animals (Hard, 1979). The
    co-administration of putrescine (Ohmori & Tabei, 1983) or
     N-3,5-dichlorophenyl-succinimide (Ito et al., 1974) with
    dimethylnitrosamine caused a dose-related incidence of up to 100%
    renal tumours after 100 weeks.  N-ethyl- N-hydroxyethylnitrosamine
    (Hiasa et al., 1979) on its own, or, especially, when administered
    with basic lead acetate (Hiasa et al., 1983) or serine (Hiasa et al.,
    1984a) causes tumours in up to 95% of animals by 32-38 weeks.

     N-Nitrosomorpholine causes oncocytomas (Bannasch et al., 1978a,b,
    1980). There are interesting differences between several of the model
    compounds, interspecies  responses, effects  of dose regimens, etc.
    However, the investigation of these models has generally permitted the
    progression of cellular injury to be described in terms of the acute
    effects, early hyperplasia, dysplasia, and different types of renal
    tumours.

    6.4.3.2  Morphological changes

         Many of these model compounds have been used to study the early,
    intermediate, and late changes at the light microscope and
    ultrastructural levels (Horning & Whittick, 1954; Butler, 1964; Butler
    & Lijinsky, 1970; Ertürk et al., 1970; Hard & Butler, 1971; Sternberg
    et al., 1972; Bennington, 1973; Hard, 1975, 1984, 1985; Bannasch et
    al., 1978a,b, 1980; Dees et al., 1980a,b; Ohmori et al., 1982; Tsuda
    et al., 1983; Eble & Hull, 1984; Hard et al., 1984; Hiasa et al.,
    1984a,b,c). The phenotypic changes associated with loss of normal
    growth control have concentrated on the focal preneoplastic changes in
    heterogeneous cells. These undergo slow changes (where it is highly
    desirable to define the origins of the neoplasm) in a limited number
    of enzyme markers and in lipids, and carbohydrates. In addition, an
    increase in cytoplasmic RNA (shown  by enhanced cytoplasmic basophilia
    or numerous ribosomes at the ultrastructural level) has been observed
    as a marker for hyperbasophilic and basophilic preneoplastic foci in
    the epithelium of the renal tubular system (Hard, 1986; Bannasch &
    Zerban, 1986).

         The nomenclature of renal parenchymal tumours is based on several
    criteria, including the size of the tumour and the
    morphological-histochemical characteristics of cells and their
    organization. The progression in tissue mass from hyperplasia, through
    dysplasia, to adenoma, adenocarcinoma, and carcinoma is a continuum,
    although  the earliest changes, particularly hyperplasia, may be
    reversible. Cells may have no cytoplasmic staining (clear cells) or
    granular acidophilic or basophilic cytoplasm staining, and tumours may
    have a mixture of cells with different staining characteristics. Where
    adenomas contain a uniform population of finely granulated
    eosinophilic cells, they are termed renal oncocytomas. Tumours can
    also be classified as tubular, solid, lobular, disorganized, invasive,
    papillary or cystadenoma, or by a composite of such terms based on
    their appearance (Hard, 1987).

         Clear and acidophilic (granular) cell kidney tumours induced by
    limited exposure of rats to  N-nitrosomorpholine are associated with
    a transient storage of glycogen (Bannasch et al., 1978a) and closely
    parallel the most common malignant renal neoplasm in man. The tumours
    originate from segments of the collecting duct system storing large
    amounts of glycogen (Nogueira et al., 1989). However, when
    microadenomas develop, the clear (glycogenotic) cells loose glycogen
    and acquire an acidophilic (granular) cytoplasm, although both cell
    types can coexist in large tumours. Lipid-storing cells are often

    found in the clear cell tumours, but the significance is not known. By
    contrast, there are no such relationships between cells storing
    glycogen and the so-called renal oncocytomas also found in NNM-induced
    rats (Bannasch et al., 1978b; Nogueira et al., 1989), although rat
    renal oncocytomas also originate from the collecting duct system
    (Nogueira & Bannasch, 1988).  They are, however, benign end-stage
    lesions where the cytoplasm is crowded with pathologically altered
    mitochondria (Krech et al., 1981). A temporary focal storage of
    glycosamino-glycans has been reported in chromophobic rat renal cell
    tubules and tumours (Bannasch et al., 1980, 1981) and in the
    corresponding type of human tumour (Thoenes et al., 1985).

    6.4.3.3  Biochemical changes in cells

         An immunohistochemical increase in glucose-6-phosphate
    dehydrogenase is associated with basophilic renal cell tumours (Tsuda
    et al., 1986) and nephroblastomas (Moore et al., 1986). The pentose
    phosphate shunt provides sugars for RNA and DNA synthesis, and the
    activation of this pathway is probably closely related to certain
    phenotypic changes, such as an increase in ribosomes and an enhanced
    cell proliferation in preneoplastic and neoplastic lesions. In line
    with this interpretation, rat renal oncocytic tubules and tumours,
    which are poor in ribosomes and grow very slowly, usually show a
    normal or even decreased activity of glucose-6-phosphate dehydrogenase
    (Tsuda et al., 1986). However, in some experimental models a reduced
    amount or activity of glucose-6-phosphate dehydrogenase was found in
    the more malignant populations, suggesting involvement of the enzyme
    in other metabolic aberrations relevant to tumorigenesis (Moore et
    al., 1986).

         Alterations in drug-metabolizing enzymes (see below) during
    carcinogenesis have been detected by immuno-histochemical methods in
    various tissues, especially in the renal tubular system, but they have
    not been correlated to the same extent with the respective enzyme
    activities and with other changes in the cellular phenotype as have
    those of carbohydrate metabolism. By contrast,
     N-ethyl- N-hydroxyethylnitrosamine-induced renal carcinomas show
    opposite alterations in drug-metabolizing enzymes in preneoplastic and
    neoplastic lesions of these tissues (Tsuda et al., 1987).  Reduced
    activities of gamma-glutamyl-transpeptidase (Ohmori et al., 1982;
    Tsuda et al., 1986), succinate dehydrogenase (Tsuda et al., 1986), and
    alkaline phosphatase (Tsuda et al., 1986) are seen as early changes
    during the development of basophilic cell tumours from hyperbasophilic
    segments of the proximal nephron. In contrast, however, there are no
    similar changes in gamma-glutamyltranspeptidase or alkaline
    phosphatase activity (but there is an increase in succinate
    dehydrogenase activity) in oncocytic tubular lesions seen in these
    animals (Tsuda et al., 1986). The increased binding of anti-cytochrome
    c oxidase to the oncocytic lesions in both man (Ortmann et al., 1988)
    and rat (Mayer et al., 1989) may be a useful marker for preneoplastic
    renal changes.

    6.4.3.4  The mechanistic basis of renal carcinoma

         The mechanistic basis for the development of renal carcinoma may
    be genotoxic or non-genotoxic. The common feature to all genotoxic
    agents is the generation of a reactive electrophilic
    (electron-deficient) species which is capable of binding to
    nucleophilic (electron-rich) sites on cellular macromolecules
    including  proteins, lipids, RNA, and especially DNA (Miller & Miller,
    1981). For example nitroso-compounds alkylate DNA in the N-7 and,
    especially (due to its prolonged stability), O-6 positions (Nicoll et
    al., 1975).

         Genotoxic compounds are either direct  alkylating agents
    (requiring no activation) or they require one or more
    biotransformation steps by the P-450-dependent mono-oxygenases (e.g.,
    chloroform) (Bailie et al., 1984; Smith & Hook, 1984). However,
    hexachloro-1,3-butadiene is transformed by ß-lyase (Elfarra & Anders,
    1984) or prostaglandin hydroperoxidase-mediated co-oxidation (Davis et
    al., 1981), which may also be involved in
     N-[4-(5-nitro-2-furyl)-2-thiazolyl]formamide transformation (Zenser
    & Davis, 1984). In addition, there may be several other renal and
    extra-renal metabolic steps. There are also some molecules that cannot
    be reliably classified into either group.  Diethystilbestol is a very
    weak alkylating agent (Lutz et al., 1982), and its mechanism of action
    is thought to be mediated by renal estrogen receptors (Li & Li, 1984). 
    However, some metabolic component may be involved, e.g., hepatic
    mixed-function oxidase (Metzler, 1981) and peroxidative activation
    (Metzler & McLachlan, 1978).

    6.5  Experimentally induced upper urothelial carcinomas
         (transitional cell carcinomas)

         There is experimental evidence to connect analgesic exposure  to 
    the development of urothelial tumours (Johansson & Angervall, 1976;
    Bengtsson et al., 1978; Bach & Bridges, 1985a). While bladder tumours
    have been studied extensively in animal models, the practical
    difficulties of looking for malignancies in the ureter or pelvis has
    limited studies in this area. Based on long-term carcinogenicity
    studies, there are few data to establish a clear experimental
    relationship between analgesic exposure and upper urothelial
    carcinoma. There is, however, experimental evidence to suggest that
    upper urothelial carcinomas can be induced using a classical two-stage
    initiation/promotion regimen (Bach & Gregg, 1988; Gregg et al., 1989).
    These data suggest that localized injury associated with papillary
    necrosis adjacent to urothelium that has already been initiated will
    result in a proliferation of changes that lead to malignancy. At
    present the full significance of these findings in terms of the human
    analgesic problem is not clear.

    7.  ASSESSMENT OF NEPHROTOXICITY

         No single  in vivo or  in vitro method of studying the
    nephrotoxicity of chemicals can address all of the questions that must
    be asked. It is therefore inadvisable to separate mechanistic research
    into target cell toxicity from the screening of novel compounds for
    their potential nephrotoxicity. The holistic approach to
    nephrotoxicity assessment also demands that  in vivo investigations
    are not separated from in vitro studies, and that data continue to be
    derived from several different animal species and related to
    accurately conducted epidemiological and clinical studies (where these
    data are available).

         Present data suggest that most  in vitro methods can provide
    information on the mechanism of primary insult and the effect on cell
    viability. However, there appears at present to be little place for
     in vitro techniques in the assessment of secondary renal changes, as
    the factors that contribute to the cascade of degenerative changes
    that follows renal insult are largely obscure. Thus there is inherent
    uncertainty as to what should be studied in  in vitro systems. There
    are, however, several approaches that can be used to provide a better
    understanding of the contribution to degenerative changes  in vivo. 
    These include harvesting tissue at different time points following an
    insult and using this tissue to study function in different cell
    types, or studying the effects of chemicals on target and non-target
    cells using pure and mixed cell cultures in the presence and absence
    of cells that are related anatomically.  It is also possible to
    exchange culture media between cells that have been insulted, and
    those that have not, in order to assess the release of factors that
    may be toxic to other cells. Different cells can be studied in the
    same media to define cell-cell interactions and how chemical insult
    affects this process.

    7.1  In vitro studies

         Despite the complexity of the kidney and nephrotoxicity, and the
    difficulty in defining what any one  in vitro system achieves, there
    are several ways to progress in the use of screening methods.

         One such rational approach could include:

    *    the careful identification of compounds with well-documented  in
          vivo nephrotoxicity, including the sequence of pathological and
         functional changes, the metabolites formed, quantities excreted,
         and the cellular pharmacodynamic effects;

    *    the choice of chemicals that target specifically for one
         anatomically discrete cell type  in vivo;

    *    the use of both more and less nephrotoxic analogues of the 
         chemical for determining structure-activity relationships and
         computer-based simulations;

    *    the systematic study of these compounds by several different  in
          vitro methods, and the use of several criteria for assessing in
         vitro nephrotoxicity for each.

    7.1.1  Choice of chemical concentrations for in vitro studies

         The validity of using the cytotoxicity  (in vitro) for a given
    concentration of a chemical as a likely indicator of a toxicological
    effect in vivo can be very difficult to establish.  It may be
    impossible to assess this in the kidney, because this organ is highly
    compartmentalized. In the intact and functioning organ, it is
    currently not possible to establish the concentrations of a xenobiotic
    (or its metabolites) associated with any one cell type. This
    uncertainty relates, for instance, to the different transport systems
    that are distributed in discrete parts of the nephron, transcellular
    pH gradients, and the selective accumulation of certain chemicals in
    cell organelles (which have heterogeneous distribution in different
    types of cells). Thus some chemicals can be selectively concentrated
    in a discrete area of the kidney to several times the plasma
    concentration. Alternatively, certain chemicals may be actively or
    selectively excluded from some cell types (Mudge, 1985).

         Drug metabolism systems that alter the physicochemical
    characteristics of xenobiotics and their metabolites (that will
    facilitate the redistribution of chemicals within and between cells)
    are also heterogeneously distributed. Thus, certain xenobiotic
    products may be selectively concentrated in (while others are excluded
    from) specific cell types.  Therefore, even though arterial and venous
    blood and urinary concentrations of chemicals can be measured, there
    is no certainty that such data relate to the concentrations of any
    specific metabolite that reaches a target cell.  Autoradiography may
    be very valuable in providing some idea on how chemicals are
    distributed, but it only shows the distribution of radiolabel-derived
    material, not its chemical nature. Once the anatomical integrity of
    the kidney has been altered, it may be difficult to relate the
    concentration of any chemical to the same cells  in vivo. While
    structure is maintained in the perfused kidney, the functional changes
    impose a similar constraint. The uncertainty as to what concentrations
    of chemicals to use in in vitro systems is exacerbated by the
    undefined influence of extrarenal and renal metabolism on the delivery
    of the proximate and ultimate toxins to the target sites of injury  in
     vivo.

    7.1.1.1  Proximate and ultimate nephrotoxicants in vitro

         In addition to the need to consider carefully the consequences of
    the changes in the route of chemical delivery when the anatomical
    integrity of the kidney has been disrupted (i.e. in all systems other
    than the isolated perfused kidney), it is important to consider how a
    chemical is delivered to the cells. This consideration should be
    expressed in terms of the chemical being free or bound (should the

    media contain protein or not?), the physicochemical characteristics of
    the solution in which the chemical is delivered (such as its pH, ionic
    concentration, and endogenous and exogenous micro- or
    macro-molecules), and the kinetics of delivery (this is generally
    zero-order in most  in vitro systems but follows first or second
    order kinetics  in vivo).

         Chemically induced nephrotoxicity may be the result of a direct
    action of the parent chemical in the kidney or may be due to an
    extrarenally formed metabolite. However, in some instances the
    chemical/metabolite has to be further metabolized in situ to form the
    ultimate nephro-toxic species. Assessing the role of  in situ
    metabolism in nephrotoxicity from  in vivo data may be difficult,
    since most nephrotoxic chemicals are also extensively metabolized in
    extrarenal tissue such as the liver. Experimental approaches using
    different species and strains of animals, inducers and inhibitors of
    drug metabolizing enzymes, and candidate proximate and ultimate
    nephrotoxic metabolites may give some evidence for the involvement of
    intrarenal metabolic activation (Rush et al., 1984). More direct
    evidence for a direct renal activation of a chemical has to come from
    in vitro studies or studies with perfused kidneys where extrarenal
    metabolism/activation can be excluded.  In vitro studies may include
    experiments with isolated tissue preparation (kidney slices and
    tubules), cells (primary cells or cell lines) or subcellular fractions
    to assess chemically induced toxicity. In order to determine the role
    of extrarenal metabolism in the formation of nephrotoxic metabolites,
    co-culture systems using liver cells (as an activation system) and
    kidney cells (as target cells) may be very useful (Moldeus et al.,
    1978). The stability of a reactive metabolite, generated by liver
    cells, may be measured by transferring, after various time intervals,
    the incubation medium to the target cell population.

    7.1.2  In vitro investigations of nephrotoxicity

         In vitro techniques can be divided into those where the
    anatomical relationship between cells is maintained (perfusion, 
    micropuncture, and slices),  those where glomeruli and tubular
    fragments are isolated, and those where cells are isolated.  The
    different techniques for assessing nephrotoxicity  in vitro have been
    reviewed and the strengths and weaknesses of each presented in broad
    terms (Bach et al., 1985, 1986; Bach & Kwizera, 1988). Some methods
    are technically difficult, depend on sophisticated equipment, are
    subject to artefacts in inexperienced hands (perfusion and
    microperfusion), and are difficult to interpret.

    7.1.2.1  Perfusion and micropuncture

         Micropuncture methodology has not been widely used to assess the
    toxicological effects of chemicals (Bank et al., 1967; Biber et al.,
    1968) because the methodology is extremely complex. However, a few of
    the problem areas should be mentioned.  Micro-puncture procedures
    succeed only when the experimentalists can collect measured small

    samples of tubular fluid and subject these to appropriate chemical
    analyses. These procedures are extremely complex, and, because of the
    small volumes involved, subject to considerable error.  To assure that
    the micropuncture collections reflect the "physiological state", most
    workers attempt to collect only very small volumes at the
    "physiological" flow rate past the point of micropuncture.  Stationary
    microperfusion has been used to assess tubular function in a
    restricted area of the nephron.

         Micro-injection into a tubular segment has been used by many
    physiologists and may be of particular toxicological interest
    (Gottschalk & Lassiter, 1973; Roch-Ramel & Peters, 1979; Diezi &
    Roch-Ramel, 1987). With this technique, various renal function
    markers, as well as potential nephrotoxicants, may be injected into a
    segment of the proximal tubule during a free-flow situation. The urine
    from the injected kidney may then be collected to assess
    nephrotoxicant effects on the injected renal function markers. For
    micropuncture specialists, this is a relatively straight-forward
    technique and does not involve the problems associated with removal of
    tubular fluid by micropuncture. This procedure might permit the
    assessment of nephrotoxicant effects on membrane permeability by
    examining, for example, inulin excretion from the injected kidney.

         The isolated perfused tubule technique represents the development
    of an  in vitro procedure that permits the assessment of intact
    tubular function under carefully controlled  in vitro conditions.
    Hence all of the advantages of  in vitro methodology are available
    while intact tubular function is being studied (Diezi & Roch-Ramel,
    1987).

         Tubule segments can also be isolated by manual dissection for
    microperfusion, where they are attached to micropipettes suspended in
    a bathing solution and perfused with an artificial tubular fluid
    (Ullrich & Greger, 1985). Relatively few attempts have been made to
    apply this sophisticated methodology to the study of nephrotoxicants,
    where it would be possible to add chemicals to either the perfusate or
    the bathing solution and examine effects on either the tubular or
    basolateral side of the cell. This technique has been used to
    demonstrate that organic anion transport across the tubular cell is
    active on the basolateral side but not on the luminal side (Tune et
    al., 1969).

         The major advantage of this procedure is that it permits an  in
     vitro assessment of renal function with a tissue segment that is
    essentially intact. The situation in an isolated nephron segment is
    obviously not identical to that in the intact kidney, but by carefully
    regulating the perfusion solution and the bathing solution one can
    approximate  in vivo physiology.  This is a potentially important
    procedure that needs to be assessed for its utility.

    7.1.2.2  Renal cortical slice

         These techniques have been reviewed extensively (Berndt,
    1976,1987; Bach & Lock, 1982; Kacew, 1987) and used to show
    deleterious effects of chemicals and drugs on the kidney. Much of the
    published information has focused on renal tubular transport as the
    criterion for establishing the nephrotoxic potential of chemicals. The
    tests are based on measuring the accumulation of the organic ions
     p-aminohippurate (PAH) and tetraethyl-ammonium (TEA) (Hirsch, 1976)
    by renal slices. The organic anion transport is a sensitive indicator
    of aminoglycoside (Kluwe & Hook, 1978; Kaloyanides & Pastoriza-Munoz,
    1980) and cephaloridine toxicity (Kuo & Hook, 1982; Kuo et al., 1982). 
    Similarly, the effects of mercuric ions, chromate ions,
    hexachlorobutadiene conjugates, and other nephrotoxins appear readily
    detectable by this approach. Organic ion accumulation and
    gluconeogenesis in renal cortical slices may be poor indicators of
    early toxic effects to the kidney resulting from cisplatin
    administration because these parameters are not affected except at
    high doses. The poor sensitivity and delayed response of renal slice
    parameters indicate that membrane function and cell metabolism may not
    be early targets of cisplatin at the cellular level. Slices maintain
    ß-lyase activity for up 12 h and have been used to study halo-alkene
    toxicity. Aminooxyacetic acid inhibits ß-lyase activity almost
    completely.  DCVC decreases PAH accumulation, but does not appear to
    use the same transport process.

    7.1.2.3  Isolated nephron segments

         Isolated tubules overcome many of the disadvantages of cortical
    slices, in that they remain in contact with substrates and toxins in
    the medium, whereas cortical slices may show lumen collapse within a
    short period (Chahwala & Harpur, 1986).

         Freshly isolated tubule or cell suspensions offer an important
    way of studying the mechanisms of nephrotoxicity and screening novel
    compounds for their potential acute effects on the kidney. A
    limitation, however, is the short  in vitro lifespan of isolated
    tubules prepared, by any technique, for studying early toxic effects.
    In general, most investigators limit incubations to no more than 2-4
    h because of loss of viability and functional capabilities. Ormstad
    (1982) reported rapid loss of viability  of isolated renal tubules and
    cells, more than 25% of the cellular lactate dehydrogenase (LDH)
    leaking to the medium during 1.5 to 2 h of incubation. Obatomi &
    Plummer (1986) observed a 40% loss in tubule cell viability during 3-h
    incubations of rat proximal tubules. Loss of renal function, such as
    O2 consumption, has also been reported for isolated tubules (Harris et
    al., 1981).

         A number of fresh tubular systems exhibiting high initial
    viability (> 90% by trypan blue exclusion) have been prepared from
    collagenase digests of rat cortical tissue (Cunnaro & Weiner, 1978;
    Belleman, 1980; Cojocel et al., 1983; Gstraunthaler et al., 1985;

    Obatomi & Plummer, 1986) using tubular fragments of proximal origin.
    The tubules can be used to study the metabolism of xenobiotics liable
    to be converted into compounds responsible for the alteration of
    normal renal metabolism (Jones et al., 1979). Tubular fragments
    obtained by collagenase treatment of dog (Baverel et al., 1978,
    1980a), human (Baverel et al., 1979), baboon (Michoudet & Baverel,
    1987), and guinea-pig (Baverel et al., 1980b) renal cortex have also
    been used for metabolic studies.  These fragments retain the
    gluconeogenic capacity that is specific to the proximal convoluted
    tubule (Guder & Ross, 1984). Human, dog, and rat renal cortex tubules
    also release ammonia from glutamine, and any drug-induced disturbance
    of renal ammoniagenesis can be studied with these models (Martin et
    al., 1987, 1989, 1990).

         Proximal tubules have been purified from the above preparations
    by centrifugation on a Percoll density gradient.  The samples obtained
    exhibit enrichment in proximal tubule cells relative to cells from
    other areas of the nephron, as indicated by the distribution of
    alkaline phosphatase and hexokinase (Vinay et al., 1981).

         Suspensions of thick ascending limb fragments have also been
    prepared from dog (Baverel et al., 1980a,b; Anand-Srivastava et al.,
    1986), rat (Trinh-Trang-Tan et al., 1986), and rabbit (Chamberlin et
    al., 1984) outer medulla.  Suspensions of collecting tubules obtained
    from the  inner  medulla (Anand-Srivastava  et al., 1986; Wirthensohn
    et al., 1987, 1989) may prove very useful in studies of the effect of
    nephrotoxic substances that interfere with the function of this renal
    zone.

         Direct addition of different concentrations of nephrotoxic agents
    such as ochratoxin A, citrinin, furosemide, and potassium chromate to
    the suspension releases enzymes specific to the proximal tubule (Table
    4), such as alanine aminopeptidase, leucine aminopeptidase and
    alkaline phosphatase, to the incubation medium in a dosedependent
    manner (Endou et al., 1985). To characterize further the intrarenal
    site(s) and mode of nephrotoxicity, definite portions of a single
    nephron can be microdissected from the collagenase-treated kidney.
    There are two different methods available for the microdissection of
    individual nephron segments; one is from lyophilized kidney sections
    and the other is from collagenase-treated fresh kidneys (Morel et al.,
    1976). In nephrotoxicity studies, fresh individual nephron segments
    are used. The following segments can be isolated: the glomerulus, the
    proximal tubule (S1, S2, S3), the thin descending limb of
    Henle's loop, the medullary and cortical thick ascending limb of
    Henle's loop, the distal convoluted tubule, the connecting tubule, and
    the cortical and medullary collecting tubules.

         Several functional parameters can be studied in nephron segments.
    Gluconeogenesis is a unique function of the proximal tubule, within
    which the S1 segment  is most active (Maleque et al., 1980; Endou et
    al., 1985). Gluconeogenesis is strongly induced by metabolic acidosis
    or by alpha1-adrenergic stimulation (Nakada et al., 1986a).

    First-generation cephalosporins, cephaloridine and cephalothin cause
    a time- and concentration-dependent decrease in gluconeogenesis, and
    it is clearly indicated that the site of nephrotoxicity of these
    antibiotics is the proximal tubule (S1, S2, S3).  Ammoniagenic
    activity is distributed in all the nephron segments, but the highest
    production rate of ammonia from glutamine is observed in the proximal
    tubule (Nonoguchi et al., 1985).  Ammoniagenesis via the purine
    nucleotide cycle from asparatate as a substrate is also high in the
    proximal tubule (Tamura & Endou, 1988).  Ammonia production is
    increased in a similar way by metabolic acidosis or potassium
    depletion (Nonoguchi et al., 1986). Cisplatin nephrotoxicity is
    morphologically known to be focussed on the S3  segment. However, this
    drug decreases ammoniagenesis from glutamine not only in S3, but
    also in S2, suggesting a discrepancy between morphological and
    biochemical evaluations, although cisplatin does not affect
    gluconeogenesis in isolated nephron segments (Nakada et al., 1986b).

         The kidney possesses various active transport processes that
    consume ATP at a high rate. Individual nephron segments require their
    own particular substrates for synthesizing the necessary ATP: this has
    been shown in both mice (Uchida & Endou, 1988) and rats (Jung et al.,
    1989). The proximal tubule cannot use glucose to produce ATP, whereas
    the other nephron segments can use it. In general, pyruvate or lactate
    is the preferred substrate in all segments.  Nephrotoxicity assessment
    by measuring cellular ATP content shows clearly that mercuric chloride
    decreases ATP content only in S2 (Jung et al., 1989) and that
    ochratoxin A nephrotoxicity localizes in S2 and S3 (Jung & Endou,
    1989).  Thus, measurement of cellular ATP in specific nephron segments
    enables possible nephrotoxicants to be evaluated. A similar principle
    can be applied by measuring intracellular free calcium (Jung & Endou,
    1990). From the biological point of view, it is essential to keep
    cellular ATP at a high level and to maintain a low concentration of
    intracellular free calcium for all living cells.  It should,
    therefore, be reasonable and useful to introduce these sensitive
    parameters to nephrotoxicity assessment, although the methods require
    special techniques for microdissecting nephron segments or special
    instruments.

         An advantage of the use of isolated tubules, as compared to  in
     vivo experiments, is that it permits a cellular environment that is
    defined both quantitatively and qualitatively.  This allows the
    relationship between the concentration of a nephrotoxin, exposure
    time, and effect to be studied. Extrarenal effects can be avoided, and
    so isolated tubules are very suitable for studying the effects of
    nephrotoxins that act directly at the tubular site.  Owing to a lack
    of polarity (Koseki et al., 1988), isolated renal cell suspension may
    have limited usefulness.

         There are several limitations associated with the use of freshly
    isolated or cultured renal cells.  Cells released by enzymic digestion
    or fresh fragments can be cultured in the presence of serum-free,
    hormonally defined culture media. This prevents fibroblast

    proliferation and encourages epithelial cell growth (Chuman et al.,
    1982), but both cell preparations lack a brush border, which may be
    critical for the active uptake of drugs and chemicals.

         Alternative approaches to obtain a preparation with an intact
    brush border include the use of different sized sieves to separate
    glomeruli and tubules (Bach et al., 1986), which may not result in a
    pure preparation of proximal tubular cells. The choice of method for
    monitoring cell viability may circumvent this, e.g., the use of
    prostaglandin synthesis (Sraer et al., 1980) to assess effects of
    chemicals selectively on glomeruli. The density gradient technique
    (Vinay et al., 1981) uses Percoll centrifugation to separate the
    different cell types to obtain a > 90% pure preparation of proximal
    tubular cells. These cells retain their viability and their GSH levels
    at > 50% for 2 h at 37 °C, and demonstrate cytochrome-P450-dependent
    mono-oxygenase activity profiles that are inducible only by
    3-methylcholanthrene (3MC).  This may be an appropriate preparation to
    study the effects of various drugs and chemicals, in both rats and
    man, that demonstrate nephrotoxicity to either the S1, S2, or S3
    regions after administration (Smith et al., 1986; Rosenberg &
    Michalopoulos, 1987).

    7.1.2.4  Primary cell cultures

         Cell to be cultured should be of well defined origin. For this,
    purification of a homogeneous glomerular or tubular cell population
    can be achieved by several methods (Jakoby & Pastan, 1979), including
    sieving techniques (Striker et al., 1980), magnetic and mechanical
    techniques (Meezan & Brendel, 1973), density gradient centrifugation
    (Scholer & Edelman, 1979; Vinay et al., 1981), and collagenase
    digestion (Curthoys & Bellemann, 1979; Belleman, 1980; Ormstad et al.,
    1981). More recently, techniques such as immunodissection, cell
    sorting, free-flow electrophoresis, and microdissection have been used
    (Pretlow & Pretlow, 1982, 1983, 1984). The advantage of using primary
    cell cultures is that it allows long exposure to xenobiotics and the
    choice of appropriate metabolites. In addition, it is possible to
    monitor a variety of cell functional, biomedical, or morphological
    responses in a dose- and time-related manner (Fry et al., 1978;
    Belleman, 1980; Fry & Perry, 1981; Bach et al., 1986).

         Mechanical or enzymic dispersal may damage cells, and once cells
    are dispersed it is generally difficult to establish  their anatomical
    identity  unless suitable markers are used. These markers include both
    the presence and absence of a range of functional and biochemical
    characteristics, such as transport systems, and an array of structural
    and functional molecules. These can best be assessed by a variety of
    histochemical and immunocytochemical methods (Bach et al., 1985,
    1987).  At present, isolated cells are generally mixtures (although
    they may be enriched) and must be used within a few hours. Primary
    cell cultures may rapidly dedifferentiate (Curthoys & Bellemann, 1979)
    or adapt to a new environment and change their characteristics as a
    result of the presence or absence of factors in the culture media,

    which may obfuscate their anatomic origins. More importantly, loss of
    a biochemical characteristic that is part of the molecular basis for
    target cell toxicity may invalidate  in vitro studies.  Changes in
    other aspects of cellular integrity can increase or decrease both the
    sensitivity and selectivity of screening methods used for cytotoxicity
    studies.

         Two approaches have been used to modulate the expression of cell
    characteristics. The polarity of epithelial cells is better expressed
    when cell are grown on permeable supports such as collagen/filters
    (Jakoby & Pastan, 1979). Similarly, the appropriate modulation of
    culture media has been used to alter rabbit proximal tubule cell
    metabolism to the gluconeogenic pathway and these cells then develop
    brush-border characteristics. Thus, media can be an important
    variable, especially because of the diverse combination of buffers and
    growth supplements used.  There are major advantages in using fully
    defined culture media (Sato & Reid, 1978), but these have not been
    widely adopted.

         Human proximal tubules have been shown to been sensitive to
    cyclosporin A (Trifillis et al., 1986), but there are no data on the
    mechanistic bases of these changes. Rat, rabbit, dog, and human
    glomerular mesangial and epithelial cells may be co-cultured or each
    type derived separately (Kreisberg et al., 1977, 1978; Foidart et al.,
    1979, 1980, 1981; Morita et al., 1980; Striker et al., 1980; Kreisberg
    & Karnovsky, 1983). Rat epithelial cells are  more sensitive to
    puromycin aminonucleoside and Adriamycin than are mesangial cells, as
    is the case  in vivo, but there is little mechanistic information.
    Rat medullary interstitial cells can be cultured at high osmolality
    and have been shown to be sensitive to a number of compounds that
    cause renal papillary necrosis.

    7.1.2.5  Established renal cell lines

         Several established renal cell lines have been studied that have
    properties reminiscent of specific parts of the nephron, such as
    LLC-PK1 (of proximal tubule type) and MDCK (of distal tubule type).
    The major disadvantage is that the exact site of origin, within the
    nephron, of each, is not known, and it may not totally represent the
    normal physiological state. However, these lines are often
    heterogeneous and there is a need to characterize them more
    systematically so as to establish where they may be useful in
    screening chemicals for toxicity or in understanding the mechanisms of
    target cell toxicity.

         Differences exist between the apical and basolateral membrane
    transport of substances into cells, which may be central to the
    mechanism of nephrotoxicity. When cells are cultured on solid
    surfaces, only apical exposure to chemicals occurs, whereas  in vivo,
    proximal tubule cells are exposed from the apical or basolateral sides
    or both. This disadvantage can be overcome by culturing renal cells on
    microporous membranes suspended in culture wells. These cells, which

    form a confluent single-cell monolayer covering the membrane within
    some days, more closely mimic the  in vivo state than those grown on
    plastic plates. They show anatomical and functional polarization. This
    culture technique allows access to the cell monolayer from both the
    apical and the basolateral sides, and apical and basolateral fluid may
    be studied simultaneously. This new experimental tool allows the study
    of transport and epithelial resistance across the cell monolayer and
    polarized uptake of various molecules, including potentially
    nephrotoxic drugs, as well as to perform a variety of analytical
    techniques.

         The various cell lines used in nephrotoxicity studies have been
    reviewed by Wilson (1986). The LLC-PK1 cell lines have a typical
    epithelial polarity and have features similar to proximal tubular
    epithelium, such as transport systems (Handler, 1983) and the enzyme
    marker gamma-glutamyltranspeptidase (Perantoni & Berman, 1979).
    Confluent LLC-PK1 cells cultured on a solid support form domes (due to
    transcellular transport), but monolayers grown on a porous membrane do
    not. More importantly these cells  have polarity and have a
    well-developed brush border. Confluent LLC-PK1 monolayers exposed to
    PCBD-GSH from the apical side are more sensitive than when exposed
    from the basolateral side. This is due to the brush border
    localization of gamma-glutamyltranspeptidase, which catalyses the
    first step of the breakdown of the conjugate to the ultimate reactive
    intermediate. Neither apical nor basolateral treatment with PCBD-NAC
    elicits any toxicity.  It is assumed that the absence of an organic
    anion transporter from these cells could explain this finding, since
    it has been established that haloalkene conjugates enter cells via the
    basolaterally located anion transporters (Lock et al., 1986). The
    absence of an organic anion transport system limits the usefulness of
    LLC-PK1 cell lines for studying nephrotoxic compounds, such as
    PCBD-NAC, that need active transport to enter the cells. However, an
    active basolateral organic cation transport system
    (gamma-glutamyltranspeptidase and dipeptidase) makes these cells
    especially useful for testing compounds that have a toxic action on
    these transport systems.

    7.1.2.6  Subcellular fractions

         It is also possible (and sometimes desirable) to use homogeneous
    or fractionated organelles, membranes, or cytoplasm  from defined
    cells for specific cell-free investigations. The constraints on the
    preparation of these systems should be apparent from the foregoing
    discussion. Subcellular fractions, such as  vesicles, nuclei,
    lysosomes, and microsomes, can be used to study subcellular 
    distribution, the interaction  between a cellular compartment and a
    chemical, and the kinetics of binding or release of substances. It is
    also possible to study specific effects, such as enzyme inhibition,
    metabolic activation, covalent binding, or the modulation of lipid
    peroxidation, using purified or commercially available biochemicals
    with appropriate cofactors and suitable techniques for monitoring
    these interactions (Bach & Bridges, 1985b, 1987).

         Many nephrotoxic agents interact with cell membranes, where they
    bind with receptors, effect transport systems, or disrupt structure
    and function  per se. Thus, membrane vesicles may be useful for
    studying these interactions and the mechanisms of cell injury. It is
    possible to isolate vesicles from the brush border and basement
    membranes to study transport systems at each site  in vitro. Williams
    et al. (1986) showed a very good correlation between the  in vitro
    binding of aminoglycosides to brush-border membrane vesicles and their
     in vivo nephrotoxicity. Inhibition of aminoglycoside membrane
    binding by polyaspartate reduces nephrotoxicity and suggests that
    binding of these antibiotics to brush-border phospholipid may be a
    crucial event in nephrotoxicity.

    7.2  in vivo experimental studies

         Current methods for diagnosing renal injury and predicting the
    health significance are not sufficient to deal with the diversity of
    possible chemical injuries (for full discussion, see Bach et al.,
    1989). This is because the kidney can undergo substantial chemically
    induced injury without any clinical indication, since subtle injury
    may be buffered within the considerable functional reserve. This masks
    a substantial amount of renal degeneration (Friedlander et al., 1989).
    Thus, for example, the single cross-sectional measurement of GFR may
    only show incipient acute or chronic renal failure.  Quantitative
    urinary enzyme excretion patterns cannot identify either the type or
    severity of renal injury, and often they do not correlate with
    morphological and functional changes (Schentag et al., 1978).

         There are a number of inherent difficulties in diagnostic
    procedures for nephropathy, which include the absence of standard
    diagnostic criteria and the inability to relate exposure to a given
    agent and the observed effect. In addition, renal functional reserve
    is a major factor that masks renal degeneration, as assessed by GFR,
    blood urea nitrogen, and creatinine, up to the point where over 75% of
    the functioning nephrons have been lost. Thus, it should be stressed
    that these factors measure incipient renal failure and that the fact
    that values are normal (something that is subject to age-related
    change and varies between the two sexes) does not signify the absence
    of renal dysfunction or even, in some cases, gross renal
    insufficiency.  Therefore, cause and effect cannot be clearly
    established on the basis of available knowledge when the renal lesion
    results from a multifactorial process with a long latency. Part of
    this uncertainty can be addressed by studies on experimental animals.

    7.2.1  Methods for assessing chemically reactive nephrotoxic
    metabolites in animals

         It is known that many nephrotoxicities that follow the
    administration of inert, relatively nontoxic chemicals are related to
    the formation of reactive electrophiles during the metabolism of these
    chemicals (Ford & Hook, 1984).  These electrophilic products can react
    covalently with nucleophilic sites on renal macromolecules such as

    protein, lipid, and DNA. The covalent binding may be measured by the
    use of radiolabelled forms of the chemical, by immunological detection
    of DNA/protein adducts (Harris et al., 1987), or by the 32P-DNA 
    postlabelling method (Reddy et al., 1984). Furthermore, chemicals that
    are metabolized to DNA-damaging intermediates may be detected  in vivo
    by measuring the alkaline elution of isolated kidney nuclei
    (Omichinski et al., 1987; Brunborg et al., 1988) or unscheduled DNA
    synthesis in isolated kidney cells (Tyson & Mirsalis, 1985) after  in
     vivo exposure of animals.

    7.2.2  Evaluation of glomerular function

         Evaluation of blood urea nitrogen is probably the most common
    procedure for indirectly evaluating GFR in experimental animals.
    Although insensitive, this test may be sufficient to establish the
    time course of chronically developing renal failure in the
    experimental setting. Serum creatinine is also used for the same
    purpose. However, owing to interferences from nonspecific chromogens,
    this test is unreliable in most experimental animals and especially in
    the rat. Although this problem may be overcome, it has not been dealt
    with adequately in most available studies, thus generating wide
    scatter in "normal" ranges.

         In animals, more subtle changes in GFR occurring during
    subchronic and chronic studies should be assessed by evaluating the
    clearance of exogenous substances such as inulin, EDTA, or
    iothalamate.  The latter may be determined  either by measuring the
    radioactivity of labelled material or by means of reliable HPLC
    methods (Prueksaritanont et al., 1984). Furthermore, the same HPLC
    method may be used to measure PAH and to assess other haemodynamic
    parameters. Two (or more) clearance periods should be calculated and
    averaged in order to ensure greater accuracy.

         There is a growing body of evidence to suggest that reduced renal
    reserve due to hypertension/hyperfunction/hyperfiltration of remnant
    nephrons is important in the course towards end-stage renal disease.
    This can in part be lowered by reducing protein intake and blood
    pressure. The concept of renal functional reserve includes the
    evaluation of renal blood flow and GFR by measuring their increase
    after protein load or the administration of aminoacids, glucagon, or
    vasodilatory drugs. At present such tests do not have defined
    standardized stimuli, and there are no data on their use for detecting
    nephrotoxicity.

    7.2.3  Evaluation of tubular functions

         In experimental animals, tubular dysfunctions are usually
    detected through simple and inexpensive tests, such as those of
    glycosuria, enzymuria, and osmolality, which may provide other useful
    information. Some of these tests are sensitive enough to detect acute
    tubular damage, although caution must be exercised in predicting
    specific effects on transport processes or cell viability on the basis

    of data obtained from  in vivo experiments (Berndt, 1981). More
    subtle renal changes occurring during chronic studies may be evaluated
    by measuring the renal clearance of lithium (Dieperink et al., 1983;
    Daugaard et al., 1988b). This non-invasive method is applicable both
    to human (Thompson et al., 1984) and animal studies. The loss of renal
    tubular functions can be assessed by test procedures that impose one
    or more stressing conditions to force compensatory changes, e.g.,
    maximal urinary dilution/concentration or
    acidification/alkalinization, and maximal tubular reabsorption of
    glucose and phosphate. The value of these tests is limited for group
    studies by practical considerations but may be useful for individuals.

    7.2.4  Proteinuria

         Proteinuria is the loss of proteins following increased
    permeability of the glomerular barrier, reduced tubular reabsorption
    of filtered proteins, or shedding of specific constituents into the
    urine as a consequence of cellular turnover or selective tissue
    damage. Since pathological changes either at the glomerular or
    tubulo-interstitial level may occur even in the absence of a
    substantial reduction in GFR, the evaluation of proteinuria may also
    be useful in some circumstances to detect renal dysfunction occurring
    either at the glomerular or at the tubular level. Proteins may be
    measured by nonspecific assays, by immunochemical methods or by their
    enzymic activity. Sensitive methods have been developed to detect
    small amounts of proteins in microlitre quantities of unconcentrated
    urine.

    7.2.4.1  Total proteinuria and electrophoretic pattern

         Measurement of total proteinuria and electrophoretic separation
    of single proteins provide a comprehensive approach to chemically
    induced renal dysfunction.

         The rationale for such an approach relies on the
    pathophysiological mechanisms controlling the renal handling of
    plasmaproteins.  Proteins of high relative molecular mass (> 45 000
    Daltons) are usually confined to the vascular compartment by basal
    membranes. Furthermore, the glomerular polyanion acts as a selective
    filter that retains negatively charged proteins, such as albumin,
    because of electrostatic interactions. The glomerular pore size is
    thought to have an important role in retaining proteins of higher
    relative molecular mass (e.g., immunoglobulins). Proteins with lower
    relative molecular mass (< 45 000 Daltons)  pass the glomerular 
    barrier with sieving coefficients inversely related to their mass.
    Filtered proteins of low relative molecular mass are efficiently taken
    up by the proximal tubules (more than 99%). Even slight decreases in
    tubular fractional reabsorption due to tissue damage or dysfunction
    will result in increased low relative molecular mass proteinuria (Fig.
    16).

    FIGURE 16

         On the basis of the electrophoretic pattern, proteinuria may
    reveal glomerular damage, tubular dysfunction or mixed patterns. The
    glomerular damage may be selective (mostly due to a loss of glomerular
    polyanion) or unselective (involving more extensive damage, and
    glomerular hyperfiltration and hypertension). However, it should be
    recognized that some features unique to experimental animals may
    account for large variations, which could lead to wrong conclusions.
    For instance, marked sex-, age-, and diet-related changes may occur,
    especially in the male rat (Neuhaus et al., 1981).  Even if most of
    these changes have a counterpart in man, they are amplified by the
    lack of variations in housing conditions. As a result, the young male
    rat may physiologically show "tubular" proteinuria, whereas the aging
    rat displays "glomerular" patterns, owing to a spontaneous
    nephropathy, which can be prevented in part by reducing dietary
    protein content. Depending on other factors, especially diet and
    concomitant treatments, such electrophoretic patterns may be accounted
    for by underlying mechanisms and related morphological changes.

         Owing to their limited affinity for most dyes, proteins of low
    relative molecular mass are better identified by the quantitative
    measurement of single components such as ß2-microglobulin  (Viau et
    al., 1986b). Because of its peculiar metabolism, the urinary excretion
    of alpha2u-globulin, a sex-related protein of low relative molecular
    mass, cannot be recommended as a marker of tubular damage.
    Furthermore,  its renal handling and disposition may interfere with
    those of other proteins, accounting for some of the spontaneous
    changes occurring in the pattern of proteinuria from the male rat.
    Thus, when evaluating proteinuria, preference should be given to
    female rats, since in this case extrapolation to man seems to be less
    affected by species-related problems.

    7.2.4.2  Urinary excretion of single plasma proteins

         Very sensitive immunochemical methods are available for measuring
    the urinary excretion of single plasma proteins, such as IgG, albumin,
    ß2-microglobulin  and alpha2u-globulin in the rat (Bernard et al.,
    1988). In addition to better analytical features, in terms of
    sensitivity, specificity, accuracy, and reproducibility, the
    quantification of single urinary proteins has two other inherent
    advantages. Firstly, single proteins may be significantly increased
    without giving rise to pathological values in total proteinuria
    (Barratt, 1983). Secondly, the power of experimental studies is
    greatly increased by quantitative data making it possible to use
    parametric tests.

    7.2.4.3  Enzymuria

         Several different enzymes have been studied (Table 8), but none
    satisfies all the criteria for nephrotoxicity (Dubach et al., 1989).
    Enzymes are not uniformly distributed along or between nephrons.
    Although it should be possible to localize the area of kidney damage
    on the basis of the pattern of enzymuria, the site-selectivity of
    single enzymes is questionable. The failure to recognize selective
    damage by measuring enzyme activity may be accounted for by two
    factors. Firstly, it is possible that chemically induced early renal
    changes are less selective than advanced lesions preceding end-stage
    renal disease. Secondly, the poor analytical features of most enzyme
    measurements in urine may give rise to aspecific patterns. This
    important question may be addressed by measuring immunoreactive
    antigens, including enzymes, since the use of reliable immunochemical
    techniques would limit the effects of analytical problems.

         Most enzymes are stable over a narrow range of pH, and their
    activity may be affected by the presence of inhibitors such as urea
    (Price, 1982). It should be stressed that all rat enzyme studies must
    be carried out under carefully controlled conditions, after an
    adequate period of acclimatization, and after changing to day instead
    of night feeding. Urine must be minimally contaminated with
    microorganisms. This is achieved by surrounding the urine collecting
    vessel with ice, so that bacteria cannot multiply as readily as in
    normal metabolic cages (Berlyne, 1984).

    7.2.4.4  Immunoreactive tissue constituents

         Tissue constituents may be released into urine due to increased
    cellular turnover or cell death and may be detected by immunochemical
    methods. Monoclonal antibodies have been produced against both rat
    (Tokoff-Rubin, 1986) and human brush-border antigens (Mutti et al.,
    1985; Mutti, 1989). In both cases, cross-reactivity between species
    has been shown (Mutti, 1987, 1989). The specificity of such an
    approach relies on the site-selectivity of target proteins and on the
    advantages of monoclonal antibodies, including monospecificity and
    reproducibility of reagents. Its sensitivity has been proved by

        Table 8.  Some enzymes used as an index of nephrotoxicity
                                                                        
    Enzymes                                 Cellular location
                                                                        
    Alanine aminopeptidase                  brush border
    Alkaline phosphatase
    gamma-Glutamyltransferase
    Maltase
    Trehalase

    Glutamic oxaloacetic transaminase       cytosol

    Glutamic pyruvic transaminase
    Lactate dehydrogenase
    Malate dehydrogenase

    N-Acetyl-ß-D-glucosaminidase            lysosome
    Acid phosphatase
    ß-galactosidase
    ß-glucosidase
    ß-glucuronidase
    Glutamate dehydrogenase                 mitochondria
                                                                        
    
    comparison with other markers in various situations, but it needs
    further validation in carefully designed chronic studies, since the
    prognostic value of slight changes in such a sensitive test is
    currently unknown.

         In general terms, it has been through the use of
    histopathological studies on kidney tissue that advanced renal lesions
    have been identified. This approach has many inherent advantages, and
    represents the method of choice, particularly if it can be used for
    the early diagnosis of renal lesions or dysfunction in experimental
    animals being used for nephrotoxicity screening studies.  For clinical
    investigations in human and population studies, it is obviously the
    least desirable of the techniques that are available.

    7.2.4.5  Urinary excretion of prostaglandins

         The urinary excretion of PGs (mainly PGE2) may reflect the rate
    of renal synthesis. This is modified in several nephropathies and may
    be affected by a number of nephrotoxic chemicals.

    7.2.5  Clinical context

         At present, the clinical diagnosis of toxic nephropathy still
    relies heavily on the case history of patients showing symptoms and/or
    laboratory abnormalities suggesting chronic renal failure without any
    obvious recent cause.

         In such circumstances, history is the cornerstone of diagnosis.
    It can only be made by excluding other known conditions or risk
    factors that lead to renal insufficiency and assessing exposure to
    nephrotoxins, which has to be consistent with known
    dose-effect/response relationships and with the temporal sequence of
    events leading to the observed effect. For immune reactions, the role
    of individual susceptibility should also be considered. In some
    circumstances, kidney toxicity may be considered as a factor
    contributing to the clinical outcome in a multifactorial process. In
    this case, it is difficult to distinguish determinants from
    predisposing and/or aggravating factors.

         There are two well-recognized clinical syndromes that result from
    immunologically mediated glomerular disease. These are the nephrotic
    syndrome and the nephritic syndrome. The nephrotic syndrome is
    characterized by massive proteinuria, hypoalbuminaemia (from
    proteinuria) and generalized oedema. The syndrome is the result of
    pathology that affects the integrity of the GBM. Although the cause of
    85% of nephrotic syndrome is not known, it is the syndrome most
    frequently occurring with toxic nephropathies, it is generally dose
    related, and it is reversible. Glomerular pathology is most often
    membranous glomerulonephritis or circulating immune-complex disease.
    The nephritic syndrome is characterized by haematuria and a decreasing
    GFR and is likely to be accompanied by some degree of hypertension. 
    The pathological changes in the glomeruli are usually caused by those
    diseases or toxic agents that produce an inflammatory proliferative
    response within the glomeruli.  The proliferation may involve
    endothelial mesangial or epithelial cells and may be associated with
    an inflammatory cell infiltration.

    7.2.6  Radiological techniques

         Radiocontrast media can be used to study the kidney either by
    conventional or by retrograde pyelography. There are a number of
    limitations to this technique, such as the need for adequate renal
    function by which to image the organ. There is adequate evidence that
    radiocontrast agents have a nephrotoxic potential. Media of high
    osmolality are especially likely to precipitate renal failure. In
    addition, those patients who are dehydrated or with reduced blood
    volume are at special risk. There is also evidence to show that
    patients who have multiple myeloma represent a special risk group.

    7.2.7  Other non-invasive renal assessment

         Whereas gamma camera renography and ultrasound are
    well-established techniques for the assessment of renal function,
    increasing use is being made of ultrasound linked with Doppler flow,
    nuclear magnetic resonance imaging, and spectoscopy; ureteral
    fibroscopy may be used to view the pelvis directly. Ultrasound
    evaluation of the kidneys provides a means of excluding obstruction as
    a cause of anuria or oliguria in ARF. It may also reveal papillary
    necrosis or perirenal haematomas and obviate the intrinsic dangers of

    pyelography. Radionuclide scans can be used to identify major
    atherothrombotic events and cortical necrosis. They can also be used
    to show reduced renal blood flow, which is usually preserved (some 50%
    of normal) in acute tubular necrosis, but is severely compromised in
    acute glomerulonephritis and vasculitis. Renal biopsy should not be
    used for ARF unless the duration exceeds three weeks and there is no
    obvious cause. It can then help diagnose glomerulonephritis,
    vasculitis, and AIN.

    8.  DETECTION OF NEPHROTOXICITY IN HUMANS

         Traditional methods of diagnosing renal damage and predicting
    their health significance are not sufficient to deal with the
    diversity of chemically induced lesions (Bach et al., 1989). This is
    because the kidney can undergo substantial chemically induced injury
    without any clinical indication, owing to its considerable functional
    reserve. This masks a substantial amount of renal degeneration
    (Friedlander et al., 1989). Thus, for example, the single common
    measurement of GFR may only show incipient acute or chronic renal
    failure.

         Most nephropathies in humans are of unknown origin. There is,
    however, some indirect evidence that toxicant exposure could be
    involved. For example, 80% of the cases of membranous nephropathy are
    of unknown origin. Among the remaining 20%, half of them have been
    found to be associated with drug exposure. It is likely, therefore,
    that a percentage of those cases of unknown origin are also related to
    toxic exposure.

    8.1  Markers of nephrotoxicity

         Over the last ten years, new biochemical and immunochemical
    methods have been developed for detecting early renal changes in
    humans. Their application has given rise to a number of markers, some
    of which also are available for studies on experimental animals and
    have been reviewed in section 7.2. Although some tests may open new
    perspectives in terms of selectivity, sensitivity, and specificity,
    most of the recently developed markers need further validation. Hence,
    some conceptual and methodological problems must be addressed and new
    methods critically evaluated. This is especially important when
    implementing screening programmes for the early detection of kidney
    damage and/or dysfunction in humans.

    8.1.1  General requirements

         Markers to be used for screening purposes in population groups
    should fulfill certain general criteria, including specificity,
    non-invasiveness, sensitivity in detecting early renal functional
    changes, and predictive value for development of renal insufficiency.
    Analytical methods must be reproducible, easy to perform, and
    applicable to a large number of samples, and the samples must be
    stored appropriately to ensure stability.

         The prevalence of true positive results among subjects who are
    actually ill indicates the sensitivity of the marker, whereas the
    prevalence of true negative findings among healthy individuals is a
    measure of its specificity.

         Some markers meet the requirement of specificity, although it may
    be difficult to establish clear-cut distinctions in individual cases.
    For instance, whereas gross changes in proteinuria involving proteins

    of low and high relative molecular mass indicate tubular and
    glomerular damage, respectively, marginal increases over reference
    limits may be due to either condition (Mutti, 1989).

         The predictive value of markers of nephrotoxicity has not been
    systematically tested, but it is usually assumed that such markers as
    the urinary excretion of single plasma  proteins  are highly 
    sensitive but somewhat aspecific.  This is due to the fact that they
    may be increased by a number of physiological conditions (e.g.,
    physical workload, posture, pharmacological effects of exogenous
    substances, or even meat meals).  On  the contrary, other markers such
    as serum creatinine are considered to be relatively specific, but
    relatively insensitive, since they reflect late changes, occurring
    when more than 50% of the renal reserve has been lost. Within these
    extremes, there are various possibilities which should be carefully
    weighed according to the methodological context dealt with.

         The predictive value of markers, in terms of sensitivity and
    specificity, should be taken into account when evaluating individual
    results. For some parameters (e.g., proteinuria involving proteins of
    low relative molecular mass) minor dysfunctions at the tubular level
    will result in deviations of several orders of magnitude from
    reference values, whereas the same relative change in other parameters
    (e.g., brush-border antigens, enzyme activities) may indicate severe
    tubular injury.

         In the clinical context, it is often difficult to establish
    correct etiological diagnoses because of the long latency between
    exposure and the development of overt disease. For the same reason, it
    is sometimes difficult in epidemiological work to assess the
    prognostic value of early changes. Nonetheless, it is clear that
    markers used in the clinical context must be specific enough to avoid
    further undue and sometimes invasive procedures, which tend to
    accumulate once a subject enters a diagnostic decision tree (Mold &
    Stein, 1986).

         The quantification of any constituents in urine may only be
    obtained by assessing urine flow rate. The accuracy of timed urine
    samples (especially when obtained during epidemiological surveys) is
    unreliable and most analytes may be unstable. There are still a number
    of factors that can confound or modify the renal response, such as
    time of sampling (owing to spontaneous rhythms), posture, physical
    work load, and diet. Thus, there is a need to use standardized
    procedures to limit possible interferences that can increase the
    variability between and within subjects. To overcome these
    methodological problems, there is an increasing tendency to use spot
    samples, generally the second sample of the day, and to express
    results as a function of creatinine.

    8.1.2  Diagnostic value

         The predictive value of a marker only in part contributes to its
    diagnostic validity, which is defined in terms of the probability that
    the classification based on test values actually corresponds to the
    subject's condition. The diagnostic value may be both positive and
    negative, the positive diagnostic value being the probability that a
    subject classified as positive by the test is actually ill, and the
    negative diagnostic value being the probability that subjects negative
    at the test are actually healthy individuals. It must be stressed that
    the diagnostic validity is only marginally affected by the predictive
    value of the markers, since in most screening programmes a low
    prevalence of disease is the major determinant of pitfalls.

         Thus, markers must be selected according to the condition under
    evaluation. When studying groups at risk, emphasis should be given to
    sensitivity, whereas specificity should receive adequate attention in
    the clinical setting.

    8.1.3  Prognostic value

         Another important property of markers, which has so far only been
    addressed in a few studies, is their prognostic  value, i.e. their
    ability to predict the "natural" course of the disease. In most cases
    early functional changes may be compensated and structural damage may
    be repaired. However, both conditions may also trigger a cascade of
    events leading to end-stage renal disease. The ability to target
    groups actually at risk and to predict the "natural" evolution of the
    disease is part of effective prevention.  This can be achieved by
    reducing exposure (primary prevention) or by establishing an
    individual diagnosis at reversible stages (secondary prevention).

         Microscopic examination of urine sediment, although impractical
    in population studies, may help in establishing individual
    differential diagnosis. Even if it is insensitive for detecting
    nephrotoxicity, this test is suitable at the individual level to
    exclude confounding factors accounting for increased excretion of
    plasma proteins (e.g., leukocyturia, haematuria, or bacteriuria).

    8.2  Screening for nephrotoxicity in humans

         In the epidemiological context, it can be assumed that negative
    results are recorded in actually healthy subjects, whereas the
    probability of false positive values is rather high because of the low
    prevalence of the disease. Thus, health surveillance programmes should
    be mainly aimed at excluding harmful situations rather than at
    identifying ill people. This goal may be achieved using very sensitive
    markers of subclinical renal impairment. A clinical diagnosis in
    individuals showing abnormal test values should then be established
    only after repeated measurements and complementary confirmatory tests.
    Practical considerations are also important when planning population

    studies, where sampling procedures that are invasive or too elaborate
    cannot be included in the protocol.

    8.2.1  Glomerular filtration

         Clearance procedures can seldom be adopted when screening
    population groups, since only single blood and spot urine samples are
    usually available. However, it must be stressed that some tests
    fulfilling the above objective have proved their diagnostic and
    prognostic validity and for this reason they are now also being
    employed in clinical practice.

         An indirect method of assessing GFR by using a single blood
    sample is the measurement of serum creatinine and the expression of
    its reciprocal value adjusted to constant body surface area. This has
    been correlated with GFR (Siersbaek-Nielson et al., 1971). The same
    concept applied to ß2-microglobulin  would increase the sensitivity of
    such an approach, owing to the higher relative molecular mass and the
    use of more accurate and reproducible analytical methods (Wibell et
    al., 1973).

    8.2.2  Tests designed to assess selective dysfunction

         Serious renal diseases (e.g., nephrotic or Fanconi's syndromes)
    may occur in the absence of haemodynamic changes. In such
    circumstances, renal insufficiency is a late event that is preceded by
    earlier, though serious, selective dysfunctions occurring at the
    glomerular and proximal tubular level. These may be revealed by
    measuring single plasma proteins in urine (see also section 8.3.3.).
    The excretion in urine of proteins of high relative molecular mass
    generally reflects glomerular dysfunction, whereas urinary excretion
    of proteins of low relative molecular mass may be indicative of
    tubular dysfunction. The assessment of protein electrophoretic
    patterns has been discussed in section 7.2.4.

    8.2.3  Tests designed to assess tissue damage

         Tubular dysfunction is not necessarily associated with
    histopathological changes. It has been shown that pharmacological
    inhibition of tubular uptake may account for proteinuria (involving
    proteins of low relative molecular mass) observed after protein load
    (Buzio et al., 1989). Furthermore, highly selective damage to a
    tubular segment may be functionally compensated by other segments of
    the nephron. Methods aimed at detecting cellular damage may thus help
    both to show subclinical lesions and to interpret associated
    dysfunctions. The urinary excretion of tissue constituents may be
    measured through the enzymic or immunochemical characteristics of
    material shed into the urine.

         Renal functional changes may be reversible, depending on the
    efficiency of repair mechanisms and the cessation of exposure to the

    offending agent. Repeated monitoring may help to distinguish
    progressive renal disease from transient lesions (Thornley et al.,
    1985).

    8.2.3.1  Enzymuria

         The general principles describe in section 7.2.4.3. also apply to
    humans. Even when using freshly voided spot samples, urine is a
    hostile environment for most enzymes. Only a few enzymes (e.g.,
     N-acetly-ß-D-glycosaminidase) show acceptable stability and
    analytical precision (Price, 1982). Table 8 lists some human urinary
    enzymes that have been used in nephrotoxic studies.

    8.2.3.2  Immunoreactive tissue constituents

         Tissue constituents (including enzymes) are physiologically shed
    into the urine as a consequence of cell turnover and metabolism. When
    they are detected by immunochemical methods (e.g., immunofluorescence,
    enzyme-linked immunosorbent assay), they are referred to as antigens.
    Tamm-Horsfall glycoprotein is a specific renal protein, localized on
    the membrane of cells of the thick ascending limb of the loop of
    Henle, which is excreted in the urine at a relatively constant rate.
    This can increase following injury to the distal part of the tubule
    and is depressed when the renal mass is reduced (Thornley et al.,
    1985). Although increased excretion of proximal tubular antigens was
    reported twenty years ago in various clinical situations such as
    tubular necrosis, allograft rejection, and
    cephalotin/gentamicin-induced nephrotoxicity (Antoine et al., 1969;
    Rosenmann et al., 1971; Scherberich et al., 1976, 1984), only recently
    has it been possible to improve significantly the specificity and
    reproducibility of such an approach, relying on the properties of
    monoclonal antibodies.

         Monoclonal antibodies to human brush-border antigens have been
    produced by Mutti et al. (1985). Monoclonal antibodies may be
    conveniently employed in ELISA procedures set up to detect trace
    amounts of antigens in biological samples. Although the BB-50
    brush-border antigen was also localized in peritubular capillaries
    (Mutti et al., 1985), subsequent work lead to the identification of a
    monoclonal antibody reacting with an antigen located specifically in
    the brush border and thus called BBA or brush-border antigen (Mutti et
    al., 1988).  Promising results have been obtained in several
    cross-sectional investigations on groups at risk of chemically induced
    renal damage (for a review, see Mutti, 1989).  Similar results were
    obtained with monoclonal antibodies to rat brush-border cross-reacting
    with the human kidney (Tokoff-Rubin et al., 1986).

         Monoclonal antibodies have also been produced that react
    specifically with other segments along the nephron (namely S3) where
    the intestinal isoform of alkaline phosphatase is located (Verpooten
    et al., 1989). They could be used to monitor the effects of chemicals
    (e.g., mercury) acting selectively on the straight part of proximal

    tubules. All of these recently developed tests need further validation
    in well-designed longitudinal studies, since their prognostic value is
    currently unknown.

    8.3  Clinical investigations

         The clinical diagnosis of toxic nephropathy frequently relies
    heavily on the history of patients who occasionally showed symptoms
    and/or laboratory abnormalities suggesting chronic failure without any
    obvious recent cause.

         Diagnosis can only be made by excluding other known conditions or
    risk factors. This has to be assessed by estimating exposure to
    nephrotoxins in relation to the known dose-effect/response
    relationship and the temporal sequence of events that follow such
    exposure. The role of individual susceptibility should also be
    considered. In some circumstances, nephrotoxicity may be one factor in
    a multifactorial process leading to clinical renal disease.
    Retrospective data about exposure and early changes in renal function
    are usually not available.

         Most progressive kidney diseases have a subtle onset. This is the
    reason why the markers designed for use in epidemiological studies are
    becoming part of clinical investigations.

         Owing to the short latency between exposure and the development
    of severe symptoms, acute renal failure should be accurately diagnosed
    in all cases. The simple evaluation of serum creatinine (increases of
    3-5 mg/litre or 50% above baseline values) makes it possible to detect
    nephrotoxic reactions.

    8.3.1  Invasive techniques

         The use of invasive techniques is limited by ethical constraints.

    8.3.1.1  Biopsies from humans

         Renal biopsy is the only way to identify glomerular, tubular, or
    interstitial renal diseases. However, the risk-to-benefit ratio must
    be considered carefully in each individual patient being evaluated,
    and there is no universal agreement on the conditions in which
    percutaneous renal biopsy may be useful.

    8.3.1.2  Autopsy in humans

         Autopsies performed on patients with end-stage renal failure only
    reveal the etiology of the renal disease in a small percentage of
    cases.

    8.3.2  Tests designed to assess glomerular filtration and renal blood flow

         Traditional methods based on inulin and PAH clearances are
    progressively being substituted by more convenient methods such as the
    clearance of 51Cr-EDTA and 99Tc-DTPA to assess GFR or the
    clearance of 125I-hippuran to measure renal plasma flow. These
    techniques are thought to be more sensitive than creatinine clearance
    and more accurate than colorimetric methods.

    8.3.3  Proteinuria

         Section 7.2.4. contains a detailed discussion of experimental
    studies involving proteinuria. Table 9 gives information on some
    proteins excreted in human urine. Values that are in excess of the
    normal range may be indicative of renal dysfunction.

        Table 9.  Excretion rates of common urine proteins
                                                                        
    Protein                     Relative molecular mass    Normal range
                                      (Daltons)
                                                                        
    Albumin                              68 000          < 30 mg/day
    ß2-microglobulin                     12 800          <  0.3 mg/day
    Retinol-binding Protein              21 400          <  0.3 mg/day
    IgG                                 160 000             2-3 mg/day
    alpha2-microglobulin               ± 30 000                 ?
                                                                        
    
    8.3.4  Tests designed to assess selective damage

         Tissue constituents may be shed into the urine following toxic
    damage to specific structures. All of these specific antigens may be
    detected by using immunochemical or biochemical methods designed to
    measure their concentration or enzymic activity, respectively.
    Although they have been designed expressly for evaluating population
    groups, these tests may also be useful in the clinical setting to
    monitor patients at risk. Even if they are very sensitive and
    relatively specific, they need further validation under carefully
    controlled clinical conditions, particularly in longitudinal studies.

    9.  SUMMARY AND CONCLUSIONS

         The kidneys are the main organs of excretion and homeostasis for
    water-soluble molecules. The functional unit of the kidney is the
    nephron, essentially a continuous tube of highly specialized
    heterogeneous cells, which exhibits marked structural, functional, and
    biochemical organization. There are pronounced differences between the
    nephrons themselves, depending on the cortical localization of the
    individual glomeruli. This complex structural organization, combined
    with differences in regional vascularity arising from the specialized
    vascular arrangement, produces a highly complex heterogeneous organ.

         Much of the anatomical and functional understanding developed
    from animals is directly applicable to the human kidney. However, the
    biomedical and metabolic processes in the human kidney, as well as the
    differences among animal species, have not been as thoroughly
    elucidated. Thus, the ability to extrapolate the effects of chemicals
    among species is limited.

         Several chemicals (both therapeutic and non-therapeutic) have
    toxic effects on one or more anatomical elements of the kidney. Toxic
    effects may be acute or chronic, and they may be direct or mediated
    indirectly through immunological mechanisms. The health impact of
    nephrotoxic chemicals is related to risk factors, which include the
    intergrade of the renal functional reserve and factors such as
    pre-existing renal damage, disease, age, sex, and diet.

         The epidemiology of chemically induced nephrotoxicity by
    individual chemicals or in mixed exposures has been inadequately
    studied. The contribution of chemicals to the overall incidence of
    nephropathy and of chronic renal failure is, with few exceptions,
    undefined. In the case of some occupationally exposed groups and
    analgesic associated renal disease, there has been extensive research
    that has shown variations in incidence between groups and countries.
    However, it is estimated that up to 18% of end-stage renal disease may
    be due to analgesic nephropathy and up to 5% to other toxic
    nephropathies. About 50% of end-stage renal disease is of unknown
    etiology. A major problem in assigning a cause for end-stage renal
    disease is the long latency and/or slow development of chronic renal
    failure, which makes retrospective identification of the causative
    agent difficult. In only a few cases is measurement of tissue (body,
    kidney) levels of chemicals relevant to the diagnosis. A further
    problem has been the lack of consistency in diagnostic and
    pathological criteria and terminology.

         Renal anatomical and physiological differences make direct
    extrapolation from experimental systems  (in vivo and  in vitro) to
    humans difficult. There are very few examples of nephrotoxic chemicals
    where there are adequate comparable data from animals and humans to
    form a firm basis for the assessment of potential human health risk.

         Chemicals may damage selectively vulnerable kidney structures or
    activate immunological mechanisms. Mechanisms of renal injury fall
    generally into two categories: (a) immunologically induced disease of
    acute interstitial nephritis; (b) those that primarily affect the
    glomerulus by either anti-GBM-mediated antibodies or immune complexes.
    Another major group is composed of diseases initiated by chemicals or
    their metabolites that interfere with cellular biochemical and
    haemodynamic effects, etc. Factors that can modify cellular injury by
    toxicants include cellular transport systems, pinocytosis, metabolic
    degradation, and interaction with cellular proteins, lipids,
    membranes, DNA, and perhaps other cellular constituents.

         The increasing use of therapeutic agents and chemicals increases
    the possibility of nephrotoxicity. Nephrotoxicity induced by
    therapeutic agents depends on the dose and duration of exposure (e.g.,
    combination analgesics leading to renal papillary necrosis).
    Nephrotoxic effects of analgesics, antibiotics (such as the
    aminoglycosides), anticancer agents (such as  cis-platinum), and a
    variety of other agents have been investigated extensively.  Chemicals
    frequently used in industry or the home, e.g., chlorinated
    hydrocarbons and ethylene glycol, also have the potential to produce
    renal damage. Environmental chemicals such as lead and cadmium are
    capable of inducing nephrotoxicity. These agents act as toxicants
    after intracellular accumulation of the parent compound or after renal
    or extrarenal hiptransformation. Multichemical exposure may result in
    antagonistic or synergistic responses.

         Turnours of the kidney and urinary tract account for less than
    2-3% of all human cancers, but the frequency of these tumours is
    increasing, suggesting a role for environmental factors. Although many
    drugs and chemicals have been shown to be carcinogenic in experimental
    models, only a few specific substances have been related to turnouts
    in man. These include asbestos (renal adenocarcinoma), analgesic abuse
    (transitional carcinomas), and lifestyle factors (tobacco use,
    alcohol, coffee). There are population groups with urinary tract
    cancer of as yet undetermined etiology. Occupational chemicals have
    been related to the etiology of cancer of the urinary bladder. Many
    drugs and chemicals cause interstitial nephritis, which may in itself
    be a factor in developing urinary tract cancers.

         No single  in vivo or  in vitro method is sufficient to assess
    chemically induced renal dysfunction. Therefore, it is advisable not
    only to screen compounds for nephrotoxic potential, but to incorporate
    mechanistic studies of target cell toxicity into the experimental
    protocols. To accomplish this, both  in vivo and  in vitro
    investigations should be utilized in concert.  In vitro studies
    involve those where anatomical cellular relationships are maintained
    (e.g., isolated perfused kidney and tubules, renal cortical slices,
    and isolated nephron segments) and those where renal cells are used
    (e.g., cell suspension, primary cell cultures, established cell lines,
    and subcellular fractions).  In vivo studies utilize both invasive
    and non-invasive techniques. Invasive procedures include

    histopathological and routine renal function measurements.
    Non-invasive procedures permit repeated assessment of renal function
    in animals through the measurement of an array of renal function tests
    (glomerular filtration, tubular function, proteinuria, enzymuria,
    etc.). Specialized biochemical tests should be used where relevant.
    The appropriate mixture of  in vivo and  in vitro experiments will
    reveal whether or not chemicals are nephrotoxic and give insights into
    potencies, sites, and mechanisms of toxicity.

         Traditional methods for the assessment of renal function in
    humans are inadequate for the timely diagnosis of chemically induced
    renal dysfunctionand prediction of its health significance. The lack
    of specific, early markers for nephrotoxicity is particularly
    troublesome. Non-invasive assessment of nephrotoxity should employ
    markers of high specificity, sensitivity for detection of early renal
    changes, and predictive value for the development of renal
    insufficiency. Present techniques for monitoring glomerular or tubular
    function are useful only when severe renal damage has developed.
    Although immuno-reactive tissue constituents are being suggested as
    appropriate markers, their suitability needs to be validated in
    well-designed longitudinal studies.

    10.  RECOMMENDATIONS

    1.   A continued effort should be made to develop and validate more
    selective and specific markers, including monoclonal antibodies, for
    assessment of renal dysfunction in animals. These markers may be
    applicable to humans.

    2.   The data base for predicting the potential of chemicals for human
    nephrotoxicity should be extended. This includes development and
    validation of experimental animal approaches ( in vivo and  in
     vitro), alternative methods for studying nephrotoxicity, information
    on interspecies differences, and experience from the preclinical
    evaluation of new therapeutic agents in humans.

    3.   Epidemiological studies (i.e. prospective studies in occupational
    and general population groups exposed to nephrotoxic chemicals or
    involved with analgesic abuse) should be reinforced.

    4.   More effort should be made to establish the role of chemicals in
    the etiology of renal disease at the earliest diagnostic stage (e.g.,
    work history, tissue monitoring for nephrotoxins).

    5.   Understanding of the mechanisms of action of nephro-toxicants
    will be helpful in the prevention and clinical management of unwanted
    renal effects, and may help in predicting the nephrotoxic potential
    for new drugs and chemicals. Areas of particular importance for
    further research are:

    *    immunological mechanisms;

    *    direct effects of chemicals on membranes, including mechanisms of
         lipid peroxidation, membrane/chemical interaction, ion shifts,
         and receptor-mediated events;

    *    activation of proto-oncogenes and cell differentiation;

    *    regulation of cellular metabolism.

    6.   There is a need to identify and correlate specific functions with
    discrete anatomical locations within the kidney.

    7.   The role of genetic variation and susceptibility to the toxic
    effects of drugs and chemicals should be studied further.

    8.   The relationship between nephrotoxicity and renal carcinogenesis
    (e.g., mycotoxins and Balkan nephropathy) needs further study.

    REFERENCES

    AHMADIZADEH, M., KUO, C.-H., & HOOK, J.B. (1981) Nephrotoxicity and
    hepatotoxicity of chloroform in mice: effect of deuterium
    substitution. J. Toxicol. environ. Health, 8: 105-111.

    AINSWORTH, S.K., SWAIN, R.P., WATABE, N., BRACKETT, N.C., PILIA, P.,
    & HENNIGAR, G.R. (1981) Gold nephropathy, ultrastructural fluorescent
    and energy-dispersive X-ray microanalysis study. Arch. Pathol. lab.
    Med., 105: 373-378.

    ALDEN, C.L. (1989) Male rat specific alpha2u-globulin nephropathy and
    tumourigenesis. In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity:
    Extrapolations from  in vitro to  in vivo, and animals to man, New
    York, London, Plenum Press, pp. 535-542.

    ALDEN, C.L., KANERVA, R.L., RIDDER, G., & STONE, L.C. (1984) The
    pathogenesis of the nephrotoxicity of volatile hydrocarbons in the
    male rat. In: Mehlman, M.A., Hemstreet, C.P., Thorpe, J.J., & Weaver,
    N.K., ed. Renal effects of petroleum hydrocarbons, Princeton, New
    Jersey, Princeton Scientific Publishers, pp. 107-120 (Advances in
    Modern Environmental Toxicology, Vol. VII).

    ANAND-SRIVASTAVA, M.B., VINAY, P., GENEST, J., & CANTIN, M. (1986)
    Effect of atrial natriuretic factor on adenylate cyclase in various
    nephron segments. Am. J. Physiol., 251: F417-F423.

    ANDERS, M.W. (1980) Metabolism of drugs by the kidney. Kidney Int., 8:
    636-647.

    ANDERSON, P.M. & SCHULTZE, M.O. (1965) Cleavage of
    S-(1,2-dichlorovinyl)-L-cysteine by an enzyme of bovine origin. Arch.
    Biochem. Biophys., 111: 593-602.

    ANDERSON, S. & BRENNER, B.M. (1986) Effects of aging on the renal
    glomerulus. Am. J. med., 80: 435-442.

    ANTOINE, B., NEVEU, T., LESKI, M., & BACH J.F. (1969) Histuria during
    renal transplantation. Transplantation, 8: 110-120.

    ARMBRECHT, H.J., BIRNBAUM, L.S., ZENSER, T.V., MATTAMMAL, M.B., &
    DAVIS, B.B. (1979) Renal cytochrome P-450's - Electrophoretic and
    electron paramagnetic resonance studies. Arch. Biochem. Biophys., 197:
    277-284.

    ASKERGREN, A. (1981) Studies on kidney function in subjects exposed to
    organic solvents. III. Excretion of cells in the urine. Acta med.
    Scand., 210: 103-106.

    ASKERGREN, A., ALLGEN, L.G., KARLSSON, C., LUNDBERG, I., & NYBERG, E.
    (1981) Studies on kidney function in subjects exposed to organic
    solvents. I. Excretion of albumin and ß2-microglobulin in the urine.
    Acta med. Scand., 209: 479-483.

    ATTALLAH, A. & STAHL, R. (1980) Inhibition of renal PGE2
    biosynthesis  in vivo: regional differences. Prostaglandins, 19:
    649-650.

    AUKLAND, K. (1980) Methods for measuring renal blood flow: Total flow
    and regional distribution. Annu. Rev. Physiol., 42: 543-555.

    AUTOR, A.P., ed. (1977) Biochemical mechanisms of paraquat toxicity,
    New York, London, San Francisco, Academic Press.

    AVENDANO, L.H. & LOPEZ-NOVOA, J.M. (1987) Glomerular filtration and
    renal blood flow in the aged. In: Macias-Nunez, I.F. & Cameron, J.S.,
    ed. Renal function and diseases in the elderly, London, Boston,
    Toronto, Butterworth, pp. 27-48.

    BACH, P.H. (1989) The detection of chemically induced renal injury,
    the cascade of degenerative morphological and functional changes that
    follow the primary nephrotoxic insult and the evaluation of these
    changes by  in vitro methods. Toxicol. Lett., 46: 237-250.

    BACH, P.H. & BRIDGES, J.W. (1984) The role of prostaglandin synthase
    mediated metabolic activation of analgesics and non-steroidal
    anti-inflammatory drugs in the development of renal papillary necrosis
    and upper urothelial carcinoma. Prostaglandins Leukotrienes Med., 15:
    251-274.

    BACH, P.H. & BRIDGES, J.W. (1985a) Chemically induced renal papillary
    necrosis and upper urothelial carcinoma. CRC crit. Rev. Toxicol., 15:
    217-439.

    BACH, P.H. & BRIDGES, J.W. (1985b) A decision tree approach for the
    application of drug metabolism and kinetics to  in vivo and  in vitro
    toxicological and pharmacological testing. Arch. Toxicol., Suppl. 8:
    173-188.

    BACH, P.H. & GREGG, N. (1988) Experimentally induced renal papillary
    necrosis and upper urothelial carcinoma. Inter. Rev. exp. Pathol., 30:
    1-54.

    BACH, P.H. & HARDY, T.L. (1985) The relevance of animal models to the
    study of analgesic associated renal papillary necrosis in man. Kidney
    Int., 28: 605-613.

    BACH, P.H. & KWIZERA, E.N. (1988) Nephrotoxicity: A rational approach
    to  in vitro target cell injury in the kidney. Xenobiotica, 16:
    685-698.

    BACH, P.H. & LOCK, E.A. (1982) The use of renal tissue slices,
    perfusion and infusion techniques to assess renal function and
    malfunction. In: Bach, P.H., Bonner, F.W., Bridges, J.W., & Lock,
    E.A., ed. Nephrotoxicity: Assessment and pathogenesis, New York,
    Chichester, Brisbane, Toronto, John Wiley and Sons, pp. 128-143.

    BACH, P.H. & LOCK, E.A., ed. (1985) Renal heterogeneity and target
    cell toxicity, New York, Chichester, Brisbane, Toronto, John Wiley and
    Sons.

    BACH, P.H. & LOCK, E.A., ed. (1987) Nephrotoxicity in the experimental
    and the clinical situation, Dordrecht, Boston, Lancaster, Martinus
    Nijhoff Publishers.

    BACH, P.H. & LOCK, E.A., ed. (1989) Nephrotoxicity: Extrapolation from
     in vitro to  in vivo, and animals to man, New York, London, Plenum
    Press.

    BACH, P.H., BONNER, F.W., BRIDGES, J.W., & LOCK, E.A., ed. (1982)
    Nephrotoxicity: Assessment and pathogenesis, New York, Chichester,
    Brisbane, Toronto, John Wiley and Sons.

    BACH, P.H., GRASSO, P., MOLLAND, E.A., & BRIDGES, J.W. (1983) Changes
    in the medullary glycosaminoglycan histochemistry and microvascular
    filling during the development of 2-bromoethanamine
    hydrobromide-induced renal papillary necrosis. Toxicol. appl.
    Pharmacol., 69: 333-344.

    BACH, P.H., KETLEY, C.P., BENNS, S.E., AHMED, I., & DIXIT., M. (1985)
    The use of isolated and cultured renal cells in nephrotoxicity -
    practice, potential and problems. In: Bach, P.H. & Lock, E.A., ed.
    Renal heterogeneity and target cell toxicity, New York, Chichester,
    Brisbane, Toronto, John Wiley and Sons, pp. 505-518.

    BACH, P.H., KETLEY, C.P., DIXIT, M., & AHMED, I. (1986) The mechanisms
    of target cell injury in nephrotoxicity. Food chem. Toxicol., 24:
    775-779.

    BACH, P.H., GREGG, N.J., & WACHSMUTH, E.D. (1987) The application of
    histochemistry at the light microscopic level to the study of
    nephrotoxicity. In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity in the
    experimental and the clinical situation, Dordrecht, Boston, Lancaster,
    Martinus Nijhoff Publishers, pp. 19-84.

    BACH, P.H., BERLIN, A., HESELTINE, E., KRUG, E., LAUWERYS, R., SMITH,
    E., & VAN DER VENNE, M.-T., ed. (1989) Proceedings of the
    International Workshop on the Health Significance of Nephrotoxicity.
    Toxicol. Lett., 46: 1-306.

    BACHUR, N.R., GORDON, S.L., GEE, M.V., & KON, H. (1979) NADPH
    cytochrome P-450 reductase activation of quinone anticancer agents to
    free radicals. Proc. Natl Acad. Sci., 76: 954-957.

    BAER, P.G. (1981) The contribution of prostaglandins to renal blood
    flow maintenance is determined by the level of activity of the
    renin-angiotensin system. Life Sci., 28: 587-593.

    BAGGETT, J.M. & BERNDT, W.O. (1984a) Renal and hepatic glutathione
    concentrations in rats after treatment with hexachloro-1,3-butadiene
    and citrinin. Arch. Toxicol., 56: 46-49.

    BAGGETT, J.M. & BERNDT, W.O. (1984b) Interaction of potassium
    dichromate with the nephrotoxins, mercuric chloride and citrinin.
    Toxicology, 33: 157-169.

    BAGGETT, J.M. & BERNDT, W.O. (1985) The effect of potassium dichromate
    on the urinary excretion, organ and subcellular distribution of
    203Hg-mercuric chloride. Toxicol. Lett., 29: 115-121.

    BAILIE, M.B., SMITH, J.H., NEWTON, J.F., & HOOK, J.B. (1984) Mechanism
    of chloroform nephrotoxicity. IV. Phenobarbital potentiation of in
    vitro chloroform metabolism and toxicity in rabbit kidneys. Toxicol.
    appl. Pharmacol., 74: 285-292.

    BAKER, E.L., GOYER, R.A., FOWLER, B.A., KHETTERY, U., BERNARD, D.B.,
    ADLER, S., DEVERE WHITE, R., BABAYAN, R., & FELDMAN, R.G. (1980)
    Occupational lead exposure, nephropathy, and renal cancer. Am. J. ind.
    Med., 1: 139-148.

    BANK, N., MUTZ, E.F., & AYNEDIJIAN, H.S. (1967) The role of "leakage"
    of tubular fluid in anuria due to mercury poisoning. J. clin. Invest.,
    46: 695-704.

    BANKI, K., ELFARRA, A.A., LASH, L.H., & ANDERS, M.W. (1986) Metabolism
    of S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine to hydrogen sulfide
    and the role of hydrogen sulfide in
    S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine-induced mitochondrial
    toxicity. Biochem. biophys. Res. Commun., 138: 707-713.

    BANNASCH, P. & ZERBAN, H. (1986) Renal cell adenoma and carcinoma. In:
    Jones, T.C., Mohr, U., Hunt, R.D., ed. Monographs on pathology of
    laboratory animals, urinary system, Berlin, Heidelberg, New York,
    Springer-Verlag, pp. 112-139.

    BANNASCH, P., KRECH, R., & ZERBAN, H. (1978a) [Morphogenesis and
    micro-morphology of epithelial kidney tumours in rats poisoned with
    nitrosomorpho-line. II. Tubular glycogenesis and the formation of
    clear-cell or acidophilic-cell tumours.] Z. Krebsforsch., 92: 63-86
    (in German).

    BANNASCH, P., KRECH, R., & ZERBAN, H. (1978b) [Morphogenesis and
    micro-morphology of epithelial kidney tumours in rats poisoned with
    nitrosomorpho-line. III. Oncocyte tubules and oncocytomas.] Z.
    Krebsforsch., 92: 87-104 (in German).

    BANNASCH, P., KRECH, R., & ZERBAN, H. (1980) [Morphogenesis and
    micro-morphology of epithelial kidney tumours in rats poisoned with
    nitrosomorpho-line. IV. Tubular lesions and basophilic tumours.] J.
    Cancer Res. clin. Oncol., 98: 243-265 (in German).

    BANNASCH, P., BENNER, U., HACKER, H.J., KLIMEK, F., MAYER, D., MOORE,
    M., & ZERBAN, H. (1981) Cytochemical and biochemical analysis of
    carcinogenesis. Histochem. J., 13: 799-820.

    BARRATT, M. (1983) Proteinuria. Br. med. J., 287: 1489-1490.

    BARRY, D.N. & BOWNESS, J.M. (1975) Identification and turnover of
    glyco-saminoglycans in rat kidneys. Can. J. Biochem., 53: 713-720.

    BATELLE, D., GAVIRIA, M., GRUPP, M., ARRUDA, J.A., WYNN, J., &
    KURTZMAN, N.A. (1982) Distal nephron function in patients receiving
    lithium therapy. Kidney Int., 21: 477-486.

    BATUMAN, V., MAESAKA, J.K., HADDAD, B., TEPPER, E., LANDY, E., &
    WEDEEN, R.P. (1981) The role of lead in gout nephropathy. New Engl. J.
    Med., 304: 520-523.

    BATUMAN, V., LANDY, E., MAESAKA, J.K., & WEDEEN, R.P. (1983)
    Contribution of lead to hypertension with renal impairment. New Engl.
    J. Med., 309: 17-21.

    BAVEREL, G., BONNARD, M., D'ARMAGNAC DE CASTANET, E., & PELLET, M.
    (1978) Lactate and pyruvate metabolism in isolated renal tubules of
    normal dogs. Kidney Int., 14: 567-575.

    BAVEREL, G., BONNARD, M., & PELLET, M. (1979) Lactate and pyruvate
    metabolism in isolated human kidney tubules. FEBS Lett., 101: 282-286.

    BAVEREL, G., FORISSIER, M., & PELLET, M. (1980a) Lactate and pyruvate
    metabolism in dog renal outer medulla. Effects of oleate and ketone
    bodies. Int. J. Biochem., 12: 163-168.

    BAVEREL, G., GENOUX, C., FORISSIER, M., & PELLET, M. (1980b) Fate of
    glutamate carbon and nitrogen in isolated guinea-pig kidney cortex
    tubules. Biochem. J., 188: 873-880.

    BECKER, C.G., BECKER, E.L., MAHER, J.F., & SCHREINER, G.E. (1962)
    Nephrotic syndrome after contact with mercury. A report of five cases,
    three after use of ammoniated mercury. Arch. intern. Med., 110:
    178-186.

    BEEUWKES, R. (1980) The vascular organisation of the kidney. Annu.
    Rev. Physiol., 42: 531-542.

    BEKERSKY, I., COLBURN, W.A., FISHMAN, L., & KAPLAN, S.A. (1980)
    Metabolism of salicylic acid in the isolated perfused rat kidney.
    Intercon-version of salicyluric and salicylic acids. Drug Metab.
    Disp., 8: 319-324.

    BELLEMAN, P. (1980) Primary monolayer cultures of liver parenchymal
    cells and kidney tubules as a useful new model for biochemical
    pharmacology and experimental toxicology. Studies  in vitro on
    hepatic membrane transport, induction of liver enzymes and adaptive
    changes in renal cortical enzymes. Arch. Toxicol., 44: 63-84.

    BENDELE, A.M., CARLTON, W.W., KROGH, P., & LILLEHOJ, E.B. (1985)
    Ochratoxin A carcinogenesis in the (C57BL/6J x C3H)F1 mouse. J. Natl
    Cancer Inst., 75: 733-742.

    BENDZ, H. (1983) Kidney function in lithium-treated patients: a
    literature survey. Acta psychiatr. Scand., 68: 303-24.

    BENGTSSON, U. (1962) A comparative study of chronic non-obstructive
    pyelonephritis and renal papillary necrosis. Acta med. Scand.,
    388(Suppl.): 1-71.

    BENGTSSON, U., JOHANSSON, S., & ANGERVALL, L. (1978) Malignancies of
    the urinary tract and their relation to analgesic abuse. Kidney Int.,
    13: 107-113.

    BENNETT, W.M. (1983) Aminoglycoside nephrotoxicity. Nephron, 35:
    73-77.

    BENNETT, W.M. (1985) Lead nephropathy. Kidney Int., 28: 212-220.

    BENNETT, W.M. (1986) Drugs and renal disease, New York, Churchill
    Livingstone.

    BENNETT, W.M., HOUGHTON, D.C., MCCARRON, D.A., ELLIOTT, W.C., PORTER,
    G.A., & GILBERT, D.N. (1983) Interventions to modify
    aminoglycoside-induced acute renal failure. In: Solez, K. & Whelton,
    A., ed. Acute renal failure: Correlations between morphology and
    function, New York, Basel, Marcel Dekker, pp. 331-358.

    BENNETT, W.M., MELA-RIKER, L., HOUGHTON, D.C., GILBERT, D.N., & BUSS,
    W.C. (1988) Microsomal protein synthesis inhibition: an early
    mani-festation of gentamicin nephrotoxicity. Am. J. Physiol., 24:
    F265-F269.

    BENNINGTON, J.L. (1973) Cancer of the kidney - etiology, epidemiology,
    and pathology. Cancer, 32: 1017-1029.

    BERGERON, M.G., TROTTIER, S., LESSARD, C., BEAUCHAMP, D., & GAGNON, P.
    (1982) Disturbed intrarenal distribution of gentamicin in experimental
    pyelonephritis due to  Escherichia coli. J. infect. Dis., 146:
    436-439.

    BERLYNE, G.M. (1984) Toxic nephropathies and current methods for early
    detection of the toxicity of the kidney. In: Mehlman, M.A., Hemstreet,
    C.P., Thorpe, J.J., & Weaver, N.K., ed. Renal effects of petroleum
    hydrocarbons, Princeton, New Jersey, Princeton Scientific Publishers,
    pp. 173-184 (Advances in Modern Environmental Toxicology, Vol. VII).

    BERNARD, A. & LAUWERYS, R. (1986) Effects of cadmium exposure in man.
    In: Foulkes, E.C., ed. Cadmium toxicology. Handbook of experimental
    pharmacology, Berlin, Heidelberg, New York, Springer-Verlag, Vol. 80,
    pp. 136-177.

    BERNARD, A., ROELS, H., HUBERMONT, G., BUCHET, J.P., MASSON, P.L., &
    LAUWERYS, R. (1976) Characterization of the proteinuria in cadmium
    exposed workers. Int. Arch. occup. environ. Health, 38: 19-30.

    BERNARD, A., BUCHET, J.P., ROELS, H., MASSON, P., & LAUWERYS, R.
    (1979a) Renal excretion of proteins and enzymes in workers exposed to
    cadmium. Eur. J. clin. Invest., 9: 11-22.

    BERNARD, A., GORET, A., BUCHET, J.P., ROELS, H., & LAUWERYS, R.
    (1979b) Comparison of sodium dodecyl sulfate polyacrylamide gel
    electro-phoresis with quantitative methods for the analysis of
    cadmium-induced proteinuria. Int. Arch. occup. environ. Health, 44:
    139-148.

    BERNARD, A., MOREAU, D., & LAUWERYS, R. (1982) Comparison of
    retinol-binding protein and beta2-microglobulin determination in urine
    for the early detection of tubular proteinuria. Clin. Chim. Acta, 126:
    1-7.

    BERNARD, A., OULED AMOR, A., & LAUWERYS, R. (1987) The effects of low
    doses of cadmium-metallothionein on the renal uptake of
    beta2-microglobulin in rats. Toxicol. appl. Pharmacol., 15: 440-445.

    BERNARD, A., OULED AMOR, A., VIAU, C., & LAUWERYS, R. (1988) The renal
    uptake of proteins: a non selective process in conscious rats. Kidney
    Int., 34: 175-185.

    BERNARD, A., LAUWERYS, R., NOEL, A., VANDELEENE, B., & LAMBERT, A.
    (1989) Urine protein 1: a sex dependent marker of tubular or
    glomerular dysfunction. Clin. Chem., 35: 2141-2142.

    BERNDT, W.O. (1976) Use of renal slice technique for evaluation of
    renal tubular transport process. Environ Health Perspect., 15: 73-88.

    BERNDT, W.O. (1981) Use of renal function tests in the evaluation of
    nephrotoxic effects. In: Hook, J.B., ed. Toxicology of the kidney, New
    York, Raven Press, pp. 1-29 (Target Organ Toxicology Series).

    BERNDT, W.O. (1987) Renal slices and perfusion. In: Bach, P.H. & Lock,
    E.A., ed. Nephrotoxicity in the experimental and the clinical
    situation, Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers,
    pp. 301-316.

    BERNDT, W.O. (1989) Potential involvement of renal transport
    mechanisms in nephrotoxicity. Toxicol. Lett., 46: 77-82.

    BERNDT, W.O. & HAYES, A.W. (1977) Effects of citrinin on renal
    transport processes. J. environ. Pathol. Toxicol., 1: 93-103.

    BERTANI, T., POGGI, A., POZZONI, R., DELAINI, F., SACCHI, G., THOUA,
    Y., MECCA, G., REMUZZI, G., & DONATI, M.D. (1982) Adriamycin-induced
    nephrotic syndrome in rats: sequence of pathologic events. Lab.
    Invest., 46: 16-23.

    BERTANI, T., ROCCHI, G., SACCHI, G., MECCA, G., & REMUZZI, G. (1986)
    Adriamycin-induced glomerulosclerosis in the rat. Am. J. Kidney Dis.,
    7: 12-19.

    BHATTACHARYA, R.K. & SCHULTZE, M.O. (1967) Enzymes from bovine and
    turkey kidneys which cleave S-(1,2-dichlorovinyl)-L-cysteine. Comp.
    Biochem. Physiol., 22: 723-735.

    BIBER, T.U.L., MYLLE, M., BAINES, A.D., & GOTTSCHALK, C.W. (1968) A
    study by micropuncture and microdissection of acute renal failure in
    rats. Am. J. Med., 44: 664-705.

    BLAIR-WEST, J.R., BOBIK, A., BROOK, A.H., ESLER, M.D., GIBSON, A.,
    MORRIS, M., MCKINLEY, M.J., & PULLAN, P.T. (1980) Renin, ADH and the
    kidney: a congeries of conundrums. Prog. biochem. Pharmacol., 17:
    20-28.

    BLUM, R.H. & CARTER, S.K. (1974) Adriamycin: a new anticancer drug
    with significant clinical activity. Ann. intern. Med., 80: 249-259.

    BOHMAN, S.-O. (1980) The ultrastructure of the renal medulla and the
    interstitial cells. In: Mandal, A.K. & Bohman, S.-O., ed. The renal
    papilla and hypertension, New York, London, Plenum Press, pp. 7-33.

    BOJESEN, I. (1974) Quantitative and qualitative analyses of isolated
    lipid droplets from interstitial cells in renal papillae from various
    species. Lipids, 9: 835-843.

    BOMHARD, E., LUCKHAUS, G., VOIGT, W.-H., & LOESER, E. (1988) Induction
    of light hydrocarbons nephropathy by p-dichlorobenzene. Arch.
    Toxicol., 61: 433-439.

    BONER, G., SHERRY, J., & RIESELBACH, R.E. (1972) Hypertrophy of the
    normal human kidney following contralateral nephrectomy. Nephron, 9:
    364-372.

    BRENNER, B.M. (1983) Hemodynamically mediated glomerular injury and
    the progressive nature of kidney disease. Kidney Int., 23: 647-655.

    BRENNER, B.M. & RECTOR, F.C., ed. (1986) The kidney, Philadelphia,
    W.B. Saunders.

    BRENNER, B.M., HOSTETTER, T.H., & HUMAS, H.D. (1978) Glomerular
    permselectivity: barrier function based on discrimination of molecular
    size and charge. Am. J. Physiol., 234: F455-F460.

    BRENNER, B.M., MEYER, T.W., & HOSTETTER, T.H. (1982) Dietary protein
    intake and the progressive nature of kidney disease: the role of
    hemodynami-cally mediated glomerular injury in the pathogenesis of
    progressive glomerular sclerosis in aging, renal ablation and
    intrinsic renal disease. New Engl. J. Med., 307: 652-659.

    BREZIS, M., ROSEN, D., SILVA, P., & EPSTEIN, F.H. (1984) Renal
    ischaemia: a new perspective. Kidney Int., 26: 375-383.

    BRIDGES, J.W., BACH, P.H., BONNER, F.W., & BEATON, E.M. (1982) Effects
    of nutritional factors on renal response to toxins. In: Bach, P.H.,
    Bonner, F.W., Bridges, J.W., & Lock, E.A., ed. Nephrotoxicity:
    Assessment and pathogenesis, New York, Chichester, Brisbane, Toronto,
    John Wiley and Sons, pp. 182-198.

    BROWN, W.W., ZENSER, T.V., & DAVIS, B.B. (1980) Prostaglandin E2
    production by rabbit urinary bladder. Am. J. Physiol., 239: F452-F458.

    BRUNBORG, G., HOLME, J.A., SODERLUND, E.J., OMICHINSKI, J.G., &
    DYBING, E. (1988) An automated alkaline elution system: DNA damage
    induced by 1,2-dibromo-3-chloropropane  in vivo and  in vitro. Anal.
    Biochem., 174: 522-536.

    BUCHET, J.-P., ROELS, H., BERNARD, A., & LAUWERYS, R. (1980)
    Assessment of renal function of workers simultaneously exposed to
    inorganic lead, cadmium or mercury vapour. J. occup. Med., 22:
    741-750.

    BUCKLEY, J.E., CLARK, V.L., MEYER, T.H., & PEARLMAN, N.W. (1984)
    Hypomagnesemia after cisplatin combination chemotherapy. Arch. intern.
    Med., 144: 2347-2348.

    BURKE, J.F., LAUCIUS, J.F., BRODOVSKY, H.S., & SORIANO, R.Z. (1977)
    Doxorubicin hydrochloride-associated renal failure. Arch. intern.
    Med., 137: 385-388.

    BURKHOLDER, P.M. (1982) Functions and pathophysiology of the
    glomerular mesangium (editorial). Lab. Invest., 46: 239-241.

    BURRY, A., CROSS, R., & AXELSEN, R. (1977) Analgesic nephropathy and
    the renal concentrating mechanism. Pathol. Annu., 12: 1-31.

    BURTON, B.T. & HIRSCHMAN, G.H. (1979) Demographic analysis: end-stage
    renal disease and its treatment in the United States. Clin. Nephrol.,
    11: 47-51.

    BUTLER, E.A. & FLYNN, F.V. (1958) The proteinuria of renal tubular
    disorders. Lancet, 2: 978-980.

    BUTLER, W.H. (1964) Acute toxicity of aflatoxin B1 in rats. Br. J.
    Cancer, 18: 756-762.

    BUTLER, W.H. & LIJINSKY, W. (1970) Acute toxicity of aflatoxin G1 to
    the rat. J. Pathol., 102: 209-212.

    BUTLER, W.H., GREENBLATT, M., & LIJINSKY, W. (1969) Carcinogenesis in
    rats by aflatoxins B1, G1, and B2. Cancer Res., 29: 2206-2211.

    BUTLER, W.T. (1966) Pharmacology, toxicity, and therapeutic usefulness
    of amphotericin B. J. Am. Med. Assoc, 195: 371-375.

    BUZIO, C., MUTTI, A., CAPANI, F., ANDRULLI, S., PERAZZOLI, F.,
    ALINOVI, R., NEGRO, A., & RUSTICHELLI, R. (1989) Circadian rhythm of
    proteinuria: effects of an evening meat meal. Nephrol. Dialysis
    Transplant., 4: 266-270.

    CALDER, I.C., YONG, A.C., WOODS, R.A., CROWN, C.A., HAM, K.N., &
    TANGE, J.D. (1979) The nephrotoxicity of p-aminophenol. II. The effect
    of metabolic inhibitors and inducers. Chem.-biol. Interact., 27:
    245-254.

    CALDICOTT, W.J.H., HOLLENBERG, N.K., & ABRAMS, H.L. (1970)
    Characteristics of response of renal vascular bed to contrast media.
    Evidence of vasoconstriction induced by renin-angiotensin system.
    Invest. Radiol., 5: 539-547.

    CALNE, R.Y., WHITE, D.J.G., THIRU, S., EVANS, D.B., MCMASTER, P.,
    DUNN, D.C., CRADDOCK, G.N., PENTLOW, B.D., & ROLLES, K. (1978)
    Cyclosporin-A in patients receiving renal allografts from cadaver
    donors. Lancet, 2: 1323-1327.

    CARLIER, M.-B., LAURENT, G., CLAES, P.J., VANDERHAEGHE, H.J., &
    TULKENS, P.M. (1983) Inhibition of lysosomal phospholipases by
    aminoglycoside antibiotics:  in vitro comparative studies.
    Antimicrob. Agents Chemother., 23: 440-449.

    CARPENTER, H.M. & MUDGE, G.H. (1981) Acetaminophen nephrotoxicity.
    Studies on renal acetylation and deacetylation. J. Pharmacol. exp.
    Ther., 218: 161-167.

    CASTEGNARO, M. & CHERNOZEMSKY, I. (1987) Endemic nephropathy and
    urinary tract tumors in the Balkans. Cancer Res., 47: 3608-3609.

    CASTOR, C.W. & GREEN, J.A. (1968) Regional distribution of acid
    mucopolysaccharides in the kidney. J. clin. Invest., 47: 2125-2132.

    CAULFIELD, J.P., REID, J.J., & FARQUBAR, M.G. (1976) Alterations of
    the glomerular epithelium in aminonucleoside nephrosis. Lab. Invest.,
    34: 43-59.

    CAVALIERE,  G., ARRIGO, G.,  D'AMICO, G., BERNASCONI,  P., SCHIAVINA,
    G., DELLAFIORE, L., & VERGNAGHI, D. (1987) Tubular nephro-toxicity
    after intravenous urography with ionic high-osmolal and nonionic
    low-osmolal contrast media in patients with chronic renal
    insufficiency. Nephron, 46: 128-133.

    CEDGARD, S., HERLITZ, H., GETERUD, K., ALTMAN, P.O., & AURELL, M.
    (1986) Acute renal insufficiency after administration of low-osmolar
    contrast media. Lancet, 2: 1281.

    CHAHWALA, S.B. & HARPUR, E.S. (1986) The use of renal tubule fragments
    isolated from the rat to investigate aspects of gentamicin
    nephrotoxicity. J. Pharmacol. Methods, 15: 21-34.

    CHAMBERLIN, M.E., LEFURGEY, A., & MANDEL, L.J. (1984) Suspension of
    medullary thick ascending limb tubules from the rabbit kidney. Am. J.
    Physiol., 247: F955-F964.

    CHARBONNEAU, M., LOCK, E.A., STRASSER, J., COX, M.G., TURNER, M.J., &
    BUS, J.S. (1987) 2,2,4-Trimethylpentane-induced nephrotoxicity. I.
    Metabolic disposition of TMP in male and female Fischer-344 rats.
    Toxicol. appl. Pharmacol., 91: 171-181.

    CHASSEAUD, L.F. (1980) Extrahepatic conjugation with glutathione. In:
    Gram, T.E., ed. Extrahepatic metabolism of drugs and other foreign
    compounds, New York, Spectrum Publications, pp. 427-443.

    CHOPRA, S., KAUFMAN, J.S., JONES, T.W., HONG, W.K., GEHR, M.K.,
    HAMBERGER, R.J., FLAMENBAUM, W., & TRUMP, B.F.  (1982)
    Cis-diamminedichloro-platinum-induced acute renal failure in the rat.
    Kidney. Int., 21: 54-64.

    CHOU, C.C., HOOK, J.B., & HSIEH, C.P. (1974) Effects of radiopaque
    dyes on renal vascular resistance. J. lab. clin. Med., 78: 705-712.

    CHUANG, E.L., REINECK, H.J., OSGOOD, R.W., LIMAI, R.T., & STEIN, J.H.
    (1978) Studies on the mechanism of reduced urinary osmolality after
    exposure of renal papilla. J. clin. Invest., 61: 633-639.

    CHUMAN, L., FINE, L.G., COHEN, A.H., & SAIER, M.H. Jr. (1982)
    Continuous growth of proximal tubular kidney epithelial cells in
    hormone-supplemented serum-free medium. J. Cell Biol., 94: 506-510.

    CLIFTON, G., KAPLOWITZ, N., WALLIN, J.D., & KUHLENKAMP, J. (1975) Drug
    induction and sex differences of renal glutathione S-transferases in
    the rat. Biochem. J., 150: 259-262.

    CLIVE, D.M. & STOFF, J.S. (1984) Renal syndrome associated with
    nonsteroidal anti-inflammatory drugs. New Engl. J. Med., 310: 563-572.

    COHEN, J.J. (1979) Is the function of the renal papilla coupled
    exclusively to an anaerobic pattern of metabolism? Am. J. Physiol.,
    236: F423-F433.

    COHEN, J.J., HARRINGTON, J.T., & KASSIREV, J.P. (1981) Lithium and the
    kidney. Kidney Int., 19: 374-387.

    COJOCEL, C., MAITA, K., PASINO, D.A., KUO, C., & HOOK, J.B. (1983)
    Metabolic heterogeneity of the proximal and distal kidney tubules.
    Life Sci., 33: 855-861.

    COLLEONI, N. & D'AMICO, G. (1986) Chronic lead accumulation as a
    possible cause of renal failure in gouty patients. Nephron, 44: 32-35.

    COLLIER, V., LIETMAN, P.S., & MITCH, W.E. (1979) Evidence for luminal
    uptake of gentamicin in the perfused rat kidney. J. Pharmacol. exp.
    Ther., 210: 247-251.

    CONNELLY, J.C. & BRIDGES, J.W. (1980) The distribution and role of
    cytochrome P-450 in extrahepatic organs. Prog. Drug. Metab., 5: 1-111.

    CONSTANTOPOULOS, G., LOUIE, M., & DEKABAN, A.S. (1973) Acid
    mucopolysaccharides (glycosaminoglycans) in normal human kidneys and
    in kidneys of patients with mucopolysaccaridoses. Biochem. Med., 7:
    376-388.

    COOPER, W.C. & GAFFEY, W.R. (1975) Mortality of lead workers. J.
    occup. Med., 17: 100-109.

    COOPER, W.C., WONG, O., & KHEIFETS, L. (1985) Mortality among
    employees of lead battery plants and lead producing plants. Scand. J.
    Work Environ. Health., 11: 331-345.

    CRAIN, R.C. (1958) Spontaneous tumors in the Rochester strain of the
    Wistar rat. Am. J. Pathol., 34: 311-336.

    CRAMER, K., GOYER, R.A., JAGENBURG, R., & WILSON, M. (1974) Renal
    ultrastructure, renal function and parameters of lead toxicity in
    workers with different periods of lead exposure. Br. J. ind. Med., 31:
    113-127.

    CROWE, C.A., CALDER, I.C., MADSEN, N.P., FUNDER, C.C., GREEN, C.R.,
    HAM, K.N., & TANGE, J.D. (1977) An experimental model of
    analgesic-induced renal damage. Some effects of p-aminophenol on rat
    kidney mitochondria. Xenobiotica, 7: 345-356.

    CROWE, C.A., YONG, A.C., CALDER, I.C., HAM, K.N., & TANGE, J.D. (1979)
    The nephrotoxicity of p-aminophenol. I. The effect of microsomal
    cyto-chromes, glutathione and covalent binding in kidney and liver.
    Chem.-biol. Interact., 27: 235-243.

    CUNARRO, J.A. & WEINER, M.W. (1978) Effects of ethacrynic acid and
    furosemide on respiration of isolated kidney tubules: the role of ion
    transport and the source of metabolic energy. J. Pharmacol. exp.
    Ther., 206: 198-206.

    CUNNINGHAM, E.E., BRENTJENS, J.R., ZIELEZREY, M.A., ANDRES, G.A., &
    VERUTO, R.C. (1980) Heroin nephropathy: a clinicopathologic and
    epidemiologic study. Am. J. Med., 68: 47-53.

    CUNNINGHAM, E.E., BARONE, P.P., NASCIMENTO, L., & VENUTO, R.C. (1986)
    Effect of a radiographic contrast agent on renal function in the rat.
    Comparison with equiosmolar mannitol. Miner. electrolyte Metab., 12:
    157-160.

    CURTHOYS, N.P. & BELLEMANN, P. (1979) Renal cortical cells in primary
    monolayer culture. Enzymatic changes and morphological observations.
    Exp. cell Res., 121: 31-45.

    CURTIS, J.R. (1979) Drug-induced renal disease. Drugs, 18: 377-391.

    CVITKOVIC, E., SPAULDING, J., BETHUNE, V., MARTIN, J., & WHITMORE,
    W.F. (1977) Improvement of cis-dichlorodiammineplatinum (NSC-119875):
    Therapeutic index in an animal model. Cancer, 39: 1357-1361.

    DARNTON, S.J. (1967) Glycogen metabolism in rabbit kidney under
    differing physiological states. Q. J. exp. Physiol., 52: 392-400.

    DARNTON, S.J. (1969) The conversion of injected glucose into renal
    glycogen and mucopolysaccharides. An autoradiographic study of rabbits
    in various states of hydration. Z. Zellforsch., 102: 273-282.

    DAUGAARD, G., ROSSING, N., & RORTH, M. (1988a) Effects of cisplatin on
    different measures of glomerular function in the human kidney with
    special emphasis on high-dose. Cancer Chemother. Pharmacol., 21:
    163-167.

    DAUGAARD, G., HOLSTEIN-RATHLOU, N.-H., & LEYSSAC, P.P. (1988b) Effects
    of cisplatin on proximal convoluted and straight segments of the rat
    kidney. J. Pharmacol. exp. Ther., 244: 1081-1085.

    DAVIES, D.F. & SHOCK, N.W. (1950) Age changes in glomerular filtration
    rate, effective renal plasma flow and tubular excretory capacity in
    adult males. J. clin. Invest., 29: 496-507.

    DAVIES, J.M. (1984) Long term mortality study of chromate pigment
    workers who suffered lead poisoning. Br. J. ind. Med., 41: 170-178.

    DAVIS, B.B., MATTAMMAL, M.B., & ZENSER, T.V. (1981) Renal metabolism
    of drugs and xenobiotics. Nephron, 27: 187-196.

    DAYAL, H. & KINMAN, J. (1983) Epidemiology of kidney cancer. Semin.
    Oncol., 10: 366-377.

    DEBROE, M.E., PAULUS, G.J., VERPOOTEN, G.A., ROELS, F., BUYSSENS, N.,
    WEDEEN, R., VANHOOF, F., & TULKENS, P.M. (1984) Early effects of
    gentamicin, tobramycin and amikacin on the human kidney. Kidney Int.,
    25: 643-652.

    DEES, J.H., HEATFIELD, B.M., REUBER, M.D., & TRUMP, B.F. (1980a)
    Adenocarcinoma of the kidney. III. Histogenesis of renal
    adenocarcinomas induced in rats by
    N-(4'-fluoro-4-biphenylyl)acetamide, J. Natl Cancer Inst., 64:
    1537-1545.

    DEES, J.H., HEATFIELD, B.M., & TRUMP, B.F. (1980b) Adenocarcinoma of
    the kidney. IV. Electron microscopic study of the development of renal
    adenocar-cinomas induced in rats by
    N-(4'-fluoro-4-biphenylyl)acetamide. J. Natl Cancer Inst., 64:
    1547-1562.

    DEKERNION, J.B. & BERRY, D. (1980) The diagnosis and treatment of
    renal carcinoma. Cancer, 45(Suppl.): 1947-1982.

    D'ELIA, J.A., GLEASON, K.E., ALDAY, M., MALARICK, C., GODLEY, K.,
    WARRAM, J., KALDANY, A., & WEINRAUCH, L.A. (1982) Nephrotoxicity from
    angiographic contrast material. A prospective study. Am. J. Med., 72:
    719-725.

    DENTINO, M., LUFT, F.C., YUM, M.W., WILLIAMS, S.D., & EINHORN, L.H.
    (1978) Long term effect of cis-diamminedichloride platinum (CDDP) on
    renal function and structure in man. Cancer, 41: 1274-1281.

    DEPNER, T.A. (1982) Nephrotic syndrome secondary to lithium therapy.
    Nephron, 30: 286-289.

    DICKER, S.E. & FRANKLIN, C.S. (1966) The isolation of hyaluronic acid
    and chondroitin sulphate from kidneys and their reaction with urinary
    hyaluronidase. J. Physiol., 186: 110-120.

    DIEPERINK, H.H. (1989) Identification of groups at risk for renal
    diseases (including nephrotoxicity). Toxicol. Lett., 46: 257-268.

    DIEPERINK, H., STARKLINT, H., & LEYSSAC, P.P. (1983) Nephrotoxicity of
    cyclosporin A - an animal model. A study of the nephrotoxic effect of
    cyclosporin on overall renal and tubular function in conscious rats.
    Transplant Proc., 15(Suppl. 1): 2736-2741.

    DIEPERINK, H., LEYSSAC, P.P., STARKLINT, H., & KEMP, E. (1985)
    Glomerulotubular function in cyclosporin A treated rats. A lithium
    clearance, occlusion time/transit time and micropuncture study. Proc.
    Eur. Dialysis Transplant Assoc., 21: 853-859.

    DIEZI, J. & ROCH-RAMEL, F. (1987) The use of single nephron techniques
    in renal toxicity studies. In: Bach, P.H. & Lock, E.A., ed.
    Nephrotoxicity in the experimental and the clinical situation,
    Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers, pp.
    317-358.

    DOBYAN, D.C. (1985) Long-term consequences of cis-platinum-induced
    renal injury: a structural and functional study. Anat. Rec., 212:
    239-245.

    DOHN, D.R., QUEBBEMANN, A.J., BORCH, R.F., & ANDERS, M.W. (1985a)
    Enzymatic reaction of chlorotrifluoroethene with glutathione: 19F NMR
    evidence for stereochemical control of the reaction. Biochemistry, 24:
    5137-5143.

    DOHN, D.R., LEININGER, J.R., LASH, L.H., QUEBBEMANN, A.J., & ANDERS,
    M.W. (1985b) S-(2-chloro-1,1,2-trifluoroethyl)glutathione and
    S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine, the glutathione and
    cysteine conju-gates of chlorotrifluoroethene. J. Pharmacol. exp.
    Ther., 235: 851-857.

    DOUGLAS, J.B. & HEALY, J.K. (1969) Nephrotoxic effects of amphotericin
    B, including renal tubular acidosis. Am. J. Med., 46: 154-162.

    DRUET, P. (1989) Contribution of immunological reactions to
    nephrotoxicity. Toxicol. Lett., 46: 55-64.

    DRUET, P., BERNARD, A., HIRSCH, F., WEENING, J.J., GENOUX, P., MAHIEU,
    P., & BIRKELAND, S. (1982) Immunologically mediated glomerulonephritis
    induced by heavy metals. Arch. Toxicol., 50: 187-194.

    DRUET, P., JACQUOT, C., BARAN, D., KLEINKNECHT, D., FILLASTRE, J.P.,
    & MERY, J.Ph. (1987) Immunologically mediated nephritis induced by
    toxins and drugs. In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity in
    the experimental and the clinical situation, Dordrecht, Boston,
    Lancaster, Martinus Nijhoff Publishers, pp. 727-770.

    DUBACH, U.C., LE HIR, M., & GANDHI, R. (1989) Use of urinary enzymes
    as markers of nephrotoxicity. Toxicol. Lett., 46: 193-196.

    DUNHILL, M.S., MILLARD, P.R., & OLIVER, D. (1977) Acquired cystic
    disease of the kidney: a hazard of long-term intermittent maintenance
    dialysis. J. clin. Pathol., 30: 868-877.

    DUNN, M.A., BLALOCK, T.L., & COUSIN, R.J. (1987) Minireview of
    metallothionein. Proc. Soc. Exp. Biol. Med., 185: 107-119.

    DUNN, M.J. (1981) Prostaglandins and Bartter's syndrome. Kidney Int.,
    19: 86-102.

    DUNN, M.J. & HOOD, V.L. (1977) Prostaglandins and the kidney. Am. J.
    Physiol. 223: F169-F184.

    DUNN, M.J. & ZAMBRASKI, E.J. (1980) Renal effects of drugs that
    inhibit prostaglandin synthesis. Kidney Int., 18: 609-622.

    DYBING, E., OMICHINSKI, J.G., SODERLUND, E.J., BRUNBORG, G., LAG, M.,
    HOLME, J.A., & NELSON, S.D. (1989) Mutagenicity and organ damage of
    1,2-dibromo-3-chloropropane (DBCP) and
    tris(2,3-dibromopropyl)phosphate (tris-BP): role of metabolic
    activation. In: Hodgson, E., Bend, J.R., & Philpot, R.M., ed. Reviews
    in biochemical toxicology, Amsterdam, Oxford, New York, Elsevier
    Science Publishers, Vol. 10, pp. 139-186.

    EARLEY, L.E. & FRIEDLER, R.M. (1964) Observations on the mechanism of
    decreased tubular reabsorption of sodium and water during saline
    loading. J. clin. Invest., 43: 1928-1936.

    EARLEY, L.E. & FRIEDLER, R.M. (1965) Changes in renal blood flow and
    possibly the intra-renal distribution of blood during natriuresis
    accompanying saline loading in the dog. J. clin. Invest., 44: 929-941.

    EBLE, J.N. & HULL, M.T. (1984) Morphologic features of renal
    oncocytoma. A light and electron microscopic study. Hum. Pathol., 15:
    1054-1061.

    EINHORN, L.H. & DONOHUE, J. (1977) Cis-diamminedichloroplatinum,
    vinblastin and bleomycin combination chemotherapy in disseminated
    testicular cancer. Ann. intern. Med., 87: 293-298.

    EISENBERG, R.L., BANK, W.O., & HEDGCOCK, M.W. (1980) Renal failure
    after major angiography. Am. J. Med., 68: 43-46.

    ELBERS, R., KAMPFFMEYER, H.G., & RABES, H. (1980) Effects and
    metabolic pathway of 4-dimethylaminophenol during kidney perfusion.
    Xenobiotica, 10: 621-632.

    ELFARRA, A.A. & ANDERS, M.W. (1984) Renal processing of glutathione
    conjugates. Role in nephrotoxicity. Biochem. Pharmacol., 33:
    3729-3732.

    ELFARRA, A.A., JACOBSON, I., & ANDERS, M.W. (1986a) Mechanism of
    S-(1,2-dichlorovinyl) glutathione-induced nephrotoxicity. Biochem.
    Pharmacol., 35: 283-288.

    ELFARRA, A.A., LASH, L.H., & ANDERS, M.W. (1986b) Metabolic activation
    and detoxification of nephrotoxic cysteine and homocysteine
    S-conjugates. Proc. Natl Acad. Sci., 83: 2667-2671.

    ELINDER, C.G., EDLING, C., LINDBERG, E., BERTIL, K., & VESTERBERG, O.
    (1985a) Beta-2-microglobulinuria among workers previously exposed to
    cadmium: follow-up and dose-response analyses. Am. J. ind. Med., 8:
    553-564.

    ELINDER, C.G., EDLING, C., LINDBERG, E., KAGEDAL, B., & VESTERBERG, O.
    (1985b) Assessment of renal function in workers previously exposed to
    cadmium. Br. J. ind. Med., 42: 754-760.

    ELLIS, K.J., MORGAN, W.C., ZANZI, I., YASUMURA, S., VARTSKY, D., &
    COHN, S.H. (1981) Critical concentration of Cd in human renal cortex:
    dose-effect studies in Cd smelter workers. J. Toxicol. environ.
    Health, 7: 691-703.

    ELSEVIERS, M.M. & DE BROE, M.E. (1988) Is analgesic nephropathy still
    a problem in Belgium? Nephrol. Dialysis Transplant., 2: 143-149.

    EMMERSON, B.T. (1973) Chronic lead nephropathy. Kidney Int., 4: 1-5.

    EMSLIE, K.R., CALDER, I.C., HART, S.J., & TANGE, J.D. (1981) Induction
    of paracetamol metabolism in the isolated perfused kidney.
    Xenobiotica, 11: 43-50.

    ENDOU, H. (1983) Cytochrome P-450 monooxygenase system in the rabbit
    kidney: its intranephron localization and its induction. Jpn. J.
    Pharmacol., 33: 423-433.

    ENDOU, H., NONOGUCHI, Y., TAKEHARA, H., YAMADA, H., & NAKADA, J.
    (1985) Intranephron heterogeneity of ammoniagenesis and
    gluconeogenesis in rats. Contrib. Nephrol., 47: 98-104.

    ENTERLINE, P.E. & VIREN, J. (1985) Epidemiologic evidence for an
    association between gasoline and kidney cancer. Environ. Health
    Perspect., 62: 303-312.

    EPSTEIN, S.M., BARTUS, B., & FARBER, E. (1969) Renal epithelial
    neoplasms induced in male Wistar rats by oral aflatoxin B1. Cancer
    Res., 29: 1045-1050.

    ERTÜRK, E., COHEN, S.M., & BRYAN, G.T. (1970) Induction, histogenesis,
    and isotransplantability of renal tumors induced by formic acid
    2-[4-(5-nitro-2-furyl)-2-thiazolyl]hydrazide in rats. Cancer Res., 30:
    2098-2106.

    EUROPEAN  DIALYSIS AND TRANSPLANT ASSOCIATION REGISTRY (l986)
    Demography of dialysis and transplantation in Europe, 1984. Nephrol.
    Dialysis Transplant., 1: 1.

    EVARTS, R.P., BROWN, C.A., & MOSTAFA, H. (1982) Production of kidney
    tumors in rats with low dose of dimethylnitrosamine after partial
    hepatectomy. J. Natl Cancer Inst., 68: 293-298.

    FAJARDO, L.F., ELTRINGHAM, J.R., STEWART, J.R., & KLAUBER, M.R. (1980)
    Adriamycin nephrotoxicity. Lab. Invest., 43: 242-253.

    FANG, L.S., SIROTA, R.A., EBERT, T.H., & LICHTENSTEIN, N.S. (1980) Low
    fractional excretion of sodium with contrast media-induced acute renal
    failure. Arch. intern. Med., 140: 531-533.

    FARBER, S.J., WALAT, R.J., BENJAMIN, R., & VAN PRAAG, D. (1971) Effect
    of increased osmolality on glycosaminoglycan metabolism of rabbit
    renal papilla. Am. J. Physiol., 220: 880-885.

    FINKELSTEIN, A., FRALEY, D.S., STACHURA, I., FELDMAN, H.A., GANDY,
    D.R., & BOURKE, E. (1982) Fenoprofen nephropathy: Lipoid nephrosis and
    interstitial nephritis. Am. J. Med., 72: 81-86.

    FISCHER, L.J., GREEN, M.D., & HARMAN, A.W. (1985) Studies on the fate
    of the glutathione and cysteine conjugates of acetaminophen in mice.
    Drug Metab. Disp., 13: 121-126.

    FJELDBORG, P., SORENSEN, J., & HELKJAER, P.E. (1986) The long-term
    effect of cisplatin on renal function. Cancer, 58: 2214-2217.

    FOA', V., CAIMI, L., AMANTI, L., ANTONINI, C., GATTINONI, A.,
    TETTAMANTI, G., LOMBARDO, A., & GIULIANI, A. (1976) Patterns of some
    lysosomal enzymes in the plasma and of proteins in urine of workers
    exposed to inorganic mercury. Int. Arch. occup. environ. Health, 37:
    115-124.

    FOA', V., COLOMBI, A., MARONI, M., BARBIERI, F., FRANCHINI, I., MUTTI,
    A., DE ROSA, E., & BARTOLUCCI, G.B. (1987) Study of kidney function of
    workers with low level exposure to inorganic arsenic. In: Foa', V.,
    Emmett, E.A., Maroni, M., & Colombi, A., ed. Occupational and
    environmental chemical hazards, Chichester, Ellis Horwood Limited, pp.
    362-367.

    FOIDART, J.B., DECKANNE, C.A., MAHIEU, P., CREUTZ, C.E., & DE MEY, J.
    (1979) Tissue culture of normal rat glomeruli: Isolation and
    morphological characterization of two homogeneous cell lines. Invest.
    cell Pathol., 2: 15-26.

    FOIDART, J.B., DUBOIS, C.H., & FOIDART, J.-M. (1980) Tissue culture of
    normal rat glomeruli: Basement membrane biosynthesis by homogenous
    epithelial and mesangial cell lines. Int. J. Biochem., 12: 197-202.

    FOIDART, J.B., DECHENNE, C.A., & MAHIEU, P. (1981) Tissue culture of
    normal rat glomeruli: Characterization of collagenous and
    non-collagenous basement membrane antigens on the epithelial and
    mesangial cells. Diagn. Histopathol., 4: 71-77.

    FORD, S.M. & HOOK, J.B. (1984) Biochemical mechanisms of toxic
    nephropathies. Semin. Nephrol., 4: 88-106.

    FORREST, J.B., HOWARDS, S.S., & GILLENWATER, I.Y. (1981) Osmotic
    effects of intravenous contrast agents on renal function. J. Urol.,
    125: 147-150.

    FOWLER, B.A. & GOYER, R.A. (1975) Bismuth localization within nuclear
    inclusions by X-ray microanalysis: Effects of accelerating voltage.
    Histochem. Cytochem., 23: 722-727.

    FRANCHINI, I. & MUTTI, A. (1988) Selected toxicological aspects of
    chromium compounds. Sci. total Environ., 71: 379-387.

    FRANCHINI, I., CAVATORTA, A., FALZOI, M., LUCERTINI, S., & MUTTI, A.
    (1983) Early indicator of renal damage in workers exposed to organic
    solvents. Int. Arch. occup. environ. Health, 51: 1-9.

    FRIBERG, L. (1948) Proteinuria and kidney injury among workers exposed
    to cadmium and nickel dust. J. ind. Toxicol., 30: 32-36.

    FRIBERG, L. (1950) Health hazards in the manufacture of alkaline
    accumulators with special reference to chronic cadmium poisoning. Acta
    med. Scand., 138 (Suppl. 240): 1-154.

    FRIBERG, L., PISCATOR, M., NORDBERG, G.F., & KJELLSTROM, T. (1974)
    Cadmium in the environment, 2nd ed., Cleveland, Ohio, Chemical Rubber
    Co. Press.

    FRIBERG, L., ELINDER, C.G., KJELLSTROM, T., & NORDBERG, G.F. (1986)
    Cadmium and health: a toxicological and epidemiological appraisal,
    Boca Raton, Florida, CRC Press, pp. 1-307.

    FRIEDLANDER, G., BLANCHET, F., & AMIEL, C. (1989) Renal functional
    reserve. Toxicol. Lett., 46: 227-235.

    FROLICH, J.C. & WALKER, L.A. (1980) Determination, source, metabolism
    and functional role of renal prostaglandins. Clin. exp. Hypertension,
    2: 709-728.

    FROLICH, J.C., NIES, A.S., & SCHRIER, R.W., ed. (1981) Prostaglandins
    and the kidney. Kidney Int., 19: 755-868.

    FRY, J.R. & PERRY, N.K. (1981) The effect of Aroclor-1254 pretreatment
    on the phase I and phase II metabolism of 7-ethoxycoumarin in isolated
    viable rat kidney cells. Biochem. Pharmacol., 30: 1197-1201.

    FRY, J.R., WIEBKIN, P., KAO, J., JONES, C.A., GWYNN, J., & BRIDGES,
    J.W. (1978) A comparison of drug-metabolising capability in isolated
    viable rat hepatocytes and renal tubule fragments. Xenobiotica, 8:
    113-120.

    FUKUSHIMA, M., ISHIZAKI, A., NOGAWA, K., SAKAMOTO, M., & KOBAYASHI, E.
    (1974) Epidemiological studies on renal failure of inhabitants in
    Itai-Itai disease endemic district. I. Some findings of inhabitants
    living in and around the endemic district of the Jinzu River basin.
    Jpn. J. pub. Health, 21: 67-73.

    GALE, M.E., ROBBINS, A.H., HAMBURGER, R.J., & WIDRICH, W.C. (1984)
    Renal toxicity of contrast agents: iopamidol, iothalamate, and
    diatrizoate. Am. J. Roentgenol., 142: 333-335.

    GARTNER, H.V. (1980) Drug-associated nephropathy. In: Berry, C.L.,
    Grundmann, E., & Kirsten, W.H., ed. Drug-induced pathology, Berlin,
    Heidelberg, New York, Springer-Verlag, pp. 144-174 (Current Topics in
    Pathology, Vol. 69).

    GEMBORYS, M.W. & MUDGE, G.H. (1981) Formation and disposition of the
    minor metabolites of acetaminophen in the hamster. Drug Metab. Disp.,
    9: 340-351.

    GILBERT,  D.N., WOOD, C.A., KOHLHEPP,  S.J., KOHNEN, P.W., HOUGHTON,
    D.B., FINKBEINER, H.C., LINDSLEY, J., & BENNETT, W.M. (1989)
    Polyaspartic acid prevents experimental aminoglycoside nephrotoxicity.
    J. infect. Dis., 159: 945-953.

    GINETZINSKY, A.G. (1958) Role of hyaluronidase in the reabsorption of
    water in renal tubules. The mechanism of action of the anti-diuretic
    hormone. Nature (Lond.), 182: 1218-1219.

    GIULIANO, R.A., PAULUS, G.J., VERPOOTEN, R.A., PATTYN, V., POLLET,
    D.E., TULKENS, P.M., & DEBROE, M.E. (1984) Recovery of cortical
    phospholipidosis and necrosis after acute gentamicin loading in rats.
    Kidney Int., 26: 838-847.

    GIULIANO, R.A., VERPOOTEN, G.A., VERBIST, L., WEDEEN, R., & DEBROE,
    M.E. (1986)  In vivo uptake kinetics of aminoglycosides in the kidney
    cortex of rats. J. Pharmacol. exp. Ther., 236: 470-475.

    GLOOR, F.J. (1978) Changing concept in the pathogenesis and morphology
    of analgesic nephropathy as seen in Europe. Kidney Int., 13: 27-33.

    GOLDBERG, J.P. & ANDERSON, R.J. (1985) Renal metabolism and excretion
    of drugs. In: Seldin, D.W. & Giebisch, G., ed. The kidney: Physiology
    and pathophysiology, New York, Raven Press, pp. 2097-2110.

    GOLDSTEIN, R.S. & MAYOR, G.H. (1983) Minireview: The nephrotoxicity of
    cis-platin. Life Sci., 32: 685-690.

    GOLDSTEIN, R.S., CONTARDI, L.R., PASINO, D.A., & HOOK, J.B. (1987)
    Mechanisms mediating cephaloridine inhibition of gluconeogenesis.
    Toxicol. appl. Pharmacol., 87: 297-305.

    GOLDSTEIN, R.S., SOZIO, R.S., TARLOFF, J.B., & HOOK, J.B. (1989) The
    role of oxidative stress in cephaloridine nephrotoxicity. In: Bach,
    P.H. & Lock, E.A., ed. Nephrotoxicity: Extrapolation from  in vitro
    to  in vivo, and animals to man, New York, London, Plenum Press, pp.
    457-461.

    GOLMAN, K. & ALMEN, T. (1985) Contrast media-induced nephrotoxicity.
    Survey and present state. Invest. Radiol., 20: S92-S97.

    GONZALEZ-VITALE, J.C., HAYES, D.M., CVITKOVIC, E., & STERNBERG, S.
    (1977) The renal pathology in clinical trials of cis-platinum (II)
    diammine-dichloride. Cancer, 39: 1362-1371.

    GOODMAN, D.G., WARD, J.M., SQUIRE, R.A., CHU, K.C., & LINHART, M.S.
    (1979) Neoplastic and non-neoplastic lesions in aging F344 rats,
    Toxicol. appl. Pharmacol., 48: 237-248.

    GOODMAN, D.G., WARD, J.M., SQUIRE, R.A., PAXTON, M.B., REICHARDT,
    W.D., CHU, K.C., & LINHART, M.S. (1980) Neoplastic and non-neoplastic
    lesions in aging Osborne-Mendel rats. Toxicol. appl. Pharmacol., 55:
    433-447.

    GOODMAN, M.T., MORGENSTERN, H., & WYNDER, E.L. (1986) A case-control
    study of factors affecting the development of renal cell cancer. Am.
    J. Epidem., 124: 926-941.

    GOTELLI, C.A., ASTOLFI, E., COX, C., CERNICHIARI, E., & CLARKSON, T.W.
    (1985) Early biochemical effects of an organic mercury fungicide on
    infants: dose makes the poison. Science, 227: 638-640.

    GOTTSCHALK, C.W. & LASSITER, W.E. (1973) Micropuncture methodology.
    In: Orloff, J. & Berliner, W.R., ed. Handbook of physiology. Section
    8: Renal physiology, Washington, DC, American Physiological Society.

    GOYER, R.A. (1982) Nephrotoxic effects of lead. In: Bach, P.H.,
    Bonner, F.W., Bridges, J.W., & Lock, E.A., ed. Nephrotoxicity:
    Assessment and pathogenesis, New York, Chichester, Brisbane, Toronto,
    John Wiley and Sons.

    GOYER, R.A. & CHERIAN, M.G. (1977) Tissue and cellular toxicology of
    metals. In: Brown, S.S., ed. Clinical chemistry and chemical
    toxicology of metals, Amsterdam, Elsevier-North Holland Biomedical
    Press, pp. 89-103.

    GOYER, R.A. & MOORE, J.F. (1974) Cellular effects of lead. Adv. exp.
    Med. Biol., 48: 447-462.

    GOYER, R.A. & RHYNE, B.C. (1973) Pathological effects of lead. Int.
    Rev. exp. Pathol., 12: 1-77.

    GOYER, R.A. & WILSON, M.H. (1975) Lead-induced inclusion bodies:
    Results of EDTA treatment. Lab. Invest., 32: 149-156.

    GOYER, R.A., FALK, H.L., HOGAN, M., FELDMAN, D.D., & RICHTER, W.
    (1981) Renal tumors in rats given trisodium nitrilotriacetic acid in
    drinking water for 2 years. J. Natl Cancer Inst., 66: 869-880.

    GRANDCHAMP, A., AYER, G., SCHERRER, J.R., & TRUNIGER, B. (1971)
    Intrarenal haemodynamics of the rat kidney determined by the Xenon
    washout technique. Nephron, 8: 33-45.

    GREEN, C.R., HAM, K.N., & TANGE, J.D. (1969) Kidney lesions induced in
    rats by p-aminophenol. Br. med. J., 1: 162-164.

    GREEN, R.M. & ELCE, J.S. (1975) Acetylation of S-substituted cysteines
    by a rat liver and kidney microsomal N-acetyltransferase. Biochem. J.,
    147: 283-289.

    GREGG, N., ELSEVIERS, M.M., DEBROE, M.E., & BACH, P.H. (1989) The
    epidemiological and mechanistic basis of analgesic associated
    nephropathy. Toxicol. Lett., 46: 141-151.

    GRITZKA, T.L. & TRUMP, B.F. (1968) Renal tubular lesions caused by
    mercuric chloride. Am. J. Pathol., 52: 1225-1227.

    GROTH, S., NIELSEN, H., SORENSEN, J.B., CHRISTENSEN, A.B., PEDERSEN,
    A.G., & RORTH, M. (1986) Acute and long-term nephrotoxicity of
    cisplatinum in man. Cancer Chemother. Pharmacol., 17: 191-196.

    GRUNFELD, J.P., RAPHAEL, J.C., BANKIR, W., KLEINKNECHT, D., & BARGER,
    A.C. (1971) Intrarenal distribution of blood flow. Adv. Nephrol., 1:
    125-143.

    GSTRAUNTHALER, G., PFALLER, W., & KOTANKO, P. (1985) Interrelation
    between oxygen consumption and Na-K-ATPase activity in rat renal
    proximal tubule suspension. Renal Physiol., 8: 38-44.

    GUDER, W.G. & ROSS, B.D. (1984) Enzyme distribution along the nephron.
    Kidney Int., 26: 101-111.

    GUDER, W.G. & SCHMIDT, U. (1974) The localization of gluconeogenesis
    in the rat nephron: determination of PEPCK in microdissected tubules.
    Hoppe Seylers Z. Physiol. Chem., 355: 273-278.

    GUENGERICH, F.P. & MASON, P.S. (1979) Immunological comparison of
    hepatic and extrahepatic cytochromes P-450. Mol. Pharmacol., 15:
    154-64.

    GYRD-HANSEN, N. (1968) Renal clearances in pigs. Acta vet. Scand., 9:
    183-198.

    HACKENTHAL, E., SCHWERTSCHLAG, U., & SEYBERTH, H.W. (1980)
    Prostaglandins and renin release studies in the isolated perfused rat
    kidney. Prog. Biochem. Pharmacol., 17: 98-107.

    HALDER, C.A., WARNE, T.M., & HATOUM, N.S. (1984) Renal toxicity of
    gasoline and related petroleum naphthas in male rats. In: Mehlman,
    M.A., Hemstreet, C.P., Thorpe, J.J., & Weaver, N.K., ed. Renal effects
    of petroleum hydrocarbons, Princeton, New Jersey, Princeton Scientific
    Publishers, pp. 73-88.

    HALDER, C.A., HOLDSWORTH, C.E., & COOKRELL, B.Y. (1985) Hydrocarbon
    nephropathy in male rats. Identification of the nephrotoxic components
    of gasoline. In: Bach, P.H. & Lock, E.A., ed. Renal heterogeneity and
    target cell toxicity, New York, Chichester, Brisbane, Toronto, John
    Wiley and Sons, pp. 481-484.

    HALES, B.F., JAEGER, V., & NEIMS, A.H. (1978) Isoelectric focusing of
    glutathione S-transferases from rat liver and kidney. Biochem. J.,
    175: 937-943.

    HALL, C.L., FOTHERGILL, N.J., BLACKWELL, M.M., HARRISON, P.R.,
    MACKENZIE, J.C., & MACIVER, A.G. (1987) The natural course of gold
    nephropathy: long-term study of 21 patients. Br. med. J., 295:
    745-748.

    HALL, P.W., III (1982) Endemic Balkan nephropathy. In: Porter, G., ed.
    Nephrotoxic mechanisms and environmental toxins, New York, London,
    Plenum Press, pp. 227-240.

    HALL, P.W., III & DAMMIN, G.J. (1978) Balkan nephropathy. Nephron, 22:
    281-300.

    HALL, R.L., WILKE, W.L., & FETTMAN, M.J. (1986) The progression of
    adriamycin-induced nephrotic syndrome in rats and the effect of
    captopril. Toxicol. appl. Pharmacol., 82: 164-174.

    HAMILTON, J.M. (1975) Renal carcinogenesis. Adv. Cancer Res., 22:
    1-56.

    HAMMOND, P.B., LERNER, S.I., GARTSIDE, P.S., HANENSON, I.B., RODA,
    S.B., FOULKES, E.C., JOHNSON, D.R., & PESCI, A.J. (1980) The
    relationship of biological indices of lead exposure to the health
    status of workers in a secondary lead smelter. J. occup. Med., 22:
    475-484.

    HANDLER, J.S. (1983) Use of cultured epithelia to study transport and
    its regulation. J. exp. Biol., 106: 55-69.

    HANSEN, H.E., HESTBECH, J., SORENSEN, J.L.,  NORGAARD, K., HEILSKOV,
    J., & AMDISEN, A. (1979) Chronic interstitial nephritis in patients on
    long-term lithium treatment. Q. J. Med., 48: 577-591.

    HANSEN, H.E., MOGENSEN, C.E., SORENSEN, J.L., NORGAARD, K., HEILSKOV,
    J., & ANDERESEN, A. (1981) Albumin and beta-2-microglobulin excretion
    in patients on long-term lithium treatment. Nephron, 29: 229-232.

    HARD, G.C. (1975) Autoradiographic analysis of proliferative activity
    in rat kidney epithelial and mesenchymal cell subpopulations following
    a carcinogenic dose of dimethylnitrosamine. Cancer Res., 35:
    3762-3773.

    HARD, G.C. (1979) Effect of age at treatment on incidence and type of
    renal neoplasm induced in the rat by a single dose of
    dimethylnitrosamine. Cancer Res., 39: 4965-4970.

    HARD, G.C. (1980) Morphological correlates of irreversible tumour
    formation. In: Holmstedt, B., Lauwerys, R., Mercier, M., & Roberfroid,
    M., ed. Mechanisms of toxicity and hazard evaluation, Amsterdam,
    Elsevier/North Holland Biomedical Press, pp. 231-240.

    HARD, G.C., (1984) High frequency, single-dose model of renal
    adenoma/carcinoma induction using dimethylnitrosamine in Crl:(W)BR
    rats. Carcinogenesis, 5: 1047-1050.

    HARD, G.C. (1985) Identification of a high frequency model for renal
    carcinoma by the induction of renal tumors in the mouse with a single
    dose of streptozotocin. Cancer Res., 45: 703-708.

    HARD, G.C. (1986) Renal carcinogenesis, rat. In: Jones, T.C., Mohr,
    U., & Hunt, R.D., ed. Monographs on pathology of laboratory animals,
    urinary system, Berlin, Heidelberg, New York, Springer-Verlag, pp.
    45-49.

    HARD, G.C. (1987) Chemically induced epithelial tumours and
    carcinogenesis of the renal parenchyma. In: Bach, P.H. & Lock, E.A.,
    ed. Nephrotoxicity in the experimental and the clinical situation,
    Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers, pp.
    211-251.

    HARD, G.C. & BUTLER, W.H. (1970) Cellular analysis of renal neoplasia:
    induction of renal tumors in dietary-conditioned rats by
    dimethylnitrosamine with a reappraisal of morphological
    characteristics. Cancer Res., 30: 2796-2805.

    HARD, G.C. & BUTLER, W.H. (1971) Morphogenesis of epithelial neoplasms
    induced in the rat kidney by dimethylnitrosamine. Cancer Res., 31:
    1496-1505.

    HARD, G.C., MACKAY, R.L., & KOCHHAR, O.S. (1984) Electron microscopic
    determination of the sequence of acute tubular and vascular injury
    induced in the rat kidney by a carcinogenic dose of
    dimethylnitrosamine. Lab. Invest., 50: 659-672.

    HARKONEN, S. & KJELLSTRAND, C.M. (1979) Intravenous pyelography in
    non-uremic diabetic patients. Nephron, 24: 268-270.

    HARKONEN, S. & KJELLSTRAND, C.M. (1981) Contrast nephropathy. Am. J.
    Nephrol., 1: 69-77.

    HARRIS, C.C., WESTON, A., WILLEY, J.C., TRIVERS, G.E., & MANN, D.L.
    (1987) Biochemical and molecular epidemiology of human cancer:
    Indicators of carcinogen exposure, DNA damage, and genetic
    predisposition. Environ. Health Perspect., 75: 109-119.

    HARRIS, S.I., BALABAN, R.S., BARRETT, L., & MANDEL, L.J. (1981)
    Mitochondrial respiratory capacity and Na+- and K+-dependent
    adenosine triphosphatase-mediated ion transport in the intact renal
    cell. J. biol. Chem., 256: 10319-10328.

    HASSALL, C.D., GANDOLFI A.J., & BRENDEL K. (1983) Correlation of the
     in vivo and  in vitro renal toxicity of
    S-(1,2-dichlorovinyl)-L-cysteine. Drug chem. Toxicol., 6: 507-520.

    HAUGLUSTAINE, D., VAN DAMME, B., DAENENS, P., & MICHIELSEN, P. (1980)
    Silicon nephropathy, a possible occupational hazard. Nephron, 26:
    219-224.

    HAWKE, R.L. & WELCH, R.M. (1985) Major differences in the specificity
    and regulation of mouse renal cytochrome P-450 dependent
    monooxygenases. Mol. Pharmacol., 28: 283-289.

    HAYES, D.M., CVITKOVIC, E., GOLBERG, R.B., SCHEINER, E., HELSON, L.,
    & KRAKOFF, I.H. (1977) High dose cis-platinum diammine chloride:
    amelioration of renal toxicity by mannitol diuresis. Cancer, 39:
    1372-1381.

    HELLSTEN, S., BERGE, T., & LINELL, F. (1983) Clinically unrecognized
    renal carcinoma. Aspects of tumor morphology, lymphatic and
    haematogenous metastatic spread. Br. J. Urol., 55: 166-170.

    HENNIS, H.L., ALLEN, R.C., HENNIGAR, G.R., & SIMMONS, M.A. (1981) A
    sensitive method for determining the nephrotoxic effects of the
    analgesic acetaminophen upon esterases using isoelectric focusing.
    Electrophoresis, 2: 187-190.

    HESTBECH, J., HANSEN, H.E., AMDISEN, A., & OLSEN, S. (1977) Chronic
    renal lesions following long-term treatment with lithium. Kidney Int.,
    12: 205-213.

    HIASA, Y., OHSHIMA, M., IWATA, C., & TANIKAKE, T. (1979)
    Histopathological studies on renal tubular cell tumours in rats
    treated with N-ethyl-N-hydroxyethylnitrosamine. Gann, 70: 817-820.

    HIASA, Y., OHSHIMA, M., KITAHORI, Y., FUJIYATA, T., YUASA, T., &
    MIYASHIRO, A. (1983) Basic lead acetate: Promoting effect on the
    development of renal tubular cell tumours in rats treated with
    N-ethyl-N-hydroxyethyl-nitrosamine. J. Natl Cancer Inst., 70: 761-765.

    HIASA, Y., ENOKI, N., KITAHORI, Y., KONISHI, N., & SHIMOYAMA, T.
    (1984a) DL-serine: Promoting activity on renal tumorigenesis by
    N-ethyl-N-hydroxyethylnitrosamine in rats. J. Natl Cancer Inst., 73:
    297-299.

    HIASA, Y., KITAHORI, Y., KONISHI, N., ENOKI, N., SHIMOYAMA, T., &
    MIYASHIRO, A. (1984b) Trisodium nitrilotriacetate monohydrate:
    promoting effects on the development of renal tubular cell tumors in
    rats treated with N-ethyl-N-hydroxyethylnitrosamine. J. Natl Cancer
    Inst., 72: 483-489.

    HIASA, Y., LIN, J.-C., KONISHI, N., KITAHORI, Y., ENOKI, N., &
    SHIMOYAMA, T. (1984c) Histopathological and biochemical analyses of
    transplantable renal adenocarcinoma in rats induced by
    N-ethyl-N-hydroxyethyl-nitrosamine. Cancer Res., 44: 1664-1670.

    HINSON, J.A., MONKS, T.T., HONG, M., HIGHET, R.J., & POHL, L.R. (1982)
    3-(Glutathion-S-yl)acetaminophen: a biliary metabolite of
    acetamino-phen. Drug Metab. Disp., 10: 47-50.

    HINTON, D.E., HEATFIELD, B.M., LIPSKY, M.M., & TRUMP, B.F. (1980)
    Animal model: chemically induced renal tubular carcinoma in rats. Am.
    J. Pathol., 100: 317-320.

    HIRAFUJI, M., SATOH, S., & OGURA, Y. (1980) Sex difference in
    stimulatory actions of cofactors on prostaglandin synthetase in
    microsomes from rat kidney medulla. Biochem. Pharmacol., 29:
    2635-2637.

    HIRSCH, G.H. (1976) Differential effects of nephrotoxic agents on
    renal transport and metabolism by use of  in vitro techniques.
    Environ. Health Perspect., 15: 89-99.

    HJELLE, J.T., PETERSON, D.R., & HJELLE, J.J. (1983) Drug metabolism in
    isolated proximal tubule cells: aldehyde dehydrogenase. J. Pharmacol.
    exp. Ther., 224: 699-706.

    HOOK, J.B., ed. (1981) Toxicology of the kidney, New York, Raven
    Press.

    HOOK, J.B., MCCORMACK, K.M., & KLUWE, W.M. (1979) Biochemical
    mechanisms of nephrotoxicity. In: Hodgson, E., Bend, J.R., & Philpot,
    R.M., ed. Reviews in biochemical toxicology, Amsterdam, Oxford, New
    York, Elsevier Science Publishers, pp. 53-78.

    HOOK, J.B., ELCOMBE, C.R., ROSE, M.S., & LOCK, E.A. (1982)
    Characterization of the effects of known hepatic monooxygenase
    inducers on male and female rat and mouse kidneys. Life Sci., 31:
    1077-1084.

    HOOK, J.B., ISHMAEL, J., & LOCK, E.A. (1983) Nephrotoxicity of
    hexachloro-1:3-butadiene in the rat: the effect of age, sex and
    strain. Toxicol. appl. Pharmacol., 67: 121-131.

    HORNING, E.S. & WHITTICK, J.W. (1954) The histogenesis of
    stilboestrol-induced renal tumours in the male golden hamster. Br. J.
    Cancer, 8: 451-457.

    HORROBIN, D.F. (1980) The regulation of prostaglandin biosynthesis:
    negative feedback mechanisms and the selective control of formulation
    of 1 and 2 series prostaglandin: relevance to inflammation and
    immunity. Med. Hypotheses, 6: 687-709.

    HOU, S.H., BUSHINSKY, D.A., WISH, J.B., COHEN, J.J., & HARRINGTON,
    J.T. (1983) Harrington, Hospital-acquired renal insufficiency: a
    prospective study. Am. J. Med., 74: 243-248.

    HULT, K. & FUCHS, R. (1986) Analysis and dynamics of ochratoxin A in
    biological systems. In: Steyn, P.S. & Vleggaar, R., ed. Mycotoxins and
    phycotoxins. Sixth International IUPAC Symposium on Mycotoxins and
    Phycotoxins, Pretoria, Republic of South Africa, 22-25 July, 1985,
    Amsterdam, Oxford, New York, Elsevier Science Publishers, pp. 365-376.

    HUMES, H.D., WEINBERG, J.M., & KNAUSS, T.C. (1982) Clinical and
    pathophysiological aspects of aminoglycoside nephrotoxicity. Am. J.
    kidney Dis., 2: 5-29.

    IESATO, K., WAKASHIN, M., WAKASHIN, Y., & TOJO, S. (1977) Renal
    tubular dysfunction in Minamata disease. Ann. intern. Med., 86:
    731-737.

    INOUE, G., SAWADA, T., & YOSHIKAWA, M. (1973) Age-related change in
    acid mucopolysaccharides level and water content in papillae.
    Gerontologia, 19: 73-78.

    ITO, N., HIASA, Y., TAMAI, A., & YOSHIDA, K. (1969) Effect of
    unilateral nephrectomy on the development of kidney tumor in rats
    treated with N-nitro-sodimethylamine. Gann, 60: 319-327.

    ITO, N., SUGIHARA, S., MAKIURA, S., ARAI, M., HIRAO, K., DENDA, A., &
    NISHIO, O. (1974) Effect of N-(3,5-dichlorophenyl)succinimide on the
    histological pattern and incidence of kidney tumors in rats induced by
    dimethylnitrosoamine. Gann, 65: 131-138.

    JACOBSON, A., GRIECO, A.J., & FARBER, S.J. (1964) Hexosamine analysis
    of renal papillae in diuretic and antiduiretic rats. Proc. Soc. Exp.
    Biol. Med., 115: 1153-1156.

    JACOBSEN, B.K., BJELKE, E., KVALE, G., & HEUCH, I. (1986) Coffee
    drinking, mortality and cancer incidence: results from a Norwegian
    prospective study. J. Natl Cancer Inst., 76: 823-831.

    JACQUES, P.J. (1975) The endocytic uptake of macromolecules. In:
    Trump, B.F. & Arstilla, U.A., ed. Pathology of cell membranes, New
    York, London, San Franciso, Academic Press, Vol. 1, pp. 225-276.

    JAFFE, N., KEIFER, R., ROBERTSON, R., CANGIR, A., & WANG, A. (1987)
    Renal toxicity with cumulative doses of
    cis-diamminedichloroplatinum-II in pediatric patients with
    osteosarcoma. Cancer, 59: 1577-1581.

    JAKOBY, W.B. & PASTAN, I.H., ed. (1979) Cell culture, methods in
    enzymology, New York, London, San Franciso, Academic Press, Vol. 58,
    642 pp.

    JASMIN, G. & RIOPELLE, J.L. (1976) Renal carcinomas and erythrocytosis
    in rats following intrarenal injection of nickel subsulfide. Lab.
    Invest., 35: 71-78.

    JOHANSSON, S. & ANGERVALL, L. (1976) Urothelial hyperplasia of the
    renal papillae in female Sprague-Dawley rats induced by long term
    feeding of phenacetin. Acta pathol. microbiol. Scand. Sect. A., 84:
    353-354.

    JOHANSSON, S., ANGERVALL, L., BENGTSSON, U., & WAHLQVIST, L. (1976) A
    clinicopathologic and prognostic study of epithelial tumors of the
    renal pelvis. Cancer, 37: 1376-1383.

    JOLLOW, D.J., KOCSIS, J.J., SNYDER, R., & VAINIO, H., ed. (1976)
    Biological reactive intermediates. Formation, toxicity and
    inactivation, New York, London, Plenum Press.

    JONES, B.R., BHALLA, R.B., MLADEK, J., KALEYA, R.N., GRALLA, R.J.,
    ALLOCK, N.N., SCHWARTZ, M.K., YOUNG, C.W., & KEIDENBERG, M.M. (1980)
    Comparison of methods of evaluating nephrotoxicity of cis-platinum.
    Clin. Pharmacol. Ther., 24: 557-562.

    JONES, D.P., SUNDBY, G.-B., ORMSTAD, K., & ORRENIUS, S. (1979) Use of
    isolated kidney cells for study of drug metabolism. Biochem.
    Pharmacol., 28: 929-935.

    JOSEPOVITZ, C., FARRUGGELLA, T., LEVINE, R., LANE, B., & KALOYANIDES,
    G.J. (1985) Effect of netilmicin on the phospholipid composition of
    subcellular fractions of rat renal cortex. J. Pharmacol. exp. Ther.,
    235: 810-819.

    JUNG, K.Y. & ENDOU, H. (1989) Nephrotoxicity assessment by measuring
    cellular ATP content II. Intranephron site of ochratoxin A
    nephrotoxicity. Toxicol. appl. Pharmacol., 100: 383-390.

    JUNG, K.Y. & ENDOU, H. (1989) Biphasic increasing effect of
    angiotensin II on intracellular free calcium in isolated rat early
    proximal tubule. Biochem. Biophys. Res. Commun., 165: 1221-1228.

    JUNG, K.Y., UCHIDA, S., & ENDOU. H. (1989) Nephrotoxicity assessment
    by measuring cellular ATP content I. Substrate specificities in the
    maintenance of ATP content in isolated rat nephron segments. Toxicol.
    appl. Pharmacol., 100: 369-382.

    KACEW, S. (1987) Detection of nephrotoxicity of foreign compounds with
    the use of  in vitro and  in vivo techniques. In: Bach, P.H. & Lock,
    E.A., ed. Nephrotoxicity in the experimental and the clinical
    situation, Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers,
    pp. 533-562.

    KAHLMETER, G. & DAHLAGER, J.I. (1984) Aminoglycoside toxicity - a
    review of clinical studies published between 1975 and 1982. J.
    antimicrob. Chemother., 13(Suppl. A): 9-22.

    KALOYANIDES, G.J. (1984a) Renal pharmacology of aminoglycoside
    nephrotoxicity. Contrib. Nephrol., 42: 148-167.

    KALOYANIDES, G.J. (1984b) Aminoglycoside induced functional and
    biochemical defects in the renal cortex. Fundam. appl. Toxicol., 4:
    930-943.

    KALOYANIDES, G.G. & PASTORIZA-MUNOZ, E. (1980) Aminoglycoside
    nephro-toxicity. Kidney Int., 18: 571-582.

    KAMINSKY, L.S., FASCO, M.J., & GUENGERICH, F.P. (1979) Comparison of
    different forms of liver, kidney and lung microsomal cytochrome P-450
    by immunological inhibition of regio- and stereoselective metabolism
    of warfarin. J. biol. Chem., 254: 9657-9662.

    KANFER, A. (1989) The role of coagulation in glomerular injury.
    Toxicol. Lett., 46: 83-92.

    KANISAWA, M. & SUZUKI, S. (1978) Induction of renal and hepatic tumors
    in mice by ochratoxin A, a mycotoxin. Gann, 69: 599-600.

    KANWAR, Y.S. & FARQUHAR, M.G. (1979) Presence of heparan sulfate in
    the glomerular basement membrane. Proc. Natl Acad. Sci., 76:
    1303-1307.

    KAPLOWITZ, N. (1980) Physiological significance of glutathione
    S-transferases. Am. J. Physiol., 239: G439-G444.

    KATZBERG, R.W., MORRIS T.W., & BURGENER, F.A. (1977) Renal renin and
    hemodynamic responses to selective renal artery catheterization and
    angiography. Invest. Radiol., 12: 381-388.

    KATZBERG, R.W., PABICO, R.C., MORRIS, T.W., HAYAKAWA,  K., MCKENNA,
    B.A., PANNER, B.J., VENTURA, J.A., & FISHER, H.W. (1986) Effects of
    contrast media on renal function and subcellular morphology in the
    dog. Invest. Radiol., 21: 64-70.

    KAUMP, D.H. (1966) Pharmacology of the fenamates: II. Toxicology in
    animals. Ann. physiol. Med., Suppl. 1: 16-23.

    KENNEDY, J.F. (1979) Proteoglycans: biological and chemical aspects in
    human life, Amsterdam, Oxford, New York, Elsevier Scientific
    Publishers.

    KHOURY, G.A., HOPPER, J.C., VARGHESE, Z., FARRINGTON,  K., DICK, R.,
    IRVING, J.D., SWENY, P., FERNANDO, O.N., & MOORHEAD, J.F. (1983)
    Nephrotoxicity of ionic and non-ionic contrast material in digital
    vascular imaging and selective renal arteriography. Br. J. Radiol.,
    56: 631-635.

    KILHAM, L., LOW, R.J., CONTI, S.F., & DALLENBACH, F.D. (1962)
    Intranuclear inclusions and neoplasms in the kidneys of wild rats. J.
    Natl Cancer Inst., 29: 863-885.

    KINCAID-SMITH, P. (1979) Analgesis nephropathy in Australia. Contrib.
    Nephrol., 16: 57-64.

    KINCAID-SMITH, P., BURROWS, G.D., DAVIES, B.M., HOLWILL, B., WALTER,
    M., & WALKER, R.G. (1979) Renal biopsy findings in lithium and
    prelithium patients. Lancet, 2(8144): 700-701.

    KITCHEN, D.N. (1984) Neoplastic renal effects of unleaded gasoline in
    Fischer 344 rats. In: Mehlman, M.A., Hemstreet, C.P., Thorpe, J.J., &
    Weaver, N.K., ed. Renal effects of petroleum hydrocarbons, Princeton,
    New Jersey, Princeton Scientific Publishers, pp. 65-71.

    KJELLEN, P., OLDBERG, A., RUBIN, K., & HOOK, M. (1977) Binding of
    heparin and heparan sulphate to rat liver cells. Biochem. biophys.
    Res. Commun., 74: 126-133.

    KJELLSTROM, T. & NORDBERG, G.F. (1978) The kinetic model of cadmium
    metabolism in the human being. Environ. Res., 16: 248-269.

    KJELLSTROM, T., SHIROISHI, K., & EVRIN, P. (1977) Urinary
    beta2-microglobulin excretion among people exposed to cadmium in the
    general environment. An epidemiological study in cooperation between
    Japan and Sweden. Environ. Res., 13: 318-344.

    KLEIN, K.L., WANG, M.S., TORIKAI, S., DAVISDON, W.D., & KUROKAWA, K.
    (1981) Substrate oxidation by isolated single nephron segments of the
    rat. Kidney Int., 20: 29-35.

    KLEINKNECHT, D., KANFER, A., MOREL-MAROGER, L., & MERY, J.P. (1978)
    Immunologically mediated drug-induced acute renal failure. Contrib.
    Nephrol., 10: 42-52.

    KLINTMALM, G.B.M., IWATSUKI, S., & STARZL, T.E. (1981) Nephro-toxicity
    of cyclosporine in liver and kidney transplant recipients. Lancet,
    1(8218): 470-471.

    KLOSS, M.W., COX, M.G., NORTON, R.M., SWENBERG, J.A., & BUS, J.S.
    (1985) Sex-dependent differences in the disposition of
    14C-5-2,2,4-trimethylpentane in Fischer 344 rats. In: Bach, P.H. &
    Lock, E.A., ed. Renal heterogeneity and target cell toxicity, New
    York, Chichester, Brisbane, Toronto, John Wiley and Sons, pp. 489-492.

    KLUWE, W.M. (1983) Chemical modulation of 1,2-dibromo-3-chloropropane.
    Toxicology, 27: 287-299.

    KLUWE, W.M. & HOOK, J.B. (1978) Analysis of gentamicin uptake by rat
    renal cortical slices. Toxicol. appl. Pharmacol., 45: 531-539.

    KLUWE, W.M. & HOOK, J.B. (1980) Effects of environmental chemicals on
    kidney metabolism and function. Kidney Int., 18: 648-655.

    KLUWE, W.M., HERRMANN, C.L., & HOOK, J.B. (1979) Effects of dietary
    polychlorinated biphenyls and polybrominated biphenyls on the renal
    and hepatic toxicities of several chlorinated hydrocarbon solvents in
    mice. J. Toxicol. environ. Health, 5: 605-615.

    KLUWE, W.M., ABDO, K.M., & HUFF, J. (1984) Chronic kidney disease and
    organic chemical exposures: evaluations of causal relationships in
    humans and experimental animals. Fundam. appl. Toxicol., 4: 889-901.

    KOCIBA,  R.J., KEYES, D.G.,  JERSEY, G.C., BALLARD,  J.J., DITTENBER,
    D.A., QUAST, J.F., WADE, C.E., HUMISTON, C.G., & SCHWARTZ, B.A. (1977)
    Results of a two year chronic toxicity study with hexachlorobutadiene
    in rats. Am. Ind. Hyg. Assoc. J., 38: 589-602.

    KOJIMA, S., HAGA, Y., KURIHARA, T., & YAMAWAKI, T. (1977) A comparison
    between fecal cadmium and urinary beta2-microglobulin, total protein
    and cadmium among Japanese farmers. Environ. Res., 14: 436-451.

    KOSEK, J.D., MAZZE, R.I., & COUSINS, M.J. (1974) Nephrotoxicity of
    gentamicin. Lab. Invest., 30: 48-57.

    KOSEKI, C., YAMAGUCHI, Y., FURUSAWA, M., & ENDOU, H. (1988) Isolation
    by monoclonal antibody of intercalated cells of rabbit kidney. Kidney
    Int., 33: 543-554.

    KOSTAKIS, A.J., WHITE, D.J.G., & CALNE, R.Y. (1977) Toxic effects in
    the use of cyclosporin A in alcoholic solution as an immunosuppressant
    of rat heart allografts. IRCS Surg. Transplant., 5: 243.

    KRECH, R., ZERBAN, H., & BANNASCH, P. (1981) Mitochondrial anomalies
    in renal oncocytes induced in rats by N-nitrosomorpholine. Eur. J.
    cell Biol., 25: 331-339.

    KREISBERG, J.I. & KARNOVSKY, M.J. (1983) Glomerular cells in culture.
    Kidney Int., 23: 439-447

    KREISBERG, J.I., PITTS, A.M., & PRETLOW, T.G. (1977) Separation of
    proximal tubule cells from suspensions of rat kidney cells in density
    gradients of ficoll in tissue culture medium. Am. J. Pathol., 86:
    591-600.

    KREISBERG, J.I., HOOVER, R.L., & KARNOVSKY, M.J. (1978) Isolation and
    characterization of rat glomerular epithelial cells  in vitro. Kidney
    Int., 14: 21-30.

    KRESSE, H. & GROSSMANN, A. (1970) [Comparative study of the
    mucopolysaccharide and collagen content in various topographical areas
    of the kidney in the rat, dog and pig.] Z. Klin. Chem., 8: 420-424 (in
    German).

    KRIJSHELD, K.R. & GRAM, T.E. (1984) Selective induction of renal
    microsomal cytochrome P-450-linked monooxygenases by
    1,1-dichloroethylene in mice. Biochem. Pharmacol., 33: 1951-1956.

    KRIZ, W. & BANKIR, L. (1988) A standard nomenclature for structures of
    the kidney, Am. J. Physiol., 254: F1-F8.

    KROGH, P., HALD, B., PLESTINA, R., & CEOVIC, S. (1977) Balkan
    (endemic) nephropathy and foodborne ochratoxin A: preliminary results
    of a survey of foodstuffs. Acta pathol. microbiol. Scand. Sect. B, 85:
    238-240.

    KUHN, J.A., ARGY, W.R., HAKOWSKI, T.A., SCHREINER, G.E., & SCHEIN,
    P.S. (1978) Nephrotoxicity of cis-diamminedichloroplatinum as measured
    by urinary beta glucuronidase. Clin. Res., 26: 776.

    KUO, C.-H. & HOOK, J.B. (1982) Depletion of renal glutathione content
    and nephrotoxicity of cephaloridine in rabbits, rats and mice.
    Toxicol. appl. Pharmacol., 63: 292-302.

    KUO, C.-H. & HOOK, J.B. (1983) Effects of age and sex on
    hexachloro-1,3-butadiene toxicity in the Fisher 344 rat. Life Sci.,
    33: 517-523.

    KUO, C.-H., BRASELTON, W.E., & HOOK, J.B. (1982) Effect of
    phenobarbital on cephaloridine toxicity and accumulation in rabbit and
    rat kidneys. Toxicol. appl. Pharmacol., 64: 244-254.

    KUROKAWA, Y., HAYASHI, Y., MAEKAWA, A., TAKAHASHI, M., KOKUBO, T., &
    ODASHIMA, S. (1983) Carcinogenicity of potassium bromate administered
    orally to F344 rats. J. Natl Cancer Inst., 71: 965-972.

    LANDRIGAN, P.J., GOYER, R.A., CLARKSON, T.W., SANDLER, D.P., SMITH,
    J.H., THUN, M.J., & WEDEEN, R.P. (l984) The work-relatedness of renal
    disease. Arch. environ. Health, 39: 225-230.

    LANGARD, S. & NORSETH, T. (1986) Chromium. In: Friberg, L., Nordberg,
    G., & Vouk, V.B., ed. Handbook on the toxicology of metals, Amsterdam,
    Oxford, New York, Elsevier Science Publishers, Vol. II, pp. 185-210.

    LAQUEUR, G.L. & SPATZ, M. (1968) Toxicology of cycasin. Cancer Res.,
    28: 2262-2267.

    LASH L.H. & ANDERS M.W. (1986) Cytotoxicity of
    S-(1,2-dichlorovinyl)-glutathione and S-(1,2-dichlorovinyl)-L-cysteine
    in isolated rat kidney cells. J. biol. Chem., 261: 13076-13081.

    LASH, L.H., ELFARRA, A.A., & ANDERS, M.W. (1986) Renal cysteine
    conjugate betalyase. Bioactivation of nephrotoxic cysteine
    S-conjugates in mitochondrial outer membrane. J. biol. Chem., 261:
    5930-5935.

    LAURENT, G., CARLIER, M.-B., ROLLMAN, B., VANHOOF, F., & TULKENS, P.M.
    1982) Mechanism of aminoglycoside-induced lysosomal phospholipidosis:
     in vitro and  in vivo studies with gentamicin and amikacin.
    Biochem. Pharmacol., 31: 3861-3870.

    LAURENT, G., TOUBEAU, G., HEUSON-STEINNON, J.A., TULKENS, P., &
    MALDAUGE, P. (1988) Kidney tissue repair after nephrotoxic injury:
    biochemical and morphologic characterization. CRC crit. Rev. Toxicol.,
    19: 147-183.

    LAUWERYS, R., ROELS, H., BERNARD, A., & BUCHET, J.-P. (1980) Renal
    response to cadmium in a population living in a non-ferrous smelter
    area in Belgium. Int. Arch. occup. environ. Health, 45: 271-274.

    LEE, J.B. (1980) Prostaglandins and the renin-angiotensin axis. Clin.
    Nephrol., 14: 159-163.

    LEEMING, B.W.A., SPOKES, K.C., & SILVA, P. (1985) Effect of meglumine
    iothalamate on renal hemodynamics and function in the diabetic rat.
    Invest. Radiol., 20: 971-977.

    LEVENSON, D.J., SIMMONS, C.E., & BRENNER, B.H. (1982) Arachadonic acid
    metabolism, prostaglandins and the kidney. Am. J. Med., 72: 354-374.

    LI, J.J. & LI, S.A. (1984) Estrogen-induced tumorigenesis in hamsters:
    roles for hormonal and carcinogenic activities. Arch. Toxicol., 55:
    110-118.

    LI, J.J., LI, S.A., KLICKA, J.K., PARSONS, J.A., & LAM, L.K.T. (1983)
    Relative carcinogenic activity of various synthetic and natural
    estrogens in the Syrian hamster kidney. Cancer Res., 43: 5200-5204.

    LILIS, R. (1981) Long-term lead exposure, chronic nephropathy, and
    renal cancer: a case report. Am. J. ind. Med., 2: 293-297.

    LIND, C., VADI, H., & ERNSTER, L. (1978) Metabolism of
    benzo(a)pyrene-3,6-quinone and 3-hydroxybenzo(a)pyrene in liver
    microsomes from 3-methyl-cholanthrene-treated rats. Arch. Biochem.
    Biophys., 190: 97-108.

    LINDBERG, E. & VESTERBERG, O. (1983) Urinary excretion of proteins in
    chromeplaters, ex-chromeplaters and referents. Scand. J. Work Environ.
    Health, 9: 505-510.

    LIPPMAN, A.J., HELSON, C., HELSON, L., & KRAKOFF, I.H. (1973) Clinical
    trials of cis-diamminedichloroplatinum (NSC-119875). Cancer Chemother.
    Rep., 57: 191-200.

    LIPPMANN, S. (1982) Lithium's effects on the kidney. Postgrad. Med.,
    71: 99-104, 107-108.

    LITTERST, C.L. & WEISS, R.B. (1987) Clinical and experimental
    nephrotoxicity of cancer chemotherapeutic agents. In: Bach, P.H. &
    Lock, E.A., ed. Nephrotoxicity in the experimental and clinical
    situation, Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers,
    pp. 771-816.

    LITTERST, C.L., MIMNAUGH, E.G., REAGAN, R.L., & GRAM, T.E. (1975a)
    Comparison of  in vitro drug metabolism by lung, liver and kidney of
    several common laboratory species. Drug Metab. Disp., 3: 259-265.

    LITTERST, C.L., MIMNAUGH, E.G., REAGAN, R.L., & GRAM, T.E. (1975b)
    Drug metabolism by microsomes from extrahepatic organs of rat and
    rabbit prepared by calcium aggregation. Life Sci., 17: 813-818.

    LITTERST, C.L., MIMNAUGH, E.G., & GRAM, T.E. (1977) Alterations in
    extrahepatic drug metabolism by factors known to affect hepatic
    activity. Biochem. Pharmacol., 26: 749-755.

    LOCK, E.A. (1979) The effect of paraquat and diquat on renal function
    in the rat. Toxicol. appl. Pharmacol., 48: 327-336.

    LOCK, E.A. (1987) Metabolic activation of halogenated chemicals and
    its relevance to nephrotoxicity. In: Bach, P.H. & Lock, E.A., ed.
    Nephrotoxicity in the experimental and clinical situation, Dordrecht,
    Martinus Nijhoff, pp. 429-461.

    LOCK, E.A. & ISHMAEL, J. (1979) The acute toxic effects of paraquat
    and diquat on the rat kidney. Toxicol. appl. Pharmacol., 50: 67-76.

    LOCK, E.A., ISHMAEL, J., & HOOK, J.B. (1984) Nephrotoxicity of
    hexachloro-1,3-butadiene in the mouse: the effect of age, sex, strain,
    monooxygenase modifiers, and the role of glutathione. Toxicol. appl.
    Pharmacol., 72: 484-494.

    LOCK, E.A., ODUM, J., & ORMOND, P. (1986) Transport of
    N-acetyl-S-pentachloro-1,3-butadienylcysteine by the renal cortex.
    Arch. Toxicol., 59: 12-15.

    LOCKARD, W.V., PHILLIPS, R.D., HAYES, A.W., BERNDT, W.O., & O'NEAL,
    R.M. (1980) Citrinin nephrotoxicity in rats: A light electron
    microscopic study. Exp. mol. Pathol., 32: 266-340.

    LONG, W.F. & WILLIAMSON, F.B. (1979) Glycosaminoglycans, calcium ions
    and the control of cell proliferation. IRCS med. Sci., 7: 429-434.

    LOURY, D., SMITH-OLIVER, T., & BUTTERWORTH, B.E. (1987) Assessment of
    the binding potential of 2,2,4-trimethylpentane to the rat
    alpha-2u-globulin. Toxicol. appl. Pharmacol., 88: 44-56.

    LUTZ, W.K, JAGGI, W., & SCHLATTER, C. (1982) Covalent binding of
    diethylstilbestrol to DNA in rat and hamster liver and kidney.
    Chem.-biol. Interact., 42: 251-257.

    MACFARLAND, H.N., ULRICH, C.E., HOLDSWORTH, C.E., KITCHEN, D.N.,
    HALLIWELL, W.H., & BLUM, S.C. (1984) A chronic inhalation study with
    unleaded gasoline vapor. J. Am. Coll. Toxicol., 3: 231-248.

    MCAULIFFE, W.G. (1978) The effects of anti-diuretic hormone on the
    morphology and histochemistry of the renal medullary interstitium of
    rats with hereditary hypothalmic diabetes insipidus (Brattleboro
    strain). Anat. Rec., 190: 474.

    MCAULIFFE, W.G. (1980) Histochemistry and ultrastructure of the
    interstitium of the renal papilla in rats with hereditary diabetes
    insipidus (Brattleboro strain). Am. J. Anat., 157: 17-26.

    MCCONKEY, D.J., HARTZELL, P., DUDDY, S.K., HÅKANSSON, H., & ORRENIUS,
    S. (1988) 2,3,7,8-Tetrachlorodibenzo-p-dioxin kills immature
    thymocytes by Ca2+- mediated endonuclease activation. Science, 242:
    256-259.

    MCINTOSH, C.S., MOSELEY, I.F., FRY, I.K., & CATTELL, W.R. (1975)
    Excretion urography toxicity studies in experimental acute renal
    failure. Nephron, 14: 373-377.

    MCKINNEY, L.L., WEAHLY, F.B., ELDRIDGE, A.C., CAMPBELL, R.E., COWAN,
    J.C., PICKEN, J.C., Jr, & BIESTER, H.E. (1957)
    S-(Dichlorovinyl)-L-cysteine: an agent causing total aplastic anemia
    in calves. J. Am. Chem. Soc., 79: 3932-3933.

    MCKINNEY, L.L., PICKEN J.C., Jr, WEAKLEY, F.B., ELDRIDGE, A.C.,
    CAMPBELL, R.E., COWAN, J.C., & BIESTER, H.E. (1959) Possible toxic
    factor of trichloroethylene-extracted soybean oil meal. J. Am. Chem.
    Soc., 81: 909-915.

    MCLACHLAN, J.R., GOYER, R.A., & CHERIAN, M.G. (1980) Formation of
    lead-induced inclusion bodies in primary rat kidney epithelial cell
    cultures. Toxicol. appl. Pharmacol., 56: 418-431.

    MCLACHLAN, M.S.F., CHICK, S., ROBERTS, E.E., & ASSCHER, A.W. (1972)
    Intravenous urography in experimental acute renal failure in the rat.
    Invest. Radiol., 7: 466-473.

    MCLEAN, A.E.M. & MAGEE, P.N. (1970) Increased renal carcinogenesis by
    dimethylnitrosamine in protein deficient rats. Br. J. exp. Pathol.,
    51: 587-590.

    MCMICHAEL, A.J. & JOHNSON, H.M. (1982) Long term mortality profile of
    heavily-exposed lead smelter workers. J. occup. Med., 24: 375-378.

    MCMURTRY, R.J., SNODGRASS, W.R., & MITCHELL, J.R. (1978) Renal
    necrosis, glutathione depletion and covalent binding after
    acetaminophen. Toxicol. appl. Pharmacol., 46: 87-100.

    MADIAS, N.E. & HARRINGTON, J.T. (1978) Platinum nephrotoxicity
    (review). Am. J. Med., 65: 307-314.

    MAEKAWA, A., KUROKAWA, Y., TAKAHASHI, M., KOKUBO, T., OGIU, T.,
    ONODERA, H., TANIGAWA, H., OHNO, Y., FURUKAWA, F., & HAYASHI, Y.
    (1983) Spontaneous tumors in F-344/DuCrj rats. Gann, 74: 365-372.

    MAHAFFEY, J.R. (1980) Nutrient-lead interactions. In: Singhal, R.L. &
    Thomas, J.A., ed. Lead toxicity, Baltimore, Maryland, Urban and
    Schwarzenberg, pp. 425-460.

    MALCOLM, D. & BARNETT, H.A.R. (1982) A mortality study of lead workers
    1925-76. Br. J. ind. Med., 39: 404-410.

    MALEQUE, A., ENDOU, H., KOSEKI, C., & SAKAI, F. (1980) Nephron
    hetero-geneity: gluconeogenesis from pyruvate in rabbit nephron. FEBS
    Lett., 116: 154-156.

    MALIS, C.D. & BONVENTRE, J.V. (1986) Mechanism of calcium potentiation
    of oxygen free radical injury to renal mitochondria. J. biol. Chem.,
    261: 14201-14208.

    MANDAL, A.K. & BOHMAN, S.O., ed. (1980) The renal papilla and
    hypertension, New York, London, Plenum Press.

    MAO, P. & MOLNAR, J.J. (1967) The fine structure and histochemistry of
    lead induced renal tumors in rats. Am. J. Pathol., 50: 571-603.

    MARKOVIC, B. (1972) Endemic nephritis and urinary tract cancer in
    Yugoslavia, Bulgaria and Rumania. J. Urol., 107: 212-219.

    MARRE, R., TARARA, N., & LOUTON, T. (1980) Age dependent
    nephrotoxicity and the pharmacokinetics of gentamicin in rats. Eur. J.
    Pediatr., 133: 25-29.

    MARTIN, G., DUROZARD, D., & BAVEREL, G. (1987) Acceleration of
    ammonia-genesis in isolated rat kidney tubules by the antiepileptic
    drug: valproic acid. In: Kovacevic, Z. & Guder, W.G., ed. Molecular
    nephrology. Biochemical aspects of kidney function, Berlin, New York,
    Walter de Gruyter & Co., pp. 287-292.

    MARTIN, G., MICHOUDET, C., & BAVEREL, G. (1989) Stimulation of
    glutamine metabolism by the antiepileptic drug, sodium valproate, in
    isolated dog kidney tubules. Biochem. Pharmacol., 38: 3947-3952.

    MARTIN, G., DUROZARD, D., BESSON, J., & BAVEREL, G. (1990) Effect of
    the antiepileptic drug, sodium valproate, on glutamine and glutamate
    metabolism in isolated human kidney tubules. Biochim. Biophys. Acta,
    1033: 261-266.

    MATSUSAKA, T., FUJII, M., NAKANO, T., TERAI., T., KURATA, A.,
    IMAIZUMI, M., & ABE, H. (1989) Germanium-induced nephropathy: Report
    of two cases and review of the literature. Clin. Nephrol., 30:
    341-345.

    MATZKE, G., LUCAROTTI, R., & SHARPIRO, H.S. (1983) Controlled
    comparison of gentamicin and tobramycin nephrotoxicity. Am. J.
    Nephrol., 3: 11-17.

    MAUNSBACH, A.N., OLSEN, T.S., & CHRISTENSEN, E.I., ed. (1980)
    Functional ultrastructure of the kidney, New York, London, Academic
    Press.

    MAYER, D., WEBER, E., KADENBACH, B., & BANNASCH, P. (1989)
    Immunocyto-chemical demonstration of cytochrome c oxidase as a marker
    for renal oncocytes and oncocytomas. Toxicol. Pathol., 17: 46-49.

    MAZZE, R.I. (1976) Methoxyflurane nephropathy. Environ. Health
    Perspect., 15: 111-119.

    MAZZE, R.I. (1981) Methoxyflurane nephropathy. In: Hook, J.B., ed.
    Toxicology of the kidney, New York, Raven Press, pp. 135-149.

    MEHLMANN, M.A., HEMSTREET, C.P., THORPE, J.J., & WEAVER, N.K., ed.
    (1984) Renal effects of petroleum hydrocarbons, Princeton, New Jersey,
    Princeton Scientific Publishers.

    MEEZAN, E. & BRENDEL, K. (1973) Effect of ethacrynic acid on oxidative
    metabolism in isolated glomeruli. J. Pharmacol. exp. Ther., 187:
    352-364.

    MEIJER, S., SLEIJFER, D., MULDER, N., SLUITER, W.J., MARRINK, J.,
    SCHRAFFORD KOOPS, H., BROUWERS, T.M., OLDHOFF, J., VAN DER HEM, G.K.,
    & MANDEMA, E. (1983) Some effects of combination chemotherapy with
    cisplatin on renal function in patients with nonseminomatous
    testicular carcinoma. Cancer, 51: 2035-2040.

    METZLER, M. (1981) Studies on the mechanism of carcinogenicity of
    diethylstilboestrol: role of metabolic activation. Food Cosmet.
    Toxicol., 19: 611-615.

    METZLER, M. & MCLACHLAN, J.A. (1978) Peroxidase-mediated oxidation, a
    possible pathway for metabolic activation of diethylstilbestrol.
    Biochem. biophys. Res. Commun., 85: 874-884.

    MICHELS, L.D., DAVIDMAN, M., & KEANE, W.F. (1983) Adriamycin nephrotic
    syndrome: glomerular haemodynamics and permselectivity. Kidney Int.,
    23: 246.

    MICHOUDET, C. & BAVEREL, G. (1987) Metabolism of acetaldehyde in human
    and baboon renal cortex. FEBS Lett., 216: 113-117.

    MIHATSCH, M.J. & KNUSLI, K. (1982) Phenacetin abuse and malignant
    tumours. An autopsy study covering 25 years (1953-1977). Klin.
    Wochenschr., 60: 1339-1349.

    MIHATSCH, M.J., HOFER, H.O., GUTZWILER, F., BRUNNER, F.P., &
    ZOLLINGER, H.U. (1980a) [Phenacetin abuse. I. Frequency, per capita
    consumption and consequential costs.] Schweiz. med. Wochenschr., 110:
    108-115 (in German).

    MIHATSCH, M.J., SCHMIDLIN, P., BRUNNER, F.P. HOFER, H.O., SIX, P., &
    ZOLLINGER, H.U. (1980b) [Phenacetin abuse. II. Chronic renal
    insufficiency and autopsy material in Basel.] Schweiz. med.
    Wochenschr., 110: 116-124 (in German).

    MIHATSCH, M.J., MANZ, T., KNUSLI, C., HOFER, H.O., RIST, M., GUETG,
    R., RUTISHAUSER, G., & ZOLLINGERS, H.U. (1980c) [Phenacetin abuse.
    III. Malignant tumours of the urinary tract associated with phenacetin
    abuse in Basel, 1963-1977.] Schweiz. med. Wochenschr., 110: 255-264
    (in German).

    MIHATSCH, M.J., HOFER, H.O., GUDAT, F., KNUSLI, C., TORHORST, J., &
    ZOLLINGER, U. (1984) Capillary sclerosis of the lower urinary tract
    and analgesic nephropathy. Clin. Nephrol., 20: 285-301.

    MIHATSCH, M.J., THIEL, G., BASLER, V., RYFFEL, B., LANDSMANN, J., VON
    OVERBECK, J., & ZOLLINGER, H.U. (1985) Morphologic patterns in
    cyclosporin-treated renal transplant recipients. Transplant. Proc.,
    17(Suppl. 1): 101-116.

    MIHATSCH, M.J., RYFFEL, B., HERMLE, M., BRUNNER, F.P., & THIEL, G.
    (1986) Morphology of cyclosporine nephrotoxicity in the rat. Clin.
    Nephrol., 25(Suppl. 1): S2-S8.

    MILLER, E.C. & MILLER, J.A. (1981) Searches for ultimate chemical
    carcinogens and their reactions with cellular macromolecules. Cancer,
    47: 2327-2345.

    MILMAN, N. & GOTTLIEB, P. (1977) Renal function after high-dose
    urography in patients with chronic renal insufficiency. Clin.
    Nephrol., 7: 250-254.

    MITCHELL, J.R., JOLLOW, D.J., POTTER, W.Z., DAVIS, D.C., GILLETTE,
    J.R., & BRODIE, B.B. (1973) Acetaminophen-induced hepatic necrosis. I.
    Role of drug metabolism. J. Pharmacol. exp. Ther., 187: 185-194.

    MITCHELL, J.R., MCMURTRY, R.J., STATHAM, C.N., & NELSON, S.D. (1977)
    Molecular basis for several drug-induced nephropathies. Am. J. Med.,
    62: 518-526.

    MITSUMORI, K., MAITA, K., SAITO, T., TSUDA, S., & SHIRASU, Y. (1981)
    Carcinogenicity of methylmercury chloride in ICR mice: preliminary
    note on renal carcinogenesis. Cancer Lett., 12: 305-310.

    MODAN, B., BOICHIS, H., BOTT-KANNER, G., BARELL, V., BAR-HOACH, N., &
    ELIAHOU, H.E. (l975) An epidemiological study of renal failure. l. The
    need for maintenance dialysis. Am. J. Epidemiol., 101: 276-280.

    MOFFAT, D.B. (1979) The mammalian kidney, Cambridge, New York,
    Cambridge University Press.

    MOFFAT, D.B. (1981) New ideas on the anatomy of the kidney. J. clin.
    Pathol., 34: 1197-1206.

    MOFFAT, D.B. (1982) Morphology of the kidney in relation to
    nephrotoxicity -portae renales. In: Bach, P.H., Bonner, F.W., & Lock,
    E.A., ed. Nephrotoxicity: Assessment and pathogenesis, New York,
    Chichester, Brisbane, Toronto, John Wiley and Sons, pp. 10-26.

    MOHANDAS, J., DUGGIN, G.G., HORVATH, J.S., & TILLER, D.J. (1981a)
    Regional differences in peroxidatic activation of paracetamol
    (acetaminophen) mediated by cytochrome P-450 and prostaglandin
    endoperoxide synthetase in rabbit kidney. Res. Commun. chem. Pathol.
    Pharmacol., 34: 69-80.

    MOHANDAS, J., DUGGIN, G.G., HORVATH, J.S., & TILLER, D.J. (1981b)
    Metabolic oxidation of acetaminophen (paracetamol) mediated by
    cytochrome P-450 mixed function oxidase and prostaglandin endoperoxide
    synthetase in rabbit kidney. Toxicol. appl. Pharmacol., 61: 252-259.

    MOHR, U., HAAS, H., & HILFRICH, J. (1974) The carcinogenic effects of
    dimethylnitrosamine and nitrosomethylurea in European hamsters
    ( Cricetus cricetus L.). Br. J. Cancer, 29: 359-364.

    MOLD, J.W. & STEIN, H.F. (1986) Sounding board: the cascade effect in
    the clinical care of patients. New Engl. J. Med., 314: 512-514.

    MOLDEUS, P., JONES, D.P., ORMSTAD, K., & ORRENIUS, S. (1978) Formation
    and metabolism of a glutathione S-conjugate in isolated rat liver and
    kidney cells. Biochem. biophys. Res. Commun., 83: 195-200.

    MOORE, J.R., GOYER, R.A., & WILSON, M.H. (1973) Lead-induced inclusion
    bodies: Solubility, amino acid content, and relationship to residual
    acidic proteins. Lab. Invest., 29: 488-494.

    MOORE, M.A., NAKAMURA, T., SHIRAI, T., & ITO, N. (1986)
    Immuno-histochemical demonstration of increased glucose-6-phosphate
    dehydrogenase in preneoplastic and neoplastic lesions induced by
    propylnitrosamines in F 344 rats and Syrian hamsters. Gann, 77:
    131-138.

    MOREAU, J.F., DROZ, D., NOEL, L.H., LEIBOWITCH, J., JUNGERS, P., &
    MICHEL, J.R. (1980) Tubular nephrotoxicity of water soluble iodinated
    contrast media. Invest. Radiol., 15: S54-S60.

    MOREL, F., CHABARDES, D., & IMBERT, M. (1976) Functional segmentation
    of the rabbit distal tubule by microdetermination of hormone-dependent
    adenylate cyclase activity. Kidney Int., 9: 264.

    MORIN, J.P., VIOTTE, G., VANDERWALLE, A., VAN HOOF, F., TULKENS, P.,
    & FILLASTRE, J.P. (1980) Gentamicin-induced nephrotoxicity: a cell
    biology approach. Kidney Int., 18: 583-590.

    MORITA, T., OITE, T., KIHARA, I., YAMAMOTO, T., HARA, M., NAKA, A., &
    OHNO, S. (1980) Culture of isolated glomeruli from normal and
    nephritic rabbits. I. Characterization of outgrowing cells. Acta
    pathol. Jpn., 30: 917-926.

    MORRISON, A.R. (1980) Prostaglandins and the kidney. Am. J. Med., 69:
    171-173.

    MOSTOFI, F.K., SESTERHENN, I.A., & SOBIN, L.H. (1981) Histological
    typing of kidney tumours, Geneva, World Health Organization, 26 pp.

    MUDGE, G.H. (1982) Analgesic nephropathy: renal drug distribution and
    metabolism. In: Porter, G.A., ed. Nephrotoxic mechanisms of drugs and
    environmental toxins, New York, London, Plenum Press, pp. 209-225.

    MUDGE, G.H. (1985) Pathogenesis of nephrotoxicity: Pharmacological
    principles. In: Bach, P.H. & Lock, E.A., ed. Renal heterogeneity and
    target cell toxicity, New York, Chichester, Brisbane, Toronto, John
    Wiley and Sons, pp. 1-12.

    MUIRHEAD, E.E. & PITCOCK, J.A. (1980) Evidence on involvement of the
    renal papilla in hypertension. In: Mandal, A.K. & Bohman, S.-O., ed.
    The renal papilla and hypertension, New York, London, Plenum Press,
    pp. 35-61.

    MURPHY, G.P., MIRAND, E.A., JOHNSTON, G.S., SCHMIDT, J.D., & SCOTT,
    W.W. (1966) Renal tumors induced by a single dose of
    dimethylnitrosamine: morphologic, functional, enzymatic and hormonal
    characterizations. Invest. Urol., 4: 39-56.

    MURRAY, B.M., PALLER, M.S., & FERRIS, T.F. (1985) Effect of
    cyclosporine administration on renal hemodynamics in conscious rats.
    Kidney Int., 28: 767-774.

    MURRAY, T. & GOLDBERG, M. (1975) Chronic interstitial nephritis:
    Etiologic factors. Ann. intern. Med., 82: 453-459.

    MUTTI, A. (1987) The study of urinary excretion of kidney antigens to
    reveal early effects of exposure to exogenous chemicals. In: Foa', V.,
    Emmett, E.A., Maroni, M., & Colombi, A., ed. Occupational and
    environmental chemical hazards, Chichester, Ellis Horwood Limited, pp.
    315-322.

    MUTTI, A. (1989) Detection of renal diseases in humans: developing
    markers and methods. Toxicol. Lett., 46: 177-191.

    MUTTI, A., CAVATORTA, A., PEDRONI, C., BORGHI, A., GIAROLI, C., &
    FRANCHINI, I. (1979) The role of chromium accumulation in the
    relationship between airborne and urinary chromium in welders. Int.
    Arch. occup. environ. Health, 43: 123-133.

    MUTTI, A., LUCERTINI, S., VALCAVI, P.P., NERI, T.M., FORNARI, M.,
    ALINOVI, R., & FRANCHINI, I. (1985) Urinary excretion of brush-border
    antigen revealed by monoclonal antibody: early indicator of toxic
    nephropathy. Lancet, 2(8461): 914-917.

    MUTTI, A., ALINOVI, R., BERGAMASCHI, E., FORNARI, M., & FRANCHINI, I.
    (1988) Monoclonal antibodies to brush-border antigens for the early
    diagnosis of nephrotoxicity. Arch. Toxicol., Suppl. 12: 162-165.

    NAKADA, J., YAMADA, H., & ENDOU, H. (1986a) Evidence that
    alpha-1-adrenergic stimuli specifically increase gluconeogenesis of
    the isolated proximal convoluted tubule in the rat. Renal Physiol., 9:
    213-222.

    NAKADA, J., MACHIDA, T., & ENDOU, H. (1986b) Nephrotoxicity of
    cisplatin in rats. In: Tanabe, T., Hook, J.B., & Endou, H., ed.
    Nephrotoxicity of antibiotics and immunosuppressants, Amsterdam,
    Oxford, New York, Elsevier Science Publishers, pp. 179-182.

    NANRA, R.S. (1980) Clinical and pathological aspects of analgesic
    nephropathy. Br. J. clin. Pharmacol., 10: 359S-368S.

    NANRA, R.S., STUART-TAYLOR, J., DELEON, A.H., & WHITE, K.H. (1978)
    Analgesic nephropathy: etiology, clinical syndrome, and
    clinicopatho-logic correlations in Australia. Kidney Int., 13: 79-92.

    NASH, J.A., KING, L.J., LOCK, E.A., & GREEN, T. (1984) The metabolism
    and disposition of hexachloro-1:3-butadiene in the rat and its
    relevance to nephrotoxicity. Toxicol. appl. Pharmacol., 73: 124-137.

    NAVRAN, S.S. & LOUIS-FERDINAND, R.T. (1975) P-Chloro-N-methylaniline
    demethylation by rat kidney subcellular fractions. Res. Commun. chem.
    Pathol. Pharmacol., 12: 713-721.

    NELSON, S.D. (1982) Metabolic activation and drug toxicity. J. med.
    Chem., 25: 753-765.

    NEUHAUS, O.W. (1986) Renal reabsorption of low-molecular weight
    proteins in adult male rats: alpha-2u globulin. Proc. Soc. Exp. Biol.
    Med., 182: 531-539.

    NEUHAUS, O.W., FLORY, W., BISWAS, N., & HOLLERMAN, C.E. (1981) Urinary
    excretion of alpha2u-globulin and albumin by adult male rats following
    treatment with nephrotoxic agents. Nephron, 28: 133-140.

    NEWSON, G.D. & VUGRIN, D. (1987) Etiologic factors in renal cell
    adeno-carcinoma. Semin. Nephrol., 7: 108-116.

    NEWTON, J.F., KUO, C.H., GEMBORYS, M.W., MUDGE, G.H., & HOOK, J.B.
    (1982) Nephrotoxicity of p-aminophenol: a metabolite of acetaminophen
    in the Fischer 344 rat. Toxicol. appl. Pharmacol., 65: 336-344.

    NEWTON, J.F., BAILIE, M.B., & HOOK, J.B. (1983a) Acetaminophen
    nephro-toxicity in the rat. Renal metabolic activation  in vitro.
    Toxicol. appl. Pharmacol., 70: 433-444.

    NEWTON, J.F., YOSHIMOTO, J., BERNSTEIN, J., RUSH, G.F., & HOOK, J.B.
    (1983b) Acetaminophen nephrotoxicity in the rat. I. Strain differences
    in nephrotoxicity and metabolism. Toxicol. appl. Pharmacol., 69:
    291-306.

    NICOLL, J.W., SWANN, P.F., & PEGG, A.E. (1975) Effect of
    dimethylnitro-samine on persistence of methylated guanines in rat
    liver and kidney DNA. Nature (Lond)., 254: 261-262.

    NOGAWA, K., KOBAYASHI, E., & HONDA, R. (1979a) A study of the
    relationship between cadmium concentrations in urine and renal effects
    of cadmium. Environ. Health Perspect., 28: 161-168.

    NOGAWA, K., ISHIZAKI, A., & KOBAYASHI, E. (1979b) A comparison between
    health effects of cadmium and cadmium concentration in urine among
    inhabitants of the Itai-Itai disease endemic districts. Environ. Res.,
    18: 397-409.

    NOGAWA, K., HONDA, R., KIDO, T., TSURITANI, I., YAMADA, Y., ISHIZAKI,
    M., & YAMAYA, H. (1989) A dose-response analysis of cadmium in the
    general environment with special reference to total cadmium intake
    limit. Environ. Res., 48: 7-16.

    NOGUEIRA, E. & BANNASCH, P. (1988) Cellular origin of rat renal
    oncocytoma. Lab. Invest., 59: 337-343.

    NOGUEIRA, E., KLIMEK, F., WEBER, E., & BANNASCH, P. (1989) Collecting
    duct: origin of rat renal clear cell tumours. Virchows Arch. B Cell
    Pathol., 57(5): 275-283.

    NOMIYAMA, K. (1980) Recent progress and perspectives in cadmium health
    effects studies. Sci. total Environ., 14: 199-232.

    NONOGUCHI, H., UCHIDA, S., SHIIGAI, T., & ENDOU, H. (1985) Effect of
    chronic metabolic acidosis on ammonia production from L-glutamine in
    microdissected rat nephron segments. Pflugers Arch., 403: 229-235.

    NONOGUCHI, H., TAKEHARA, Y., & ENDOU, H. (1986) Intra- and
    inter-nephron heterogeneity of ammoniagenesis in rats: Effect of
    chronic metabolic acidosis and potassium depletion. Pflugers Arch.,
    407: 245-251.

    NORBY L.H. & DIBONA, G.F. (1975) The renal vascular effects of
    meglumine diatrizoate. J. Pharmacol. exp. Ther., 193: 932-940.

    NTP (1983) Technical report on the carcinogenesis bioassay of
    penta-chloroethane (CAS No. 76-01-7) in F344/N rats and B6c3F1 mice
    (gavage studies), Research Triangle Park, North Carolina, US
    Department of Health and Human Services, National Toxicology Program
    (NIH Publication No. 83-1788).

    NTP (1986) Technical report on the toxicology and carcinogenesis
    studies of 1,4-dichlorobenzene (CAS No. 106-46-7) in F344/N rats and
    B6c3F1 mice (gavage studies), Research Triangle Park, North Carolina,
    US Department of Health and Human Services, National Toxicology
    Program (NIH Publication No. 86-2575).

    NTP (1987) Toxicology and carcinogenesis studies of
    1,4-dichlorobenzene in F344/N rats and B6C3F1 mice, Research Triangle
    Park, North Carolina, US Department of Health and Human Services,
    National Toxicology Program (NTP TR 319) (NIH Publication No.
    87-2575).

    NTP (1988) NTP draft technical report on the toxicology and
    carcinogenesis studies of ochratoxin A (CAS No. 303-47-9) in F344/N
    rats (gavage studies), Research Triangle Park, North Carolina, US
    Department of Health and Human Services, National Toxicology Program
    (NIH Publication No. 88-2813).

    NUYTS, G.D., ELSEVIERS, M.M., & DE BROE, M.E. (1989) Health impact of
    renal disease due to nephrotoxicity. Toxicol. Lett., 46: 31-44.

    OBATOMI, D.K. & PLUMMER, D.T. (1986) Oxygen uptake and glutamate
    dehydro-genase release in isolated renal tubules from rat kidney.
    Biochem. Soc. Trans., 14: 1281-1282.

    OHI, G., YAMAMOTO, S., FUJINO, T., ITAI, Y., UENO, K., SAITO, H.,
    INOBA, Y., NAKAJIMA, K., & SHIROISHI, K. (1982) Urinary
    beta2-microglobulin does not serve as diagnostic tool for Minamata
    disease. Arch. environ. Health, 37: 336-341.

    OHMORI, T., HIASA, Y., MURATA, Y., & WILLIAMS, G.M. (1982) Gamma
    glutamyl transpeptidase activity in carcinogen-induced epithelial
    lesions of rat kidney. Gann, 73: 543-548.

    OHMORI, T. & TABEI, R. (1983) Modulation of N-nitrosodimethylamine
    kidney tumorigenesis by unilateral hydronephrosis and multiple
    putrescine adminis-trations. J. Natl Cancer Inst., 71: 787-793.

    OLORUNSOGO, O.O., MALOMO, S.O., & BABABUNMI, E.A. (1985)
    Proto-nophoric properties of fluorinated arylalkylsulphonamides:
    Observations with perfluidone. Biochem. Pharmacol., 34: 2945-2952.

    OLSON, C.T., YU, K.O., & SERVE, M.P. (1986) Metabolism of nephrotoxic
    cis-and trans-decalin in Fischer-344 rats. J. Toxicol. environ.
    Health, 18: 285-292.

    OLSON, M.J., GARG, B.D., MURTY, C.V.R., & ROY, A.K.  (1987)
    Accumulation of alpha2u-globulin in the renal proximal tubules of male
    rats exposed to unleaded gasoline. Toxicol. appl. Pharmacol., 90:
    43-51.

    OMICHINSKI, J.G., BRUNBORG, G., SODERLUND, E.J., DAHL, J.E., BAUSANO,
    J.A., HOLME, J.A., NELSON, S.D., & DYBING, E. (1987) Renal necrosis
    and DNA damage caused by selectively deuterated and methylated analogs
    of 1,2-dibromo-3-chloropropane in the rat. Toxicol. appl. Pharmacol.,
    91: 358-370.

    ORLOFF, J. & BERLINER, R.W., ed. (1973) Handbook of physiology.
    Section 8: Renal physiology, Washington, DC, American Physiological
    Society.

    ORMSTAD, K. (1982) The use of isolated kidney cells, glomeruli and
    subcellular fractions for assessing renal function. In: Bach, P.H.,
    Bonner, F.W., Bridges, J.W., & Lock, E.A., ed. Nephrotoxicity:
    Assessment and pathogenesis, New York, Chichester, Brisbane, Toronto,
    John Wiley and Sons, pp. 161-168.

    ORMSTAD, K. (1987) Metabolism of glutathione in the kidney. In: Bach,
    P.H. & Lock, E.A., ed. Nephrotoxicity in the experimental and the
    clinical situation, Dordrecht, Boston, Lancaster, Martinus Nijhoff
    Publishers, pp. 405-428.

    ORMSTAD, K., JONES, D.P., & ORRENIUS, S. (1981) Preparation and
    characterisation of isolated kidney cells. Methods Enzymol., 77:
    137-146.

    ORTMANN, M., VIERBUCHEN, M., KOLLER, G., & FISCHER, R. (1988) Renal
    oncocytoma. I. Cytochrome c oxidase in normal and neoplastic renal
    tissue as detected by immunohistochemistry - a valuable aid to
    distinguish oncocytomas from renal cell carcinomas. Virchows Arch. B
    Cell Pathol., 56: 165-173.

    OTTOSEN,  P.D., SIGH, B., CHRISTENSEN, J., OLSEN, S., & CHRISTENSEN,
    S. (1984) Lithium induced interstitial nephropathy associated with
    chronic renal failure. Acta pathol. microbiol. immunol. Scand., 92(6):
    447-54.

    OZOLS, R.F., CORDEN, B.J., JACOB, J., WESLEY, M.N., OSTCHAGA, Y., &
    YOUNG, R.C. (1984) High dose cisplatin in hypertonic saline. Ann.
    intern. Med., 100: 19-24.

    PALESTINE, A.G., AUSTIN, H.A., BALOW, J.E., ANTONOVYCH, T.T., SABNIS,
    S.G., PREUSS, H.G., & NUSSENBLATT, R.B. (1986) Renal histo-pathologic
    alterations in patients treated with cyclosporine for uveitis. New
    Engl. J. Med., 314: 1293-1298.

    PARKER, R.A., BENNETT, W.M., & PORTER, G.A. (1982) Animal models in
    the study of aminoglycoside nephrotoxicity. In: Whelton, A. & Neu,
    H.S., ed. The aminoglycosides: Microbiology, clinical use and
    toxicology, New York, Basel, Marcel Dekker, pp. 235-267.

    PEARSON, R.M. (1979) Methods for the assessment of the effects of
    drugs on renal blood flow. Br. J. clin. Pharmacol., 7: 129-138.

    PEPELJNJAK, S. & CVETNIC, Z. (1985) The mycotoxicological chain and
    contamination of food by ochratoxin A in the nephropathic and
    non-nephropathic areas in Yugoslavia. Mycopathologia, 90: 147-153.

    PERANTONI, A. & BERMAN, J.J. (1979) Properties of Wilms' tumor line
    (TuWi) and pig kidney line (LLC-PK1) typical of normal kidney tubular
    epithelium. In Vitro, 15: 446-454.

    PETERSON, C.D., COLLINS, A.J., HINES, J. M., BULLOCK, M.L., & KEANE,
    W.F. (1981) Ethylene glycol poisoning: pharmacokinetics during therapy
    with ethanol and hemodialysis. New Engl. J. Med., 304: 21-23.

    PETKOVA-BOCHAROVA, T. & CASTEGNARO, M. (1985) Ochratoxin A
    con-tamination of cereals in an area of high incidence of Balkan
    endemic nephropathy in Bulgaria. Food Addit. Contam., 2: 267-270.

    PHILLIPS R.D. & COCKRELL, B.Y. (1984a) Kidney structural changes in
    rats following inhalation exposure to C10-C11 isoparaffinic solvent.
    Toxicology, 33: 261-273.

    PHILLIPS, R.D. & COCKRELL, B.Y. (1984b) Effect of certain light
    hydrocarbons on kidney function and structure in male rats. In:
    Mehlman, M.D., Hemstreet, C.O., Thrope, J.J., & Weaver, N.K., ed.
    Renal effects of petroleum hydro-carbons, Princeton, New Jersey,
    Princeton Scientific Publishers, pp. 89-105.

    PHILLIPS, R.D. & EGAN, G.F. (1984a) Effect of C10-C11 isoparaffinic
    solvent on kidney function in Fischer 344 rats during eight weeks of
    inhalation. Toxicol. appl. Pharmacol., 73: 500-510.

    PHILLIPS, R.D. & EGAN, G.F. (1984b) Subchronic inhalation exposure of
    dearomatized white spirit and C10-C11 isoparaffinic hydrocarbon in
    Sprague-Dawley rats. Fundam. appl. Toxicol., 4: 808-818.

    PHILLIPS, R.D., HAYES, A.W., & BERNDT W.O. (1979) Disposition of
    14C-citrinin in the rat. Toxicology, 12: 285-298.

    PHILLIPS, R.D., HAYES, A.W., & BERNDT, W.O. (1980a) High pressure
    liquid chromatographic analysis of citrinin and its application to
    biological systems. J. Chromatogr., 190: 419-425.

    PHILLIPS, R.D., HAYES, A.W., BERNDT, W.O., & WILLIAMS. W.L. (1980b)
    Effects of citrinin on renal function and structure. Toxicology, 16:
    123-137.

    POLLACK, E.S. & HORM, J.W. (1980) Trends in cancer incidence and
    mortality in the United States, 1969-76. J. Natl Cancer Inst., 64:
    1091-1103.

    POMMER, W., GLAESKE, G., & MOLZAHN, M. (1986) The analgesic problem in
    the Federal Republic of Germany: analgesic consumption, frequency of
    analgesic nephropathy and regional differences. Clin. Nephrol., 26:
    273-278.

    PORTER, G.A., ed. (1982) Nephrotoxic mechanisms of drugs and
    environmental toxins, New York, London, Plenum Press.

    PORTER, G.A. (1989) Risk factors for toxic nephropathies. Toxicol.
    Lett., 46: 269-279.

    PORTER, G.A. & BENNETT, W.M. (1989) Drug-induced renal effects of
    cyclosporine, aminoglycoside antibiotics and lithium: extrapolation of
    animal data to man. In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity:
    Extrapolation from  in vitro to  in vivo, and animals to man, New
    York, London, Plenum Press, pp. 147-170.

    PORTER, N.A., WOLF, R.A., & WEENEN, H. (1980) The free radical
    oxidation of polyunsaturated lecithins. Lipids, 15: 163-167.

    POTTER, C.L., GANDOLFI, A.J., NAGLE, R.B., & CLAYTON, J.W. (1981)
    Effects of inhaled chlorotrifluoroethylene and and hexafluoropropene
    on the rat kidney. Toxicol. appl. Pharmacol., 59: 431-440.

    POUNDS, J.G. (1984) Effect of lead intoxication on calcium-mediated
    cell function: a review. Neurotoxicology, 5: 295-332.

    POUNDS, J.G. & ROSEN, J.F. (1988) Cellular Ca2+ Homeostasis and
    Ca2+-mediated cell processes as critical targets for toxicant
    action: con-ceptual and methodological pitfalls. Toxicol. appl.
    Pharmacol., 94: 331-341.

    POUR, P., ALTHOFF, J., SALMASI, S.Z., & STEPAN, K. (1979) Spontaneous
    tumors and common diseases in three types of hamsters. J. Natl Cancer
    Inst., 63: 797-811.

    POWELL, W.S. (1980) Distribution of prostaglandin - hydroxylases in
    different tissues. Prostaglandins, 19: 701-710.

    POWLES, R.L., BARRETT, A.J., CLINK, H., KAY, H.E.M., SLOANE, J., &
    MCELWAIN, T.J. (1978) Cyclosporin A for the treatment of
    graft-versus-host disease in man. Lancet, 2: 1327-1331.

    PRESCOTT, L.F. (1982) Analgesic nephropathy: a reassessment of the
    role of phenacetin and other analgesics. Drugs, 23: 75-149.

    PRETLOW, T.G., II & PRETLOW, T.P., ed. (1982) Cell separation: Methods
    and selected applications, New York, London, San Francisco, Academic
    Press, Vol. 1.

    PRETLOW, T.G., II & PRETLOW, T.P., ed. (1983) Cell separation: Methods
    and selected applications, New York, London, San Francisco, Academic
    Press, Vol. 2.

    PRETLOW, T.G., II & PRETLOW, T.P., ed. (1984) Cell separation: methods
    and selected applications, New York, London, San Francisco, Academic
    Press, Vol. 3.

    PRICE, R.G. (1982) Urinary enzymes, nephrotoxicity and renal disease.
    Toxicology, 23: 99-134.

    PRUEKSARITANONT, T., CHEN, M.-L., & CHIOU, W.L. (1984) Simple and
    micro high-performance liquid chromatographic method for simultaneous
    analysis of p-aminohippuric acid and iothalamate in biological fluids.
    J. Chromatogr., 306: 89-97.

    RAHIMI, A., EDMONDSON, R.P.S., & JONES, N.F. (1981) Effect of
    radiocontrast media on kidney of patients with renal disease. Br. Med.
    J., 282: 1194-1195.

    RAKIETEN, N., GORDON, B.S., COONEY, D.A., DAVIS, R.D., & SCHEIN, P.S.
    (1968) Renal tumorigenic action of streptozotocin (NSC-85998) in rats.
    Cancer Chemother. Rep., 52: 563-567.

    RAMSAMMY, L.S., JOSEPOVITZ, C., LING, K.-Y., LANE, B.P., &
    KALOYANIDES, G.J. (1986) Effects of diphenyl-phenylenediamine on
    gentamicin-induced lipid peroxidation and toxicity in rat renal
    cortex. J. Pharmacol. exp. Ther., 238: 83-88.

    RAMSAMMY, L.S., JOSEPOVITZ, C., LING, K.-Y., LANE, B.P., &
    KALOYANIDES, G.J. (1987) Failure of inhibition of lipid peroxidation
    by vitamin E to protect against gentamicin nephrotoxicity in the rat.
    Biochem. Pharmacol., 36: 2125-2132.

    RAMSAMMY, L.S., JOSEPOVITZ, C., & KALOYANIDES, G.J. (1988a) Gentamicin
    inhibits agonist stimulation of the phosphatidylinositol cascade in
    primary cultures of rabbit proximal tubular cells and in rat renal
    cortex. J. Pharmacol. exp. Ther., 247: 989-996.

    RAMSAMMY, L.S., JOSEPOVITZ, C., LANE, B., & KALOYANIDES, G.J. (1988b)
    Failure of hydroxyl radical scavengers to protect against
    gentamicin-induced acute renal failure in the rat. Fed. Am. Soc. Exp.
    Biol. J., 2: 407A.

    RAMSAMMY, L.S., JOSEPOVITZ, C., LANE, B., & KALOYANIDES, G.J. (1989a)
    Effect of gentamicin on phospholipid metabolism in cultured rabbit
    proximal tubular cells. Am. J. Physiol., 256: C204-C213.

    RAMSAMMY, L.S., JOSEPOVITZ, C., LANE, B.P., & KALOYANIDES, G.J.
    (1989b) Polyaspartic acid protects against gentamicin nephrotoxicity
    in the rat. J. Pharmacol. exp. Ther., 250: 149-153.

    RANSLEY, P.G. & RISDON, R.A. (1979) The pathogenesis of reflux
    nephropathy. Contrib. Nephrol., 16: 90-97.

    RAPP, N.S., ZENSER, T.V., BROWN, W.W., & DAVIS, B.B. (1980) Metabolism
    of benzidine by a prostaglandin-mediated process in renal inner
    medullary slices. J. Pharmacol. exp. Ther., 215: 401-406.

    RASMUSSEN, H.H. & IBELS, L.S. (1982) Acute renal failure: multivarite
    analysis of causes and risk factors. Am. J. Med., 73: 211-218.

    RAVNSKOV, U. (1985) Possible mechanism of hydrocarbon-associated
    glomerulo-nephritis. Clin. Nephrol., 23: 294-298.

    RECKNAGEL, R.O. (1983) Carbon tetrachloride hepatotoxicity: status quo
    and future prospects. Trends Pharmacol. Sci., March: 129-131.

    REDDY, M.V., GUPTA, R.C., RANDERATH, E., & RANDERATH, K. (1984)
    32P-Postlabeling  test for covalent DNA binding of chemicals  in
    vivo: Application to a variety of aromatic carcinogens and methylating
    agents. Carcinogenesis, 5: 231-243.

    REED, D.J. & BEATTY, P.W. (1980) Biosynthesis and regulation of
    glutathione: toxicological implications. In: Hodgson, E., Bend, J.R.,
    & Philpot, R.M., ed. Reviews in biochemical toxicology, Amsterdam,
    Oxford, New York, Elsevier Science Publishers, Vol. 2, pp. 213-241.

    REED, J.R., WILLIAMS R.H., & LUKE, R.G. (1983) The renal hemodynamic
    response to diatrizoate in normal and diabetic rats. Invest. Radiol.,
    18: 536-540.

    REICHERT, D., SPENGLER, V., ROMEN, W., & HENSCHLER, D. (1984)
    Carcinogenicity of dichloroacetylene: an inhalation study.
    Carcinogenesis, 5: 1411-1420.

    RENNICK, G. (1978) Renal excretion of cationic drugs - Tubule
    transport and metabolism. In: Fillastre, J.P., ed. Nephrotoxicity
    interaction of drugs with membrane systems, mitochondria-lysosomes,
    New York, Masson Publishing Co., pp. 1-41.

    REZNIK, G., WARD, J.M., HARDISTY, J.F., & RUSSFIELD, A. (1979) Renal
    carcinogenic and nephrotoxic effects of the flame retardant
    tris(2,3-dibromopropyl) phosphate in F344 rats and (C57BL/6N X
    C3H/HeN)F1 mice. J. Natl Cancer Inst., 63: 205-212.

    RICHMAN, A.V., MASCO, H.L., RIFKIN, S.I., & ACHARYA, M.K. (1980)
    Minimal-change disease and the nephrotic syndrome associated with
    lithium therapy. Ann. intern. Med., 92: 70-72.

    RIMPELA, A.H. & PUKKALA, E.I. (1987) Cancers of affluence: positive
    social class gradient and rising incidence trend in some cancer forms.
    Soc. Sci. Med., 24: 601-606.

    RITCHIE, A.W.S., KEMP, I.W., & CHISHOLM, G.D. (1984) Is the incidence
    of renal carcinoma increasing? Br. J. Urol., 56: 571-573.

    ROBBINS, M.E.C., CAMPLING, D., WHITEHOUSE, E., HOPEWELL, J.W., &
    MICHALOWSKI, A. (1990) Cisplatin-induced reductions in renal reserve
    uncovered by unilateral nephrectomy: an experimental study in the pig.
    Cancer Chemother. Pharmacol., 27: 211-218.

    ROCH-RAMEL, F. & PETERS, G. (1979) Micropuncture techniques as a tool
    in renal pharmacology. Annu. Rev. Pharmacol. Toxicol., 19: 323-345.

    ROELS, H., LAUWERYS, R., BUCHET, J.P., BERNARD, A., CHETTLE, D.R.,
    HARVEY, T.C., & AL HADDAD, I.K. (1981a) In vivo measurement of liver
    and kidney cadmium in workers exposed to this metal. Environ. Res.,
    26: 217-240.

    ROELS, H., LAUWERYS, R., BUCHET, J.P., & BERNARD, A. (1981b)
    Environmental exposure to cadmium and renal function of aged women in
    three areas of Belgium. Environ. Res., 24: 117-130.

    ROELS, H., DJUBGANG, J., BUCHET, J.P., BERNARD, A., & LAUWERYS, R.
    (1982) Evolution of cadmium-induced renal dysfunction in workers
    removed from exposure. Scand. J. Work Environ. Health, 8: 191-200.

    ROELS, H., GENNART, J.P., LAUWERYS, R., BUCHET, J.P., MALCHAIRE, J.,
    & BERNARD, A. (1985) Surveillance of workers exposed to mercury
    vapour: validation of a previously proposed biological threshold limit
    value for mercury concentration in urine. Am. J. ind. Med., 7: 45-71.

    ROELS, H., LAUWERYS, R., BUCHET, J.P., BERNARD, A., VOS, A., &
    OVERSTEYNS, M. (1989) A prospective study of proteinuria in cadmium
    workers. In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity:
    Extrapolation from  in vitro to  in vivo, and animals to man, New
    York, London, Plenum Press, pp. 33-36.

    ROMAN-FRANCO,  A.A., TWIRELLO, M.,  ALBINI, B., OSSI,  E., MILGROM,
    F., & ANDRES, G.A. (1978) Anti-basement membrane antibodies and
    antigen-antibody complexes in rabbits injected with mercuric chloride.
    Clin. Immunol. Immunopathol., 9: 404-481.

    ROSENBERG, M.R. & MICHALOPOULOS, G. (1987) Kidney proximal tubular
    cells isolated by collagenase perfusion grow in defined media in the
    absence of growth factors. J. cell Physiol., 131: 107-113.

    ROSENBERG, M.R., NOVICKI, D.L., JIRTLE, R.L., NOVOTNY, A., &
    MICHALOPOULOS, G. (1985) Promoting effect of nicotinamide on the
    development of renal tubular cell tumours in rats initiated with
    diethylnitrosamine. Cancer Res., 45: 809-814.

    ROSENMAN, K.D., VALCIUKAS, J.A., GLICKMAN, L., MYERS, B.R., & CINOTTI,
    A. (1986) Sensitive indicators of inorganic mercury toxicity. Arch.
    environ. Health, 41: 208-215.

    ROSENMANN, E., DISHON, T., DURST, A., & BOSS, J.H. (1971) Urinary
    excretion of kidney antigens in experimental renal disease of the rat.
    Br. J. exp. Pathol., 52: 388-394.

    ROSS, B., TANGE, J., EMSLIE, K., HART, S., SMAIL, M., & CALDER, I.
    (1980) Paracetamol metabolism by the isolated perfused rat kidney.
    Kidney Int., 18: 562-570.

    ROSSOF, A.H., SLAYTON, R.E., & PERLIA, C.P. (1972) Preliminary
    clinical experience with cis-diamminedichloroplatinum (NSC-119875
    CACP). Cancer, 30: 1451-1456.

    ROXE, D.M. (1980) Toxic nephropathy from diagnostic and therapeutic
    agents. Am. J. Med., 69: 759-766.

    RUSH, G.F., SMITH, J.H., NEWTON, J.F., & HOOK, J.B. (1984) Chemically
    induced nephrotoxicity: role of metabolic activation. CRC crit. Rev.
    Toxicol., 13: 99-160.

    RYFFEL, B., DONATSCH, P., MADORIN, M., MATTER, B.E., RUTTIMAN, G.,
    SCHON, H., STOLL, R., & WILSON, J. (1983) Toxicological evaluation of
    cyclosporin A. Arch. Toxicol., 53: 107-141.

    RYFFEL, B., HIESTAND, P., FOXWELL, B., DONATSCH, P., BOELSTERLI, H.J.,
    MAURER, G., & MIHATSCH, M.J. (1986) Nephrotoxic and immunosuppressive
    potentials of cyclosporine metabolites in rats. Transplant. Proc.,
    18(Suppl. 5): 41-45.

    SAFIRSTEIN, R., WINSTON, J., GOLDSTEIN, M., MOEL, D., DICKMAN, S., &
    GUTTENPLAN, J. (1986) Cisplatin nephrotoxicity. Am. J. kidney Dis., 8:
    356-367.

    SAITO, H., SHIOJI, R., HURUKAWA, Y., NAGAI, K., ARIKEWA, T., SAITO,
    T., SASAKI, Y., FURUYAMA, T., & YOSHINAGA, K. (1977) Cadmium-induced
    proximal tubular dysfunction in a cadmium polluted area. Contrib.
    Nephrol., 6: 1-12.

    SANDLER, D.P. (1987) Epidemiology in the assessment of nephrotoxicity
    In: Bach, P.H. & Lock, E.A., ed. Nephrotoxicity in the experimental
    and clinical situation. Part 2, Dordrecht, Boston, Lancaster, Martinus
    Nijhoff Publishers, pp. 847-883.

    SANDLER, D.P., SMITH, J.C., WEINBERG, C.R., BUCKALEW, V.M., DENNIS,
    V.W., BLYTHE, W.B., & BURGESS, W.P. (1989) Analgesic use and chronic
    renal disease. New Engl. J. Med., 320: 1238-1243.

    SANDSTEAD, H.H., MICHELAKIS, A.M., & TEMPLE, T.E. (1970) Lead
    intoxication. Its effect on the renin-aldosterone response to sodium
    deprivation. Arch. environ. Health, 20: 356-363.

    SASTRASINH, M., KNAUSS, T.C., WEINBERG, J.M., & HUMES, H.D. (1982)
    Identification of the aminoglycoside binding site in rat renal brush
    border membranes. J. Pharmacol. exp. Ther., 222: 350-358.

    SATO, G. & REID, L. (1978) Replacement of serum in cell culture by
    hormones. Int. Rev. Biochem., 20: 219-251.

    SCHENTAG, J.J. (1983) Specificity of renal tubular damage criteria for
    aminoglycoside nephrotoxicity in critically ill patients. J. clin.
    Pharmacol., 23: 473-483.

    SCHENTAG, J.J., SUFTEN, T.A., & PLAUT, M.E. (1978) Early detection of
    aminoglycoside nephrotoxicity with urinary beta-2-microglobulin. J.
    Med., 9: 201-210.

    SCHERBERICH, J.E., FALKENBERG, F., MONDORF, W., PFLEIDERER, G., &
    SCHEOPPE, W. (1976) Isolation of kidney tissue antigens from urine by
    biospecific immunosorbents. Verh. Dtsch. Ges. Inn. Med., 82:
    1469-1473.

    SCHERBERICH, J.E., MONDORF, W., FALKENBERG, F.W., PIERARD, D., &
    SCHOEPPE, W. (1984) Monitoring drug nephrotoxicity. Quantitative
    estimation of human kidney brush border antigens in urine as a
    specific marker of tubular damage. Contrib. Nephrol., 42: 81-92.

    SCHMIDT, U. & GUDER, W.G. (1976) Sites of enzyme activity along the
    nephron. Kidney Int., 9: 233-242.

    SCHOLER, D.W. & EDELMAN, I.S. (1979) Isolation of rat kidney cortical
    tubules enriched in proximal and distal segments. Am. J. Physiol.,
    237: F350-F359.

    SCHRIER, R.W. & GOTTSCHALK, C.W. (1987) Diseases of the kidney,
    Boston, Little, Brown and Co.

    SCHULTZ, S.G., LAVELLE K.J., & SWAIN, R. (1982) Nephrotoxicity of
    radiocontrast media in ischemia renal failure in rabbits. Nephron, 32:
    113-117.

    SCHUMANN, G.B., LERNER, S.I., WEISS, M.A., GAWRONSKI, L., & LOHIYA,
    G.K. (1980) Inclusion-bearing cells in industrial workers exposed to
    lead. Am. J. clin. Med., 74: 192-196.

    SCHWARTZ, S.L., JOHNSON, C.B., & DOOLAN, P.D. (1970) Study of the
    mechanism of renal vasculogenesis induced in the rat by EDTA.
    Comparison of the cellular activities of calcium and chromium. Mol.
    Pharmacol., 6: 54-59.

    SCHWARZ, A. (1987) Analgesic-associated nephropathy. Klin.
    Wochenschr., 65: 1-16.

    SCHWERTZ, D.W., KREISBERG, J.I., & VENKATACHALEM, M.A.  (1984) Effects
    of aminoglycosides on proximal tubule brush border membrane
    phosphatidylinositol-specific phospholipase C. J. Pharmacol. exp.
    Ther., 231: 48-55.

    SELDIN, D.D. & GIEBISCH, G. (1985) The kidney, physiology and
    pathophysiology, New York, Raven Press, Vol. 1.

    SELEVAN, S.G., LANDRIGAN, P.J., STERN, F.B., & JONES, J.H. (1985)
    Mortality of lead smelter workers. Am. J. Epidemiol., 122: 673-683.

    SELLI, C., HINSHAW, W.M., WOODARD, B.H., & PAULSON, D.F. (1983)
    Stratification of risk factors in renal cell carcinoma. Cancer, 52:
    899-903.

    SHINOHARA, Y., ARAI, M., HIRAO, K., SUGIHARA, S., NAKANISHI, K.,
    TSUNODA, H., & ITO, N. (1976) Combination effect of citrinin and other
    chemicals on rat kidney tumorigenesis. Gann, 67: 147-155.

    SHIRACHI, D.Y., JOHANSEN, M.G., MCGOWAN, J.P., & TU, S.-H. (1983)
    Tumorigenic effect of sodium arsenite in rat kidney. Proc. West.
    Pharmacol. Soc., 26: 413-415.

    SHIRAI, T., OHSHIMA, M., MASUDA, A., TAMANO, S., & ITO, N. (1984)
    Promotion of 2-(ethylnitrosamino)ethanol-induced renal carcinogenesis
    in rats by nephrotoxic compounds: positive responses with folic acid,
    basic lead acetate, and N-(3,5-dichlorophenyl)succinimide but not with
    2,3-dibromo-1-propanol phosphate. J. Natl Cancer Inst., 72: 477-482.

    SHIROISHI, K., KJELLSTROM, T., KUBOTA, K., EVRIN, P.E., ANAYAMA, M.,
    VESTERBERG, O., SHIMADA, T., PISCATOR, M., IWATA, T., & NISHINO, H.
    (1977) Urine analysis for detection of cadmium-induced renal changes,
    with special reference to beta2-microglobulin. Environ. Res., 13:
    407-424.

    SHORT, B.G., BURNETT, V.L., & SWENBERG, J.A. (1986) Histopathology and
    cell proliferation induced by 2,2,4-trimethylpentane in the male rat
    kidney. Toxicol. Pathol., 14: 194-203.

    SIEGERS C.-P. & KLAASSEN, C.D. (1984) Biliary excretion of
    acetaminophen in ureter-ligated rats. Pharmacology, 28: 177-180.

    SIERSBAEK-NIELSON, K., MOLHOLM HANSEN, J.M., KAMPMANN, J., &
    KRISTENSEN, M. (1971) Rapid evaluation of creatinine clearance.
    Lancet, 1: 1133-1134.

    SIKIC, B.I., MIMNAUGH, E.G., LITTERST, C.L., & GRAM, T.E. (1977) The
    effects of ascorbic acid deficiency and repletion on pulmonary, renal
    and hepatic drug metabolism in the guinea-pig. Arch. Biochem.
    Biophys., 179: 663-671.

    SILVERBLATT, F.J. & KUEHN, C. (1979) Autoradiography of gentamicin
    uptake by the rat proximal tubule cell. Kidney Int., 15: 335-345.

    SMITH, H.J., LEVORSTAD, K., BERG, K.J., ROOTWELT, K., & SVEEN, K.
    (1985) High dose urography in patients with renal failure. A double
    blind investigation of iohexol and metrizoate. Acta radiol. Diagn.,
    26: 213-220.

    SMITH, J.H. & HOOK, J.B. (1984) Mechanism of chloroform
    nephrotoxicity. III. Renal and hepatic microsomal metabolism of
    chloroform in mice. Toxicol. appl. Pharmacol., 73: 511-524.

    SMITH, J.H., MAITA, K., SLEIGHT, S.D., & HOOK, J.B. (1983) Mechanism
    of chloroform nephrotoxicity. I. Time course of chloroform toxicity in
    male and female mice. Toxicol. appl. Pharmacol., 70: 467-479.

    SMITH, J.H., MAITA, K., SLEIGHT, S.D., & HOOK, J.B. (1984) Effect of
    sex hormone status on chloroform nephrotoxicity and renal mixed
    function oxidases in mice. Toxicology, 30: 305-316.

    SMITH, J.H., RUSH, G.F., & HOOK, J.B. (1986) Induction of renal and
    hepatic mixed function oxidases in the hamster and guinea-pig.
    Toxicology, 38: 209-218.

    SMITH, R.L. (1974) The excretory function of bile: the elimination of
    drugs and toxic substances in bile - factors affecting biliary
    excretion, London, Chapman & Hall, pp. 16-34.

    SNYDER, R., PARKE, D., KOCSIS, J.J., JOLLOW, D.J., & GIBSON, G.G., ed.
    (1981) Biological reactive intermediates. Part II, New York, London,
    Plenum Press.

    SOBERON, L., BOWMAN, R.L., PASTORIZA-MUNOZ, E., & KALOYANIDES, G.J.
    (1979) Comparative nephrotoxicities of gentamicin, netilmicin and
    tobramycin in the rat. J. Pharmacol. exp. Ther., 210: 334-343.

    SOOSE, M., HABERSTROH, U., ROVIRA-HOLBACH, G., BRUNKHORST, R., &
    STOLTE, H. (1988) Heterogeneity of glomerular barrier function in
    early adriamycin nephrosis of MWF rats. Clin. Physiol. Biochem., 6:
    310-315.

    SPATARO, R. (1984) New contrast agents for urography. Radiol. Clin.
    North Am., 22: 365-380.

    SRAER, J., FOIDART, J., CHANSEL, D., MAHIEU, P., & ARDAILLOU, R.
    (1980) Prostaglandin synthesis by rat isolated glomeruli and
    glomerular cultured cells. Int. J. Biochem., 12: 203-207.

    STERNBERG, S.S., PHILIPS, F.S., & CRONIN, A.P. (1972) Renal tumors and
    other lesions in rats following a single intravenous injection of
    Daunomycin. Cancer Res., 32: 1029-1036.

    STOLTE, H. & ALT, J., ed. (1980) Research animals and experimental
    design in nephrology. Contrib. Nephrol., 19: 1-249.

    STOLTE, H. & ALT, J. (1982) The choice of animals for nephrotoxic
    investigations. In: Bach, P.H., Bonner, F.W., Bridges, J.W., & Lock,
    E.A., ed. Nephrotoxicity: Assessment and pathogenesis, New York,
    Chichester, Brisbane, Toronto, John Wiley and Sons, pp. 102-112.

    STONARD, M.D., CHATER, B.V., DUFFIELD, D.P., NEVITT, A.L., O'SULLIVAN,
    J.J., & STEEL, G.T. (1983) An evaluation of renal function in workers
    occupationally exposed to mercury vapour. Int. Arch. occup. environ.
    Health, 52: 177-189.

    STONARD, M.D., PHILLIPS, P.G.N., FOSTER, J.R., SIMPSON, M.G. & LOCK,
    E.A. (1986) Alpha-2u-globulin: Measurement in rat kidney following
    administration of 2,2,4-trimethylpentane. Toxicology, 41: 161-168.

    STRIKER, G.E., KILLEN, P.D., & FARIN, F.M. (1980) Human glomerular
    cells  in vitro: Isolation and characterization. Transplant. Proc.,
    12: 88-99.

    STUVE, J. & GALLE, P. (1970) Role of mitochondria in the handling of
    gold by the kidney. J. cell. Biol., 44: 667-676.

    SUFRIN, G. & BECKLEY, S.A. (1980) Renal adenocarcinoma, Geneva,
    International Union Against Cancer, pp. 133-155 (Technical Report No.
    49).

    SUN, C.N. (1980) Effect of anti-diuretic hormone to the connective
    tissue of rat renal papilla. Exp. Pathol., 18: 469-473.

    SUN, C.N., WHITE, H.J., & TOWBIN, E.J. (1972) Histochemistry and
    electron microscopy of renal papilla in a genetic strain of rats with
    diabetic insipidus. Nephron, 9: 308-317.

    SUNDERMAN, F.W., Jr (1981) Present research on nickel carcinogenesis.
    Environ. Health Perspect., 40: 131-142.

    SUNDERMAN, F.W., MCCULLY, R.S., & HOPFER, S.M. (1984) Association
    between erythrocytosis and renal cancers in rats following intrarenal
    injection of nickel compounds. Carcinogenesis, 5: 1511-1517.

    SUZUKI, S., SUZUKI, S., NAKAMURA, N., & KOIZUMI, T. (1976) The
    heterogeneity of dermatan sulfate and heparan sulfate in rat liver and
    a shift in the glycosaminoglycan contents in carbon tetrachloride
    damaged liver. Biochim. Biophys. Acta, 428: 166-181.

    SUZUKI, S., SHAPIRO, R., MULROW, P.J., & TAN, S.Y. (1980) Urinary
    prostaglandin E2 excretion in chronic renal disease. Prostaglandins
    Med., 4: 377-382.

    SWANN, P.F., KAUFMAN, D.G., MAGEE, P.N., & MACE, R. (1980) Induction
    of kidney tumours by a single dose of dimethylnitrosamine: dose
    response and influence of diet and benzo(a)pyrene pretreatment. Br. J.
    Cancer, 41: 285-294.

    SWENBERG, J.A., SHORT, B., BORGHOFF, S., STRASSER, J., & CHARBONNEAU,
    J. (1989) The comparative pathobiology of alpha-2u-globulin
    nephropathy. Toxicol. appl. Pharmacol., 97: 35-46.

    SZEFLER, S.J. & ACARA, M. (1979) Isoproterenol excretion and
    metabolism in the isolated perfused rat kidney. J. Pharmacol. exp.
    Ther., 210: 295-300.

    TALIERCIO, C.P., VLIETSTRA, R.E., FISHER, L.D., & BURNETT, J.C. (1986)
    Risks for renal dysfunction with cardiac angiography. Ann. intern.
    Med., 104: 501-504.

    TAMURA, T. & ENDOU, H. (1988) Contribution of purine nucleotide cycle
    to intranephron ammoniagenesis in rats. Am. J. Physiol., 255:
    F1122-F1127.

    TANAKA, H., ISHIKAWA, E., TESHIMA, S., & SHIMUZI, E. (1986)
    Histopathological study of human cisplatin nephropathy. Toxicol.
    Pathol., 14: 247-257.

    TARLOFF, J.B., GOLSTEIN, R.S., & HOOK, J.B. (1987) Xenobiotic
    metabolism in the mammalian kidney. In: Bach, P.H. & Lock, E.A., ed.
    Nephro-toxicity in the experimental and the clinical situation,
    Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers, pp.
    371-404.

    TEIXEIRA, R.B., KELLEY, F., ALPERT, H., PARDOV, V., & VAAMONDE, C.A.
    (1982) Complete protection from gentamicin-induced acute renal failure
    in the untreated streptozotocin diabetes mellitus rat. Kidney Int.,
    21: 600-612.

    TERRACINI, B. & PARKER, V.H. (1965) A pathological study on the
    toxicity of S-(dichlorovinyl)-L-cysteine. Food Cosmet. Toxicol., 3:
    67-74.

    TERUEL, J.L., MARCEN, R., ONAINDIA, J.M., SERRANO, A., QUEREDA, C., &
    ORTUNO, J. (1981) Renal function impairment caused by intravenous
    urography, a prospective study. Arch. intern. Med., 141: 1271-1274.

    TESTA, B. & JENNER, P. (1976) Drug metabolism: chemical and
    biochemical aspects, New York, Basel, Marcel Dekker.

    THOENES, W., STÖRKEL, S., & RUMPELT, H.-J. (1985) Human chromophobe
    cell renal carcinoma. Virchows Arch. B Cell Pathol., 48: 207-217.

    THOMPSON, W.M., FOSTER, W.L., Jr, HALVORSEN, R.A., DUNNICK, N.R.,
    ROMMEL, A., & BATES, M. (1984) Iopamidol: new, nonionic contrast agent
    for excretory urography. Am. J. Roentgenol., 142: 329-332.

    THORNLEY, C., DAWNAY, A., & CATTEL, W.R. (1985) Human Tamm-Horsfall
    glycoprotein: urinary and plasma levels in normal subjects and
    patients with renal disease determined by a fully validated
    radioimmunoassay. Clin. Sci., 68: 529-535.

    THUN, M.J., BAKER, D.B., STEENLAND, K., SMITH, A.B., HALPERIN, W., &
    BERL, T.B. (1985) Renal toxicity in uranium mill workers. Scand. J.
    Work Environ. Health, 11: 83-90.

    THURAU, K. (1964) Renal hemodynamics. Am. J. Med., 36: 698-719.

    TOKOFF-RUBIN, N.E. (1986) Diagnosis of tubular injury in renal
    transplant patients by urinary assay for proximal tubular antigen, the
    adenosine-deaminase-binding protein. Transplantation, 41: 593-599.

    TOLEDO, O.M.S. & DIETRICH, C.P. (1977) Tissue specific distribution of
    sulfated mucopolysaccharides in mammals. Biochim. Biophys. Acta, 498:
    114-122.

    TOLINS, J.P. & RAIJ, L. (1988) Chronic amphotericin B nephrotoxicity
    in the rat; protective effect of prophylactic salt loading. Am. J.
    kidney Dis., 11: 313-317.

    TOUBEAU, G., LAURENT, G., CARLIER, M.-B., ABID, S., MALDAGUE, P.,
    HEUSON-STIENNON, J.A., & TULKENS, P.M. (1986) Tissue repair in rat
    kidney cortex after short treatment with aminoglycosides at low doses.
    Lab. Invest., 54: 385-393.

    TRIFILLIS, A.L., REGEC, A., HALL-CRAGGS, M., & TRUMP, B.F. (1986)
    Effects of cyclosporine on cultured human renal tubular cells.
    Toxicol. Pathol., 14: 210-212.

    TRINH-TRANG-TAN, M.M., BOUBY, N., COUTAUD, C., & BANKIR, L. (1986)
    Quick isolation of rat medullary thick ascending limbs. Enzymatic and
    metabolic characterization. Pflügers Arch., 407: 228-234.

    TRUMP, B.F., BEREZESKY, I.K., SMITH, M.W., PHELPS, P.C., & ELLIGET,
    K.A. (1989) The relationship between cellular ion deregulation and
    acute and chronic toxicity. Toxicol. appl. Pharmacol., 97: 6-22.

    TSUDA, H., SAKATA, T., TAMANO, S., OKUMURA, M., & ITO, N. (1983)
    Sequential observations on the appearance of neoplastic lesions in the
    liver and kidney after treatment with
    N-ethyl-N-hydroxyethylnitrosamine followed by partial hepatectomy and
    unilateral nephrectomy. Carcinogenesis, 4: 523-528.

    TSUDA, H., HACKER, H.J., KATAYAMA, H., MASUI, T., ITO, N., & BANNASCH,
    P. (1986) Correlative histochemical studies on preneoplastic and
    neoplastic lesions in the kidney of rats treated with nitrosamines.
    Virchows Arch. B Cell Pathol., 51: 385-404.

    TSUDA, H., MOORE, M.A., ASAMOTO, M., INOUE, T., FUKUSHIMA, S., ITO,
    N., SATOH, K., AMELIZAD, Z., & OESCH, F. (1987) Immunohistochemically
    demonstrated altered expression of cytochrome P-450 molecular forms
    and epoxide hydrolase in N-ethyl-N-hydroxyethylnitrosamine-induced rat
    kidney and liver lesions. Carcinogenesis, 8: 711-717.

    TUBBS, R.R., GEPHARDT, G.N., MCMAHON, J.T., POHL, M.C., VIDT, D.G.,
    BARENBERG, S.A., & VALENZUELA, R. (1982) Membranous
    glomerulo-nephritis associated with industrial mercury exposure. Am.
    J. clin. Pathol., 77: 409-413.

    TULKENS, P.M. (1989) Nephrotoxicity of aminoglycoside antibiotics.
    Toxicol. Lett., 46: 107-123.

    TUNE, B.M. (1975) Relationship between the transport and toxicity of
    cephalosporine in the kidney. J. infect. Dis., 132: 189-194.

    TUNE, B.M. (1986) The nephrotoxicity of cephalosporin antibiotics -
    structure-activity relationships. Comments Toxicol., 1: 145-170.

    TUNE, B.M. & FRAVERT, D. (1980) Mechanisms of cephalosporin
    nephrotoxicity. A comparison of cephaloridine and cephaloglycin.
    Kidney Int., 18: 591-600.

    TUNE, B.M., BURG, M.D., & PATLAK, C.S. (1969) Characteristics of
    p-aminohippurate transport in proximal renal tubules. Am. J. Physiol.,
    217: 1057-1063.

    TUNE, B.M., SIBLEY, R.K., & HSU, C.-Y. (1988) The mitochondrial
    respiratory toxicity of cephalosporin antibiotics. An inhibitory
    effect on substrate uptake. J. Pharmacol. exp. Therap., 245:
    1054-1059.

    TYSON, C.K. & MIRSALIS, J.C. (1985) Measurement of unscheduled DNA
    synthesis in rat kidney cells following in vivo treatment with
    genotoxic agents. Environ. Mutagen., 7: 889-899.

    UCHIDA, S. & ENDOU, H. (1988) Substrate specificity to maintain
    cellular ATP along the mouse nephron. Am. J. Physiol., 255: F977-F983.

    ULLRICH, K.J. & GREGER, R. (1985) Approaches to the study of tubule
    transport function. In: Seldin, D.G. & Giebisch, G., ed. The kidney,
    physiology and pathology, New York, Raven Press, Chapter 20, pp.
    427-455.

    VAAMONDE, C.A., BIER, R., GOUVEA, W., ALPERT, H., KELLY, J., & PARDO,
    V. (1984) Effect of duration of diabetes on the protection observed in
    the diabetic rat against gentamicin-induced acute renal failure.
    Miner. Electrolyte Metab., 10: 209-216.

    VALTIN, H., ed. (1973) Renal functions: mechanisms preserving fluid
    and solute balance in health, Boston, Toronto, Little Brown and Co.

    VAN DORP, D. (1971) Recent developments in the biosynthesis and the
    analyses of prostaglandins. Ann. NY Acad. Sci., 180: 181-199.

    VAN ESCH, G.J. & KROES, R. (1969) The induction of renal tumours by
    feeding basic lead acetate to mice and hamsters. Br. J. Cancer, 23:
    765-771.

    VAN FLEET, J.F., GREENWOOD, L.A., & FERRANO, F.J. (1979) Pathologic
    features of adriamycin toxicosis in young pigs: non skeletal lesions.
    Am. J. vet. Res., 40: 1537-1552.

    VERPOOTEN, G.F., NOUWEN, E.J., HOYLAERTS, M.F., HENDRIX, P.G., & DE
    BROE, M. (1989) Segment-specific localization of intestinal-type
    alkaline phosphatase in human kidney. Kidney Int., 36: 617-625.

    VERSCHOOR, M., WIBOWO, A., HERBER, R., VAN HEMMEN, J., & ZIELHUIS, R.
    (1987) Influence of occupational low-level lead exposure on renal
    parameters. Am. J. ind. Med., 12: 341-351.

    VIAU, C., BERNARD, A., GUERET, F., MALDAGUE, P., GENGOUX, P., &
    LAUWERYS, R. (1986a) Isoparaffinic solvent-induced nephrotoxicity in
    the rat. Toxicology, 38: 227-240.

    VIAU, C., BERNARD, A., & LAUWERYS, R. (1986b) Determination of rat
    beta2-microglobulin in urine and in serum. I. Development of an
    immunoassay based on latex particles agglutination. J. appl. Toxicol.,
    6: 185-189.

    VIAU, C., BERNARD, C., LAUWERYS, R., QUAGHEBEUR, L., CORNU, M.E.,
    PHILLIPS, S.C., MUTTI, A., LUCERTINI, S., & FRANCHINI, I. (1987) A
    cross-sectional survey of kidney function in refinery employees. Am.
    J. ind. Med., 11: 177-187.

    VICTERY, W., VANER, A.J., SHULAK, J.M., SCHOEPS, P., & JULIUS, S.
    (1982) Lead, hypertension and the renin-angiotension system in rats.
    J. lab. clin. Med., 99: 354-362.

    VIOL, G.W., MINIELLY, J.A., & BISTRICKI, T. (1977) Gold nephropathy.
    Tissue analysis by X-ray fluorescent spectroscopy. Arch. Pathol. lab.
    Med., 101: 635-640.

    VINAY, P., GOUGOUX, A., & LEMIEUX, G. (1981) Isolation of a pure
    suspension of rat proximal tubules. Am. J. Physiol., 241: F403-F411.

    VOGELZANG, N.J., TORKELSON, J.L., & KENNEDY, B.J. (1985)
    Hypo-magnesemia, renal dysfunction and Raynaud's phenomenon in
    patients treated with cisplatin, vinblastine and bleomycin. Cancer,
    56: 2765-2770.

    WALKER, P.D. & SHAH, S.V. (1987) Evidence suggesting a role for
    hydroxyl radical in gentamicin-induced acute renal failure in rats. J.
    clin. Invest., 81: 334-341.

    WALKER, R.G., DAVIES, B.M., HOLWILL, B.J., DOWLING, J.P., &
    KINCAID-SMITH, P. (1982) A clinicopathological study of lithium
    nephrotoxicity. J. chron. Dis., 35: 685-695.

    WALKER, R.G., ESCOTT, M., BIRCHALL, I., DOWLING, J.P., &
    KINCAID-SMITH, P. (1986) Chronic progressive renal lesions induced by
    lithium. Kidney Int., 29: 875-881.

    WALLACE, A.C. & NAIRN, R.C. (1972) Renal tubular antigens in kidney
    tumors. Cancer, 29: 977-981.

    WAN, S.H. & RIEGELMAN, S. (1972a) Renal contribution to overall
    metabolism of drug. I: Conversion of benzoic acid to hippuric acid. J.
    pharm. Sci., 61: 1278-1284.

    WAN, S.H. & RIEGELMAN, S. (1972b) Renal contribution to overall
    metabolism of drugs. II: Biotransformation of salicylic acid to
    salicyluric acid. J. pharm. Sci., 61: 1284-1287.

    WAN, S.H., VON LEHMANN, B., & RIEGELMAN, S. (1972) Renal contribution
    to overall metabolism of drugs. III: Metabolism of p-aminobenzoic
    acid. J. pharm. Sci., 61: 1288-1292.

    WARD, J.M. & FAUVIE, K.A. (1976) The nephrotoxic effects of
    cis-diammine-dichloroplatinum (II) (NSC-119875) in male F344 rats.
    Toxicol. appl. Pharmacol., 38: 535-547.

    WARD, J.M., GOODMAN, D.G., SQUIRE, R.A., CHU, R.C., & LINHART, M.S.
    (1979) Neoplastic and non-neoplastic lesions in aging (C57BL/6N X
    CH3/HeN)F1 (B6C3F11) mice. J. Natl Cancer Inst., 63: 849-854.

    WEBER, P.C. (1980) Renal prostaglandins, kidney function and effect of
    aspirin on metabolism of acetaminophen and benzidine by renal inner
    medulla prostaglandin hydroperoxidase. J. lab. clin. Med., 101: 58-65.

    WEDEEN, R.P., MALLIK, D.K., & BATUMAN, V. (1979) Detection and
    treatment of occupational lead nephropathy. Arch. intern. Med., 139:
    53-57.

    WEINBERG, J.M. & HUMES, H.D. (1980) Mechanisms of gentamicin-induced
    dysfunction of renal cortical mitochondria. I. Effects on
    mitochondrial respiration. Arch. Biochem. Biophys., 205: 222-231.

    WHO (1979) IPCS Environmental Health Criteria 11: Mycotoxins, Geneva,
    World Health Organization, 127 pp.

    WHO COLLABORATING CENTRE FOR THE HISTOLOGICAL CLASSIFICATION OF RENAL
    DISEASES (1982) Renal disease: Classification and atlas of glomerular
    diseases, Tokyo, Igaku-Shoin, 359 pp.

    WHO (1983) IPCS Environmental Health Criteria 27: Guidelines on
    studies in environmental epidemiology, Geneva, World Health
    Organization, pp. 184-185.

    WHO (1984) IPCS Environmental Health Criteria 39: Paraquat and Diquat,
    Geneva, World Health Organization, 181 pp.

    WHO (1990) International classification of diseases for oncology, 2nd
    ed., Geneva, World Health Organization, 144 pp.

    WHO COLLABORATING CENTRE FOR THE HISTOLOGICAL CLASSIFICATION OF RENAL
    DISEASES (1985) Renal disease: Classification and atlas of
    tubulo-interstitial diseases, New York, Igaku-Shoin, 221 pp.

    WHO COLLABORATING CENTRE FOR THE HISTOLOGICAL CLASSIFICATION OF RENAL
    DISEASES (1987) Renal disease: Classification and atlas, New York,
    Igaku-Shoin, 291 pp.

    WHO COLLABORATING CENTRE FOR THE HISTOLOGICAL CLASSIFICATION OF RENAL
    DISEASES (1988) Classification and atlas of infectious and tropical
    diseases, Chicago, American Society of Clinical Pathologists, 273 pp.

    WIBELL, L., EVRIN, P.E., & BERGGARD, I. (1973) Serum
    beta2-microglobulin in renal disease. Nephron, 10: 320-331.

    WILLIAMS, P.D., HOLOHAN, P.D., & ROSS, C.R. (1981) Gentamicin
    nephro-toxicity. II. Plasma membrane changes. Toxicol. appl.
    Pharmacol., 61: 243-251.

    WILLIAMS, P.D., HOTTENDORF, G.H., & BENNETT, D.B. (1986) Inhibition of
    renal membrane binding and nephrotoxicity of aminoglycosides. J.
    Pharmacol. exp. Ther., 237: 919-925.

    WILSON, P.D. (1986) Use of cultured renal tubular cells in the study
    of cell injury. Miner. Electrolyte Metab., 12: 71-83.

    WING, A.J., BRUNNER, F.P., GEERLINGS, W., BROYER, M., BRYNGER, H.,
    FASSBINDER, W., RISSONI, G., SELWOOD, N.H., & TUFESON, G. (1989)
    Contribution of toxic nephropathies to end-stage renal failure in
    Europe: a report form the EDTA-ERA registry. Toxicol. Lett., 46:
    281-292.

    WIRTHENSOHN, G., BECK, F.X., & GUDER, W.G. (1987) Role and regulation
    of glycerophosphorylcholine in rat renal papilla. Pflügers Arch., 409:
    411-415.

    WIRTHENSOHN, G., LEFRANK, S., SCHMOLKE, M., & GUDER, W.G. (1989)
    Regulation of organic osmolyte concentrations in tubules from rat
    renal inner medulla. Am. J. Physiol., 256: F128-F135.

    WITSCHI, H.P., ALDRIDGE, W.N., AUST, S.D., AUTOR, A.P., DE MATTEIS,
    F., DRUET, P., ELBERS, R., FOWLER, B.A., GRONIOWSKI, J.A., HAGMAN, W.,
    ORRENIUS, S., PETERING, D.H., SEINEN, W., SUMMER, K.H., & TRUMP, B.F.
    (1987) Mechanisms and target cell injury. In: Fowler, B.A., ed.
    Mehanisms of cell injury: Implications for human health, New York,
    Chichester, Brisbane, Toronto, John Wiley and Sons, pp. 385-403.

    WOLD, J.S., TURNIPSEED, S.A., & MILLER, B.L. (1979) The effect of
    renal cation transport inhibition on cephaloridine nephrotoxicity.
    Toxicol. appl. Pharmacol., 47: 115-122.

    WOLGAST, M., PERSSON, E., SCHNERMANN, J., ULFENDAHL, H., & WUNDERLICH,
    P. (1973) Colloid osmotic pressure of the subcapsular interstitial
    fluid of rat kidneys during hydropenia and volume expansion. Pflügers
    Arch., 340: 123-131.

    YOUNG, D.M. (1975) Pathologic effects of adriamycin in experimental
    systems. Cancer Chemother. Rep., 6: 159-175.

    ZENSER, T.V. & DAVIS, B.B. (1984) Enzyme systems involved in the
    formation of reactive metabolites in the renal medulla: cooxidation
    via prostaglandin H synthase. Fundam. appl. Toxicol., 4: 922-929.

    ZENSER, T.V., MATTAMMAL, M.B., & DAVIS, B.B. (1978a) Differential
    distribution of the mixed function oxidase in rabbit kidney. J.
    Pharmacol. exp. Ther., 207: 719-725.

    ZENSER, T.V., MATTAMMAL, M.B., HERMAN, C.A., JOSHI, S., & DAVIS, B.B.
    (1978b) Effect of acetaminophen on prostaglandin E2 and
    prostaglandin F2 alpha synthesis in the renal inner medulla of rat.
    Biochim. Biophys. Acta, 542: 486-495.

    ZENSER, T.V., MATTAMMAL, M.B., & DAVIS, B.B. (1979a) Demonstration of
    separate pathways for the metabolism of organic compounds in rabbit
    kidney. J. Pharmacol. exp. Ther., 208: 418-421.

    ZENSER, T.V., MATTAMMAL, M.B., & DAVIS, B.B. (1979b) Cooxidation of
    benzidine by renal medullary prostaglandin cyclooxygenase. J.
    Pharmacol. exp. Ther., 211: 460-464.

    ZENSER, T.V., MATTAMMAL, M.B., BROWN, W.W., & DAVIS, B.B. (1979c)
    Cooxygenation by prostaglandin cyclooxygenase from rabbit inner
    medulla. Kidney Int., 16: 688-694.

    ZENSER, T.V., MATTAMMAL, M.B., RAPP, N.S., & DAVIS, B.B. (1983) Effect
    of aspirin on metabolism of acetaminophen and benzidine by renal inner
    medulla prostaglandin hydroperoxidase. J. lab. clin. Med., 101: 58-65.

    ZIMMERMAN, S.W. & NORBACH, D.H. (1980) Nephrotoxic effects of
    long-term carbon tetrachloride administration in rats. Arch. Pathol.
    lab. Med., 104: 94-99.

    ZUSMAN, R.M. (1980) Prostaglandin E2 biosynthesis by renomedullary
    inter-stitial cells:  In vitro studies and pathophysiological
    correlations. In: Mandal, A.K. & Bohman, S.-O., ed. The renal papilla
    and hypertension, New York, London, Plenum Press, pp. 187-207.

    RESUME ET CONCLUSIONS

         L'excrétion et l'homéostase des molécules hydrosolubles sont
    principalement assurées par les reins.  Ceuxci sont constitués
    d'unités fonctionnelles appelées néphrons, qui consistent
    essentiellement en un tube continu formé de cellules hétérogènes
    hautement spécialisées qui présentent une organisation structurale,
    fonctionnelle et biochimique poussée.  Il existe des différences
    marquées entre les néphrons eux-mêmes, qui dépendent de la
    localisation plus ou moins profonde du glomérule dans le cortex. 
    C'est l'agencement complexe de cette structure qui, joint à des
    différences régionales de vascularisation tenant à la disposition
    particulière des vaisseaux, constitue cet organe hétérogène et
    hautement organisé qu'est le rein.

         Une grande partie des connaissances sur l'anatomie et le
    fonctionnement du rein, obtenues grâce à l'expérimentation animale,
    est directement transposable à l'homme.  Toutefois, les processus
    biologiques et métaboliques qui se déroulent dans le rein humain ne
    sont pas encore complètement élucidés, non plus que les différences
    constatées d'une espèce animale à une autre.  Dans ces conditions, on
    comprend qu'on ne puisse pas toujours extrapoler d'une espèce à une
    autre les effets toxicologiques observés.

         Plusieurs substances (qu'elles aient ou non un effet
    thérapeutique) exercent des effets toxiques sur les divers éléments
    anatomiques du rein.  Les effets toxiques peuvent être aigus ou
    chroniques; ils peuvent être directs ou se produire par
    l'intermédiaire de mécanismes immunologiques.  L'impact qu'une
    substance néphrotoxique peut avoir sur la santé dépend de plusieurs
    facteurs de risque: état fonctionnel du rein, lésion préexistante,
    maladie, âge, sexe, régime alimentaire, etc..

         On n'a pas suffisamment étudié l'épidémiologie des cas de
    néphrotoxicité imputables à des diverses substances chimiques agissant
    seules ou en association.  A quelques exceptions près, la contribution
    des produits chimiques à l'incidence globale des néphropathies et de
    l'insuffisance rénale chronique reste mal définie.  D'importantes
    recherches ont été consacrées à certains groupes exposés de par leur
    profession ainsi qu'aux néphropathies résultant de la consommation
    d'analgésiques : elles ont montré l'existence de variations entre les
    différents groupes et les différents pays.  Toutefois on estime que
    jusqu'à 18% des néphropathies terminales pourraient être dues à la
    prise d'analgésiques et jusqu'à 5% avoir une origine toxique
    quelconque.  Environ 50% des néphropathies terminales sont d'origine
    inconnue.  L'un des principaux problèmes qui se posent dans la
    recherche de l'étiologie des néphropathies terminales tient à la
    longue période de latence ou à l'évolution lente de l'insuffisance
    rénale chronique, qui rendent difficile l'identification rétrospective
    de l'agent causal.  Ce n'est que dans quelques cas que le dosage de
    certaines substances dans les tissus (organisme, rein) peut conduire
    au diagnostic.  Par ailleurs on se heurte également à un certain

    manque de cohérence dans les critères et la terminologie utilisés sur
    le plan diagnostique et anatomopathologique.

         Les différences anatomiques et physiologiques rendent difficile
    une extrapolation directe à l'homme des résultats obtenus sur des
    systèmes expérimentaux ( in vivo ou  in vitro).  On ne connaît que
    très peu de cas où les données de néphrotoxicité sont comparables chez
    l'homme et chez l'animal et permettent d'évaluer de façon fiable le
    risque pour la santé humaine.

         Certaines substances chimiques peuvent provoquer des lésions
    sélectives au niveau des structures rénales vulnérables ou déclencher
    des mécanismes immunologiques.  On considère qu'il existe en général
    deux types de mécanismes à la base des lésions rénales : a) les
    mécanismes immunologiques qui conduisent à la néphrite aiguë
    interstitielle et b) les mécanismes qui affectent principalement le
    glomérule par l'intermédiaire d'anticorps anti-GBM ou
    d'immunocomplexes.  Il existe un autre grand groupe de maladies dues
    à des substances chimiques ou à leurs métabolites qui perturbent
    l'activité biochimique cellulaire ou modifient les propriétés
    hémodynamiques, etc.  Parmi les facteurs susceptibles d'influer sur
    les lésions cellulaires provoquées par les produits toxiques figurent
    les systèmes de transport cellulaire, la pinocytose, la dégradation
    métabolique, l'interaction avec les protéines, les lipides, les
    membranes et l'ADN cellulaires et éventuellement avec d'autres
    constituants de la cellule.

         Plus on utilise de produits pharmaceutiques et de substances
    chimiques, plus on accroît les risques de néphrotoxicité.  La
    néphrotoxicité des médicaments dépend de la dose et de la durée du
    traitement (par exemple dans le cas des associations d'analgésiques
    qui conduisent à une nécrose papillaire).  On a beaucoup étudié les
    effets néphrotoxiques des analgésiques, des antibiotiques (comme les
    aminoglycosides), des agents anticancéreux (comme le cis-platine) et
    de divers autres agents.  Nombre de substances fréquemment utilisées
    dans l'industrie ou à la maison comme les hydrocarbures chlorés ou
    l'éthylèneglycol peuvent également avoir une action toxique sur le
    rein.  D'autres substances, présentes dans l'environnement comme le
    plomb ou le cadmium sont effectivement néphrotoxiques.  L'action
    toxique de ces substances intervient après accumulation
    intracellulaire du produit initial ou métabolisation intra- ou
    extrarénale.  L'exposition à plusieurs substances chimiques peut
    entraîner des réactions antagonistes ou synergistiques.

         Les tumeurs du rein et des voies urinaires représentent moins de
    2 à 3 % de l'ensemble des cancers humains, mais leur fréquence est en
    augmentation, ce qui incite à penser que des facteurs environnementaux
    pourraient être à l'oeuvre.  Nombre de produits pharmaceutiques et de
    substances chimiques se sont révélés cancérogènes sur des modèles
    expérimentaux mais seuls quelques-uns d'entre eux sont responsables de
    l'apparition de cancers chez l'homme.  Il s'agit de l'amiante

    (adénocarcinome rénal), de l'abus d'analgésiques (carcinomes de type
    transitionnel) et de facteurs tenant au mode de vie (tabagisme,
    consommation d'alcool et de café).  Il existe des groupes de
    population chez lesquels certains cancers des voies urinaires ont une
    étiologie encore inconnue.  Certaines substances chimiques utilisées
    dans le cadre professionnel sont à l'origine de cancers de la vessie. 
    De nombreuses substances chimiques ou pharmaceutiques peuvent
    entraîner une néphrite interstitielle, affection qui serait
    susceptible de favoriser l'apparition de cancers des voies urinaires.

         Il n'existe pas de méthode qui permette d'évaluer in vivo ou in
    vitro dans des conditions satisfaisantes une insuffisance rénale
    provoquée par des substances chimiques.  Aussi faut-il non seulement
    contrôler l'activité néphrotoxique des produits chimiques, mais
    également inclure dans les protocoles expérimentaux l'étude du
    mécanisme toxique au niveau des cellules cibles.  Pour y parvenir il
    faut faire appel tant à l'expérimentation in vivo qu'à des études in
    vitro.  Les études in vitro doivent porter sur des éléments où les
    relations anatomiques et cellulaires sont maintenues (par
    microperfusion de reins et de tubules isolés, coupes de cortex,
    fragments isolés de néphrons, etc.) et sur des cellules rénales (par
    exemple suspensions cellulaires, cultures cellulaires primaires,
    lignées cellulaires et fractions infracellulaires).  Pour les études
    in vivo, on peut faire appel à des techniques effractives ou non
    effractives.  Parmi les techniques effractives figurent les examens
    histopathologiques ainsi que les techniques classiques de bilan de la
    fonction rénale.  Les techniques non effractives permettent un bilan
    répété de la fonction rénale chez l'animal par la mesure d'une série
    de paramètres fonctionnels (filtration glomérulaire, fonction
    tubulaire, protéinurie, enzymurie, etc.).  Des épreuves biochimiques
    spécialisées devront être effectuées le cas échéant.  En associant de
    façon convenable les épreuves in vivo et les épreuves in vitro on
    pourra déterminer si telle ou telle substance chimique est
    néphrotoxique et avoir une idée de son activité, de son site d'action
    et du mécanisme de son action toxique.

         Les méthodes habituellement utilisées pour l'exploration
    fonctionnelle du rein chez l'homme ne permettent pas un diagnostic
    suffisamment rapide d'une insuffisance rénale due à des substances
    chimiques ni d'en apprécier le retentissement sur la santé.  L'absence
    de marqueurs spécifiques précoces de la néphrotoxicité est
    particulièrement gênante.  Les méthodes non effractives devront faire
    appel à des marqueurs très spécifiques et sensibles, qui permettent de
    déceler rapidement les modifications qui se produisent au niveau rénal
    afin de prévoir l'apparition éventuelle d'une insuffisance
    fonctionnelle.  Les techniques actuelles de contrôle de la fonction
    glomérulaire ou tubulaire ne sont utilisables que lorsque les lésions
    sont déjà très importantes.  On a proposé comme marqueurs
    l'utilisation de constituants tissulaires immunoréactifs mais encore
    faut-il en confirmer la valeur au moyen d'études longitudinales bien
    conçues.

    RECOMMANDATIONS

    1.   Il faut faire un effort soutenu pour développer et valider des
    marqueurs plus sélectifs et plus spécifiques et notamment des
    anticorps monoclonaux en vue d'étudier l'insuffisance rénale chez
    l'animal.  Ces marqueurs pourraient être utilisables chez l'homme.

    2.   Il faudrait compléter la base de données utilisables pour
    déterminer le potentiel néphrotoxique des substances chimiques pour
    l'homme.  A cet effet, on développera et on validera diverses méthodes
    d'expérimentation animale ( in vivo et  in vitro), on mettra au
    point des méthodes nouvelles d'étude de la néphrotoxicité, on étudiera
    les différences interspécifiques et on prendra en compte l'expérience
    tirée des essais précliniques de nouveaux médicaments chez l'homme.

    3.   Il faudrait renforcer les études épidémiologiques (c'est-à-dire
    les études prospectives sur des groupes professionnels ou des groupes
    de la population générale exposés à des substances néphrotoxiques ou
    qui font une consommation excessive d'analgésiques).

    4.   Des efforts plus importants devront être consentis afin d'établir
    le rôle de certaines substances chimiques dans l'étiologie des
    maladies rénales au stade le plus précoce possible (antécédents
    professionnels, recherches de néphrotoxines dans les tissus).

    5.   L'élucidation du mode d'action des substances néphrotoxiques peut
    être utile à la prévention et au traitement des effets rénaux
    indésirables et pourrait contribuer à la prévision du pouvoir
    néphrotoxique des nouveaux médicaments et des nouveaux produits. 
    Parmi les secteurs de recherche particulièrement importants on peut
    citer :

    *    les mécanismes immunologiques

    *    les effets directs des substances chimiques sur les membranes et
         notamment les mécanismes de peroxydation des lipides, les
         interactions membranes/substances chimiques, les déplacements
         d'ions, les événements au niveau des récepteurs

    *    l'activation des proto-oncogènes et la différenciation cellulaire

    *    la régulation du métabolisme cellulaire.

    6.   Il faut également préciser la localisation anatomique de
    certaines fonctions rénales.

    7.   Il faudrait étudier de façon plus approfondie le rôle des
    variations et de la prédisposition génétiques aux effets toxiques des
    médicaments et des produits chimiques.

    8.   La relation entre néphrotoxicité et cancer (mycotoxines et
    néphropathie des Balkans par exemple) doit être étudiée plus à fond.

    RESUMEN Y CONCLUSIONES

         Los riñones son los principales órganos de excreción y
    homeostasis de las moléculas hidrosolubles.  La unidad funcional del
    riñón es el nefrón, que consiste esencialmente en un tubo continuo de
    células heterogéneas sumamente especializadas, y que exhibe una
    notable organización estructural, funcional y bioquímica.  Existen
    diferencias pronunciadas entre unos nefrones y otros, atendiendo a la
    localización de los glomérulos individuales correspondientes en la
    corteza.  Esta compleja organización estructural, combinada con las
    diferencias en la vascularidad regional debida a lo especializado de
    la disposición vascular, da lugar a un órgano sumamente complejo y
    heterogéneo.

         Gran parte de los conocimientos anatómicos y funcionales
    obtenidos en los animales pueden aplicarse directamente al riñón
    humano.  No obstante, los procesos biomédicos y metabólicos del riñón
    humano, así como las diferencias entre las especies animales, no se
    han elucidado de forma tan detallada.  Así, los efectos de las
    sustancias químicas sólo se pueden extrapolar hasta cierto punto de
    unas especies a otras.

         Varias sustancias químicas (tanto terapéuticas como no
    terapéuticas) tienen efectos tóxicos en uno o más elementos anatómicos
    del riñón.  Los efectos tóxicos pueden ser agudos o crónicos, y pueden
    ser directos o estar indirectamente mediados por mecanismos
    inmunológicos.  El impacto que tienen en la salud las sustancias
    químicas nefrotóxicas está relacionado con los factores de riesgo,
    entre los que figuran el estado de la reserva funcional del riñón y
    factores como las lesiones renales ya existentes, las enfermedades, la
    edad, el sexo y la dieta.

         No se ha estudiado bastante la epidemiología de la nefrotoxicidad
    inducida con sustancias químicas aisladas o en exposiciones mixtas. 
    La contribución de las sustancias químicas a la incidencia global de
    la nefropatía y del fallo renal crónico está, salvo raras excepciones,
    sin definir.  En el caso de algunos grupos expuestos por su profesión
    y de enfermedad renal asociada a los analgésicos, se han hecho amplios
    estudios que han demostrado la existencia de la incidencia entre
    grupos y países.  Sin embargo, se estima que hasta el 18% de las
    enfermedades renales en fase terminal pueden deberse a nefropatía por
    analgésicos y hasta el 5% a otras nefropatías tóxicas.  Alrededor del
    50% de las enfermedades renales en fase terminal son de etiología
    desconocida.  Uno de los principales problemas para atribuir una causa
    a la enfermedad renal en fase terminal es la larga latencia y/o la
    lenta evolución del fallo renal crónico, lo que dificulta la
    identificación retrospectiva del agente causal.  Sólo en algunos
    casos, la medida de los niveles tisulares (organismo, riñón) de
    sustancias químicas es útil para el diagnóstico.  Otro problema ha
    sido la falta de coherencia en la terminología y los criterios
    patológicos y de diagnóstico.

         Las diferencias anatómicas y fisiológicas del riñón dificultan la
    extrapolación directa al ser humano a partir de los sistemas
    experimentales ( in vivo e  in vitro).  Existen muy pocos ejemplos
    de sustancias químicas nefrotóxicas para las que se disponga de datos
    adecuados y comparables entre los animales y el hombre como para
    formar una base sólida sobre la que evaluar el riesgo potencial para
    la salud humana.

         Las sustancias químicas pueden dañar de modo selectivo las
    estructuras vulnerables del riñón o activar mecanismos inmunológicos. 
    Los mecanismos de lesión renal pueden dividirse generalmente en dos
    categorías: a) nefritis intersticiales agudas inmunológicamente
    inducidas; b) aquellos que afectan primordialmente al glomérulo por
    conducto de anticuerpos mediados por la anti-GBM o por conducto de
    complejos inmunes.  Otro grupo principal está compuesto por las
    enfermedades desencadenadas por sustancias químicas o sus metabolitos
    que interfieren con los efectos bioquímicos y hemodinámicos en la
    célula, entre otros.  Entre los factores que pueden modificar la
    lesión celular producida por sustancias tóxicas figuran los sistemas
    de transporte celular, la pinocitosis, la degradación metabólica y la
    interacción con las proteínas, los lípidos, las membranas y el ADN
    celulares, y tal vez otros constituyentes de la célula.

         El uso cada vez más extendido de agentes y sustancias químicas
    terapéuticas aumenta las posibilidades de aparición de nefrotoxicidad. 
    La nefrotoxicidad inducida por agentes terapéuticos depende de la
    dosis y del tiempo de exposición (por ejemplo, analgésicos de
    combinación que producen necrosis papilar del riñón).  Los efectos
    nefrotóxicos de los analgésicos, los antibióticos (como los
    aminoglucósidos), los agentes anticancerosos (como el cis-platino), y
    varios agentes más han sido objeto de amplios estudios.  Las
    sustancias químicas de uso frecuente en la industria o el hogar, como
    los hidrocarburos clorurados y el etilenglicol, también tienen
    potencial para producir lesiones renales.  Las sustancias químicas del
    medio ambiente como el plomo y el cadmio son capaces de inducir
    nefrotoxicidad.  Estos agentes tienen efectos tóxicos tras la
    acumulación intracelular del compuesto original o tras la
    biotransformación renal o extrarrenal.  La exposición multiquímica
    puede dar lugar a respuestas antagonistas o sinérgicas.

         Aunque los tumores del riñón y del tracto urinario representan
    menos del 2-3% de todos los cánceres humanos, la frecuencia de esos
    tumores está aumentando, lo que indica que los factores ambientales
    ejercen cierta influencia.  Si bien se ha demostrado que muchos
    fármacos y sustancias químicas son carcinogénicos en modelos
    experimentales, sólo algunas sustancias concretas se han relacionado
    con tumores en el hombre.  Entre ellos figuran el amianto
    (adenocarcinoma renal), el uso indebido de analgésicos (carcinomas de
    transición), y los factores relacionados con el modo de vida (uso de
    tabaco, alcohol, café).  Existen grupos de población con cáncer del
    tracto urinario de etiología aún por determinar.  Las sustancias
    químicas presentes en el medio profesional se han relacionado con la

    etiología del cáncer de la vejiga urinaria.  Muchos fármacos y
    sustancias químicas provocan nefritis intersticial, que en sí misma
    puede ser un factor desencadenante de cánceres en el tracto urinario.

         No hay ningún método  in vivo o  in vitro que por sí solo baste
    para evaluar la disfunción renal químicamente inducida.  Así pues, es
    aconsejable no sólo estudiar compuestos en busca de su potencial
    nefrotóxico, sino incorporar a los protocolos experimentales estudios
    mecanicistas sobre la toxicidad para células diana.  Para conseguirlo,
    deben realizarse investigaciones concertadas  in vivo e  in vitro. 
    Los estudios  in vitro son aquellos en los que se mantienen las
    relaciones anatómicas entre células (por ejemplo, riñón y túbulos
    aislados y perfundidos, secciones de la corteza renal y segmentos
    aislados de nefrones) y aquellos en los que se utilizan células
    renales (por ejemplo, suspensión de células, cultivos de células
    primarias, líneas celulares establecidas y fracciones subcelulares). 
    En los estudios  in vivo se utilizan técnicas tanto invasivas como no
    invasivas.  Los procedimientos invasivos comprenden las mediciones
    histopatológicas y ordinarias de la función renal.  Los procedimientos
    no invasivos permiten evaluar de modo repetido la función renal en los
    animales midiendo una serie de pruebas de la función renal (filtración
    glomerular, función tubular, proteinuria, enzimuria, etc).  Cuando
    convenga, deben realizarse ensayos bioquímicos especializados.  La
    combinación adecuada de experimentos  in vivo e  in vitro revelará
    si las sustancias químicas son nefrotóxicas o no y dará idea de la
    potencia, la localización y los mecanismos de la toxicidad.

         Los métodos tradicionales de evaluación de la función renal en el
    hombre no bastan para diagnosticar de modo oportuno la disfunción
    renal inducida por sustancias químicas y la predicción de su
    importancia para la salud.  La falta de marcadores específicos y
    precoces en la nefrotoxicidad resulta particularmente problemática. 
    La evaluación no invasiva de la nefrotoxicidad debe hacer uso de
    marcadores de gran especificidad, sensibilidad para la detección de
    las alteraciones renales precoces y valor predictivo para la evolución
    de la insuficiencia renal.  Las técnicas actuales de seguimiento de la
    función glomerular o tubular sólo son útiles en los casos en que se ha
    producido una lesión renal grave.  Aunque actualmente se señalan los
    constituyentes tisulares inmunorreactivos como marcadores apropiados,
    es preciso validar su idoneidad en estudios longitudinales bien
    diseñados.

    RECOMENDACIONES

    1.   Debe hacerse un esfuerzo continuado por desarrollar y validar
    marcadores más selectivos y específicos, inclusive anticuerpos
    monoclonales, para evaluar la disfunción renal en animales.  Es
    posible que esos marcadores sean aplicables al hombre.

    2.   Debe ampliarse la base de datos para predecir el potencial
    nefrotóxico de las sustancias químicas para el hombre.  Ello comprende
    el desarrollo y la validación de criterios experimentales en animales
    ( in vivo e  in vitro), métodos alternativos para estudiar la
    nefrotoxicidad, información sobre diferencias interespecíficas, y
    experiencia de la evaluación preclínica de nuevos agentes terapéuticos
    en el hombre.

    3.   Deben reforzarse los estudios epidemiológicos (es decir, estudios
    prospectivos en grupos profesionales y de la población general
    expuestos a sustancias químicas nefrotóxicas o que hagan uso indebido
    de analgésicos).

    4.   Deben intensificarse los esfuerzos por establecer el papel de las
    sustancias químicas en la etiología de las enfermedades renales en la
    etapa más temprana de diagnóstico (por ejemplo, pasado laboral,
    vigilancia de los tejidos en busca de nefrotoxinas).

    5.   La comprensión de los mecanismos de acción de los nefrotóxicos
    ayudará a prevenir y gestionar desde el punto de vista clínico los
    efectos renales no deseados, y puede ayudar a predecir el potencial
    nefrotóxico de nuevos fármacos y sustancias químicas.  Los aspectos de
    particular importancia para las investigaciones futuras son:

    *    los mecanismos inmunológicos;

    *    los efectos directos de las sustancias químicas en las membranas,
         inclusive los mecanismos de peroxidación lipídica, la interacción
         membrana/sustancias químicas, los cambios de iones, y los
         procesos mediados por receptores;

    *    la activación de los proto-oncogenes y de la diferen ciación
         celular;

    *    la regulación del metabolismo celular.

    6.   Es necesario determinar y correlacionar las funciones específicas
    con localizaciones anatómicas discretas dentro del riñón.

    7.   Debe estudiarse más a fondo el papel de la variación genética y
    la susceptibilidad a los efectos tóxicos de fármacos y sustancias
    químicas.

    8.   Debe estudiarse con más detalle la relación entre la
    nefrotoxicidad y la carcinogénesis renal (por ejemplo micotoxinas y
    nefropatía de los Balcanes).



    See Also:
       Toxicological Abbreviations