INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY

        WORLD HEALTH ORGANIZATION



        SAFETY EVALUATION OF CERTAIN
        FOOD ADDITIVES AND CONTAMINANTS



        WHO FOOD ADDITIVES SERIES 40





        Prepared by:
          The forty-ninth meeting of the Joint FAO/WHO Expert
          Committee on Food Additives (JECFA)



        World Health Organization, Geneva 1998



    tert-BUTYLHYDROQUINONE (TBHQ)

    First draft prepared by
    Ms Elizabeth Vavasour and Ms J. Eastwood
    Chemical Health Hazard Assessment Division
    Bureau of Chemical Safety
    Food Directorate, Health Protection Branch
    Health Canada, Ottawa, Ontario, Canada

         1.   Explanation
         2.   Biological data
              2.1  Biochemical aspects
                   2.1.1   Absorption, distribution and excretion
                           2.1.1.1   Rats
                           2.1.1.2   Dogs
                           2.1.1.3   Humans
                   2.1.2   Biotransformation
                   2.1.3   Effects on enzymes and other biochemical
                           parameters
              2.2  Toxicological studies
                   2.2.1   Acute toxicity studies
                   2.2.2   Short-term toxicity studies
                           2.2.2.1   Mice
                           2.2.2.2   Rats
                   2.2.3   Long-term toxicity/carcinogenicity  studies
                           2.2.3.1   Mice
                           2.2.3.2   Rats
                           2.2.3.3   Dogs
                   2.2.4   Reproductive toxicity studies
                   2.2.5   Special studies on teratogenicity
                   2.2.6   Special studies on genotoxicity
                   2.2.7   Special studies on lung toxicity
                   2.2.8   Special studies on the forestomach
                           2.2.8.1   Rats
                           2.2.8.2   Hamsters
                   2.2.9   Special studies on the liver
                   2.2.10  Special studies on the kidney and urinary
                           bladder
                   2.2.11  Special studies on potentiation and inhibition
                           of cancer
                           2.2.11.1  Liver
                           2.2.11.2  Urinary bladder
         3.   Comments
         4.   Evaluation
         5.   References

    1.  EXPLANATION

          tert-Butylhydroquinone (TBHQ) was evaluated by the Committee at
    its nineteenth, twenty-first, thirtieth, thirty-seventh and
    forty-fourth meetings (Annex 1, references 38, 44, 73 94 and 116). At
    the forty-fourth meeting, the previously-established temporary ADI of

    0-0.2 mg/kg bw was extended pending results from ongoing long-term
    studies in rodents. This ADI was derived from a NOEL of 1500 mg/kg of
    feed (equivalent to 37.5 mg/kg bw per day) in a 117-week feeding study
    in dogs on the basis of haematological changes observed at the next
    highest dose level of 5000 mg/kg feed (Annex 1, reference 39).

         At its present meeting, the Committee reviewed the results of the
    long-term studies in mice and rats. In addition, new information
    relating to metabolism of TBHQ, its effects on enzyme induction and
    its short-term and reproductive toxicity in rodents was available for
    review. The results from the long-term study in dogs and the
    genotoxicity studies relating to clastogenic potential of TBHQ were
    also re-evaluated.

         The following consolidated monograph is a compilation of studies
    from the previous monographs and those reviewed for the first time at
    the present meeting.

    2.  BIOLOGICAL DATA

    2.1  Biochemical aspects

    2.1.1  Absorption, distribution, excretion

    2.1.1.1  Rats

         In a single dose study, rats received 14C-labelled TBHQ
    equivalent to 15, 48, 92, 383, 380 or 400 mg/kg bw. Urine and faeces
    were collected daily as was expired CO2. At the end of the test
    period the animals were sacrificed and blood, brain, kidneys, liver,
    gastrointestinal tract, and perirenal, omental and subcutaneous fat
    removed for assay. Of the administered radioactivity, 78-88% was
    recovered in the urine, the bulk of this being excreted within the
    first 24 hours (55-82.7% of the administered dose). Of the recovered
    radioactivity, 70-76% was in the form of the  o-sulfate conjugate and
    1-2% as the  o-glucuronide. Faecal excretion was 2-6%. Only traces of
    radioactivity were detected in the tissues at the 92 mg/kg bw level;
    no values were given for the higher levels (Astill  et al., 1967a).

         In another experiment, rats (body weight 200-250 g) were
    maintained on a daily diet that allowed an intake of 5.7 mg/kg bw
    (0.029% level) of 14C-TBHQ daily for 17 days. Urine and faeces were
    collected throughout the experiment. At the end of the test period the
    rats were starved overnight before sacrifice, and brain, liver,
    kidney, and fat samples collected. Tissue levels were as follows (mg
    TBHQ/g wet tissue): liver, 0.06-0.34; kidney, 0.09-0.38; brain,
    0.06-0.56; fat, 0.06-0.37 (Astill  et al., 1967a).

         Male and female rats (body weight 250 g) were given a single dose
    of TBHQ, dissolved in corn oil (l0% w/w), by intubation at dose levels
    equivalent to 100, 200, 300 or 400 mg/kg bw. At the 400 mg/kg bw level
    there was a rapid onset of ataxia, followed by recovery in 2-3 hours.

    Urine samples were collected daily for 3 days before dosing and then
    for 6 days after dosing. At all dose levels excretion appeared to be
    complete in 3-4 days. About 66% of the dose was excreted as the
     o-sulfate conjugate and less than 10% as the glucuronide. At the 100
    mg/kg bw level, urinary excretion accounted for almost all the dose.
    At higher levels about 33% could not be accounted for in the urine or
    be detected in the faeces. Excretion of the free TBHQ at the 100 mg/kg
    bw level was about 12%, but this decreased at the higher dose levels
    (2% at 400 mg/kg bw). No other major metabolites were detected (Astill
     et al., 1968).

         Urine samples were collected from two animals of each of the 0,
    0.16 and 0.5% dietary TBHQ groups of a long-term feeding study at
    months 12 and 20. Serum samples were collected from groups of five
    rats at months 6, 12, 20 and at autopsy. Samples of perirenal, omental
    and subcutaneous fat were removed at autopsy, and pooled by sex and
    dose. At 12 months, males at both levels excreted about equal amounts
    of the conjugates in the urine  (o-sulfate and  o-glucuronide). Most
    (about two-thirds) of the excretory products in females was the
     o-sulfate form and the remainder the  o-glucuronide. At 20 months
    in both male and female, most of the conjugate excreted was in the
     o-sulfate form with little evidence for glucuronide excretion. Only
    negligible amounts of TBHQ were detected in serum or fat (Astill 
     et al., 1968).

         Portions of the fat of control animals and animals that had been
    maintained on 0.5% TBHQ were examined for stability by the active
    oxygen method (oxidative stability) and also for TBHQ content. There
    were no apparent differences in the oxidative stability of fats from
    treated and control animals, nor did polarographic and colorimetric
    methods of analysis (sensitive to 5 mg/kg) indicate the presence of
    TBHQ in the fat of test animals (Eastman Chemical Products, 1968a).

         Pregnant albino SD rats (380-440 g body weight; age 48 weeks)
    were selected from the third litter of second generation females in a
    reproductive toxicity study that had received 0.5% of TBHQ in the diet
    since weaning. Animals were given one day before term an oral dose of
    14C-TBHQ (40 mg/kg bw) as a 10% solution in corn oil. Urine and
    faeces were collected up to time of sacrifice (7.6-16.7 hours after
    dosing). Fetuses were removed by Caesarean section. The uterus,
    amniotic fluid, gastrointestinal tract, liver, brain, kidneys and fat
    specimens were collected for radioassay. About 74% of the dose was
    excreted in the urine in the 16.7-hour period. Only 10% of the dose
    was detected in the gastrointestinal tract at 7.4 hours after dosing,
    and 8.5% at 17.6 hours. The level of radioactivity in fetuses was 0.2%
    of the dose at 7.6 hours and 0.02% at 16.7 hours. Similar small
    proportions of the dose were present in the uterus and amniotic fluid
    and other tissues examined. Based on these results, extrapolation to
    possible known exposures suggest that at the highest possible intake
    (0.1 mg/kg bw per day), the human fetus would be exposed to the order
    of 1% of the daily intake in the form of unchanged TBHQ and probably
    higher levels of the conjugate (Astill & Walton, 1968).

         Oral doses of 4 ml of 0.01, 0.1 or 1.0% butylated hydroxyanisole
    (BHA) were administered to male F344 rats. After 3 hours, the
    concentrations of  tert-butylquinone (TBQ) (the oxidation product of
    TBHQ) detected by HPLC in the forestomach mucosa were found to be
    0.00453, 0.04504 and 0.05520 µg/animal, respectively, compared with
    1.77, 18.84 and 216.28 µg BHA/animal, respectively (Morimoto 
     et al., 1991).

         TBHQ was not detected in homogenates of forestomach mucosa from
    male F344 rats that had received 4 ml of 0.01-2.0% 14C-BHA.
    Forestomach homogenates were therefore treated with sodium dodecyl
    sulfate in order to reduce TBQ to TBHQ, in which form it could be more
    easily measured. The TBHQ content thus generated in the forestomach
    homogenates was proportional to the dose of BHA. The ratios of the
    total tissue content of TBHQ to the total amount of covalent binding
    of 14C in forestomach were 0.01-0.03% at oral BHA doses of 0.1-2.0%.
    The authors concluded that the covalent binding level was an important
    indicator of reactive metabolites of BHA (Morimoto  et al., 1992).

    2.1.1.2  Dogs

         Male beagle dogs (about 11 kg weight) were fed Purina Chow and
    TBHQ as a single 100 mg/kg bw oral dose via ground meat capsule. Urine
    was collected 3 days before dosing and 6 days after dosing. Excretion
    was essentially complete within 48 hours. The major urinary excretory
    products were the  o-sulfate, and  o-glucuronide conjugates and a
    small amount of unchanged TBHQ. Total recoveries ranged from 77 to
    98%. Most (about two-thirds) of this was as the  o-sulfate and
    one-third as the  o-glucuronide (Astill  et al., 1967b).

         In another study, 26 male and female dogs were used. The dogs
    were maintained on diets containing TBHQ dissolved in corn oil at
    levels equivalent to 0, 0.05, 0.1 or 0.5%. Urine and serum samples
    were collected on day 9 and one day before commencement of feeding
    TBHQ, and at months 3, 6, 12, 13 and 24 of the test period. Serum was
    collected 23 hours after feeding. At autopsy, performed on one dog of
    each sex at each dose level at 12 months, and on the remaining dogs at
    24 months, samples of perirenal, omental and subcutaneous fat were
    removed. Chromatographic studies of the urine indicated excretion of
    both the  o-sulfate and  o-glucuronide conjugates, at all dose
    levels. In the case of males, the  o-sulfate/glucuronide ratios were
    2/1, whereas in females the bulk of the conjugate was in the form of
    the  o-sulfate. Only insignificant quantities of TBHQ were detected
    in the fat (the maximum in males was 7 mg/kg and in females 17 mg/kg,
    but in most cases the value was 0), and serum (0-0.7 mg/litre) (Astill
     et al., 1967b).

         Portions of the fat from test animals and animals that had been
    maintained on the highest level of TBHQ (0.5%) for 2 years were
    examined for stability by the active oxygen method. There was no
    apparent difference in the oxidative stability of fats from treated or
    control animals (Eastman Chemical Products, 1968a). In another study

    TBHQ residues were assayed in fat, brain, liver and kidney of dog and
    rats from the long-term feeding studies. Storage appeared to be
    negligible (Astill & Jones, 1969).

    2.1.1.3  Humans

         Human subjects (males) received TBHQ under the following
    conditions: (1) a gelatin capsule containing 150 mg TBHQ; (2) a
    mixture of TBHQ (2%) in corn oil and graham cracker crumbs, equivalent
    to a dose of 125 mg TBHQ; (3) 100 mg dissolved in cottonseed oil
    contained in a gelatin capsule; (4) 20 g of mixture containing TBHQ,
    2% cottonseed oil and 2% confectioners' sugar in graham cracker
    crumbs. Doses of TBHQ ranged from 20 to 70 mg. Subjects one, two and
    three drank milk immediately after ingesting test material; subject
    four ate doughnuts and drank coffee.

         Urine was collected from subjects 24 hours before dosing and
    during the 72-hour period after dosing. Blood was collected by
    veni-puncture at 3 or 5 and 24 hour after-dosing. Clinical
    observations were made immediately before ingestion and 3 to 6 hours
    after, and consisted of blood pressure, pulse response, condition of
    pharynx, conjunctivae and pupils and neurological effects.
    Haematological studies consisted of haemoglobin, cell volume, WBC,
    differentials, reticulocyte and platelet counts, and total protein.
    Urinalysis consisted of SpGr, albumin, reducing sugars, ketone bodies,
    occult blood, pH and sediment. Levels of TBHQ in serum and metabolites
    of TBHQ in urine were also determined.

         There was no evidence of any systemic effect following ingestion
    of TBHQ. No significant changes were observed in haematological
    studies or urinalysis. Examination of urine indicated that TBHQ was
    excreted as the  o-sulfate and  o-glucuronide conjugates (ratio
    approximately 3:1). These were mainly recovered during the first 24
    hours. No free TBHQ was detected at any time. The manner of ingestion
    had a marked effect on the proportion of the dose recovered from
    urine. TBHQ administered by methods 1 and 3 resulted in only 22-4% of
    the dose being recovered in the urine, whereas method 2 resulted in
    90-100% recovery. In all cases, the same metabolic products were
    present in urine. High recoveries of TBHQ metabolites in urine were
    accompanied by a serum level of 31-37 mg TBHQ/litre at 3 hours for
    subject two, compared to 4-12 mg/litre for subjects one and three. At
    24 hours these levels had fallen to 15 mg/litre for subject two and
    2-12 mg/litre for subjects one and three (Astill  et al., 1967c).

    2.1.2  Biotransformation

         Following the intraperitoneal administration of 400 mg BHA/kg bw
    or 200 mg TBHQ/kg bw TBHQ to male Wistar rats, two previously
    undocumented metabolites, 3- tert-butyl-5-methylthiohydroquinone
    (TBHQ-5-SMe) and 3- tert-butyl-6-methylthiohydroquinone (TBHQ-6-SMe),
    were detected in the urine using GC-MS. The authors suggested that
    these metabolites resulted from the metabolic conversion of
    glutathione conjugates of a quinone or semiquinone form of TBHQ. In

    rat liver microsomal preparations, the formation of two GSH conjugates
    at the 5- and 6- positions of TBHQ in the presence of an NADPH-
    generating system, molecular oxygen and GSH was confirmed. It appeared
    that glutathione S-transferase (GST) is not required for the reaction.
    While inhibitors of cytochrome P-450 markedly reduced formation of
    TBHQ-GSH conjugates, indicating its role in the activation of TBHQ to
     tert-butylquinone (TBQ) autooxidation was also shown to play a
    partial role in this reaction (Tajima  et  al., 1991).

         Benzylthiol derivatives synthesized from TBQ had higher first
    reduction potentials than the parent compound. The authors concluded
    that TBQ maintained its potential for the generation of active oxygen
    species even after its addition to cellular thiols (Morimoto 
     et al., 1991).

         The  tert-butyl semiquinone radical was shown to be formed from
    TBHQ in aerobic rat liver microsomes in the presence of NADPH. A
    concentration of 500 µM TBHQ and 5 µM TBQ produced similar reductions
    in SOD-inhibitable cytochrome c, which was used as an indication of
    excess superoxide anion radical production. The authors concluded that
    autooxidation of the semiquinone formed from the quinone was
    responsible for superoxide formation and that the hydroquinone entered
    the redox cycle via autooxidation. TBQ, but not TBHQ, induced toxic
    injury to rat hepatocyte plasma membrane as indicated by LDH release
    into the culture medium. The authors speculated that
    semiquinone-dependent superoxide formation was responsible for the
    toxic action (Bergmann  et al., 1992).

         Incubation of TBHQ with horseradish peroxidase and hydrogen
    peroxide resulted in its rapid oxidation to TBQ. TBQ epoxide was also
    produced at hydrogen peroxide concentrations of 2.5 mM or more. The
    presence of horseradish peroxidase was not a requirement for the
    production of TBQ epoxide from TBQ (Tajima  et al., 1992).

         Three GSH conjugates were generated by the incubation of TBHQ
    with GSH; two of these were monoconjugates at the 5 or 6 positions
     (tert-butyl group at position 2) and one was a 5,6 diconjugate. The
    redox potentials for the conjugates were twice those for the
    unconjugated hydroquinone. The monoconjugates showed an approximately
    10-fold increase in redox cycling activity (oxygen consumption in the
    presence of a reducing agent) compared with TBHQ, whereas the
    diconjugate showed a 2-fold increase compared with TBHQ. None of the
    major GST isoenzymes were required for the formation of glutathione
    conjugates from TBHQ (van Ommen  et al., 1992).

         Incubation of TBHQ in phosphate-buffered saline resulted in the
    generation of the semiquinone radical through autooxidation,
    accompanied by the formation of superoxide anion, hydroxyl radical and
    hydrogen peroxide as detected by electron spin resonance (ESR)
    spectroscopy. The addition of prostaglandin H synthase resulted in a
    substantial increase of semiquinone production with concomitant
    production of reactive species. Under the conditions of the assay,
    lipoxygenase had no effect on the formation of the semiquinone. The

    presence of either prostaglandin H synthase or lipoxygenase was found
    to accelerate substantially the metabolism of TBHQ to TBQ compared
    with the rates of autooxidation. In an  in vivo study, male Wistar
    rats were fed diets containing 1.5% BHA for 14 days, with concurrent
    administration of the prostaglandin H synthase inhibitors
    acetylsalicylic acid (0.2%) or indomethacin (0.002%) in the
    drinking-water. Both agents produced a significant decrease in the
    amount of TBQ excreted into the urine, compared with controls
    receiving drinking-water only, while the combined urinary excretion of
    BHA and its metabolites, TBHQ and TBQ, was similar for the various
    groups (46.9%, 45.4% and 43.5% of the ingested dose during urine
    collection in the control, indomethacin and acetylsalicylic acid
    groups, respectively). The results suggested an  in vivo role for
    prostaglandin H synthase in the metabolism of TBHQ to TBQ (Schilderman
     et al., 1993a).

         Following intraperitoneal administration of TBHQ (1.0 mmol/kg bw)
    three glutathione metabolites, 2- tert-butyl-5-glutathione- S-
    ylhydroquinone, 2- tert-butyl-6-glutathione- S-ylhydroquinone and 
    2- tert-butyl-3,6-bisglutathion- S-ylhydroquinone, were identified
    in the bile of male F344 rats. Sulfur-containing metabolites of TBHQ
    were identified in the urine. The results indicated that TBHQ
    undergoes oxidation and GSH conjugation  in vivo in the male F344
    rat. These conjugates are excreted into the bile and undergo further
    metabolism prior to excretion in the urine. The authors suggested that
    the sulfur-containing metabolites of TBHQ may occur in amounts
    sufficient of play a role in the toxicity of TBHQ for kidney and
    bladder (Peters  et al., 1996a).

    2.1.3  Effects on enzymes and other biochemical parameters

         Adult male rats (SD strain) were maintained on standard diets
    containing the following additions: (1) none; (2) 5% heated cottonseed
    oil; (3) DL-ethionine, 2.5% level for 10 days; (4) 100 mg/kg bw per
    day phenobarbital for five days (intraperitoneal injection), (5) 1%
    corn oil + 0.05% BHA; (6) 4% corn oil + 0.2% BHA; (7) 1% corn oil +
    0.05% TBHQ; (8) 4% corn oil + 0.2% TBHQ; and (9) 5% heated cotton-seed
    oil + 0.025% TBHQ. A liver microsomal fraction was prepared from each
    group of animals and glucose-6-phosphatase (G-6-Pase),
     p-nitroanisole demethylase ( pNaD) and aniline hydroxylase (AHase)
    activities determined. The expected elevation of  pNaD (5x) and AHase
    (3x) occurred with phenobarbital, but DL-ethionine had no significant
    effect on these enzymes. Phenobarbital produced a depression of
    G-6-Pase activity (25%), and TBHQ at 0.05% level produced a 25%
    depression in G-6-Pase which was absent at the 0.2% level. TBHQ had no
    effect on  pNaD at the 0.05% level, but produced at the 0.2% level a
    60% elevation of  pNaD. There was no clear effect on AHase. In
    contrast, BHA produced a 30% decrease in G-6-P at both levels, a 50%
    increase in  pNaD at 0.05% and a 700% increase of  pNaD at the 0.2%
    level. There was no effect on AHase. Inclusion of heated oil in the
    diet had no marked effect on previous changes. In another experiment
    in which measurements were made of enzyme activities in microsomal
    preparation from livers of rats fed for 180 days diets containing 0.5%

    TBHQ dissolved in either heated or unheated cottonseed oil, no
    significant differences were observed that could be attributed to heat
    treatment of oil before addition to the diet (Tischer & Walton, 1968).

          pNaD, AHase, and G-6-Pase activities of microsomal fractions
    from dogs that had been maintained on diets containing 0, 0.05, 0.16
    and 0.5% TBHQ for 2 years were within the range of control values
    (Tischer & Walton, 1968).

         Electron microscopy studies of liver and kidney tissue from both
    dog and rat showed that long-term administration of TBHQ did not
    significantly alter the subcellular constituents or cause a
    proliferation of the endoplasmic reticulum of liver cells (Wolf &
    Fassett, 1968a,b).

         The effects of antioxidants, including TBHQ, on prostaglandin
    biosynthesis were examined by determining the production of
    prostaglandin E1 (PGE1) and prostaglandin E2 (PGE2) by incubated
    microsomal fractions of bovine seminal vesicles. All the antioxidants
    tested proved to be concentration-dependent inhibitors of
    prostaglandin biosynthesis. A 50% inhibition of PGE1 and PGE2
    biosynthesis was observed at TBHQ concentrations of 5.5 and 6.1 µM,
    respectively. A 50% inhibition was also observed with BHA at
    comparable concentrations, while much higher concentrations of
    butylated hydroxytoluene (BHT) and ascorbate were required to have the
    same inhibitory effect (Boehme & Branen, 1977).

         Six phenols, including TBHQ and BHA, were examined for their
    ability to induce hepatic mono-oxygenase and detoxication enzyme
    activities in female CD-1 mice. TBHQ treatment (42 nmol/kg diet for 12
    days) had no effect on relative liver weight, cytochrome P-450 content
    or on the activities of AHase, aminopyrine N-demethylase or
    peroxidase. TBHQ treatment was shown to increase cytosolic GST
    activity 2-fold while UDP-glucuronyl transferase (UGT) activity was
    reduced by one third. In an additional  in vitro experiment with
    mouse liver microsomes, TBHQ was shown to inhibit benzo (a)pyrene
    (BP) metabolism and its DNA-binding capacity. These effects were not
    observed in the  in vivo study (Rahimtula  et al., 1982). 

         Groups of 50 rainbow trout were fed diets containing 0 or 5.6
    mmol/kg TBHQ (equal to 0.1%), BHT, BHA or ethoxyquin for 6 weeks. The
    treated trout had reduced liver/body weight ratios. Compared to
    controls, TBHQ treatment led to a decrease in hepatic microsomal
    protein and cytochrome P-450 content, and  pNaD activity. In
    contrast, the hepatic activities of BP hydroxylase, epoxide hydratase,
    ethoxycoumarin-O-deethylase and NADPH-cytochrome c reductase were
    elevated following TBHQ treatment (Eisele  et al., 1983).

         Male albino rats (CRL:COBS.CD(SD)BR) were dosed daily (i.p.) on 3
    consecutive days with TBHQ at dose levels equivalent to 50, 100 or 150
    mg/kg bw. No spontaneous haemorrhage was observed in any of the test
    animals on autopsy, nor was there any effect on prothrombin time. It

    should be noted that in a similar study on this strain of rats, BHT at
    dose levels up to 1520 mg/kg bw failed to produce deaths due to
    haemorrhage, and only at this high dose did BHT cause a significant
    increase in prothrombin time (Krasavage, 1984).

         TBHQ was found to be an inducer of quinone reductase in cell
    mutants and mouse strains that lack the Ah receptor. In these target
    tissues, agents that induce both phase I and phase II enzymes (such as
    polycyclic aromatics) were ineffective (Prochaska, 1987).

         TBHQ was found to stimulate the production of superoxide,
    hydrogen peroxide and hydroxyl radicals in microsomes from rat liver
    and forestomach. The oxidation product of TBHQ, i.e. TBQ, exceeded
    TBHQ in its capacity to induce oxygen radical formation (Kahl 
     et al., 1989).

         TBHQ (30 µM) was found to induce quinone reductase activity in
    cultures of bone marrow stromal cells from C57BL/6 and DBA/2 mice
    (Twerdok & Trush, 1990).

         An assay was described for measuring induction of quinone
    reductase in Hep 1c1c7 cells as a screen for Phase II enzyme inducers.
    TBHQ, a representative monofunctional inducer, was shown to induce
    quinone reductase in wild-type Hepa 1c1c7 cells and its mutant
    subclones defective in either functional  Ah receptor or cytochrome
    P-450 gene product (Prochaska, 1994).

         The induction of a human oxidoreductase (H-37) in wild type HT29
    colon cells by Michael reaction acceptors, including TBHQ, was
    investigated. H-37, also called dihydrodiol dehydrogenase (DDH), has
    been identified as a human hepatic bile acid binding protein, which
    may be an important determinant of net hepatic bile acid transport.
    These investigations demonstrated that DDH is inducible by  de novo 
    synthesis from mRNA by Michael reaction acceptors. Indirect evidence
    suggested that an element similar to hARE (human antioxidant response
    element) is involved in the regulation of DDH by these agents (Ciaccio
     et al., 1994). 

         For comparative purposes, the effects of TBHQ on the regulation
    of  Ah gene battery enzymes and glutathione levels were investigated
    in mouse hepatoma cell lines. The cell lines used included a wild
    type, a CYP1A1 metabolism-deficient mutant and an aromatic hydrocarbon
    receptor nuclear translocation (ARNT)-deficient mutant. In wild type
    cells, TBHQ treatment resulted in increased activities of all Phase II
    enzymes in the  Ah gene battery: NAD(P)H:menadione oxidoreductase
    (NMO1), cytosolic aldehyde dehydrogenase (ALDH3c), UDP-glucuronosyl
    transferase (UGT1*06) and glutathione  S-transferase (GST). TBHQ
    treatment induced the activities of NMO1, ALDH3c and UGT1*06 in
    ARNT-deficient cells, indicating that the induction of these Phase II
    enzymes by TBHQ was through the electrophilic response element (EpRE)
    and independent of the Ah receptor. No significant induction of Phase
    II enzymes was observed when CYP1A1 metabolism-deficient cells were
    treated with TBHQ. However, all responses were muted in this cell line

    because in the absence of CYP1A1, the activities of all Phase II
    enzymes were elevated in these cells compared to the wild type. In all
    cell lines, TBHQ treatment resulted in an elevation of intracellular
    GSH levels. Increases in intracellular cysteine, the precursor for GSH
    synthesis, and the activity of gamma-glutamyl cysteine synthetase
    (GCS), the rate-limiting enzyme in GSH synthesis, appeared to be
    involved in the elevation of GSH levels. The authors predicted that
    the GCS gene regulatory region would contain an EpRE but not an AhRE
    (Ah response element) (Liu  et al., 1994).

         The hARE-mediated regulation of type 1 NAD(P)H:quinone
    oxidoreductase (NQO1) gene expression was investigated by
    transforming mouse hepatoma (Hepa 1) cells with a
    hARE-tk-choramphenicol acetyltransferase (CAT) recombinant plasmid.
    Various compounds, including ß-naphthoflavone, BHA, TBHQ, tumour
    promoters and hydrogen peroxide, were shown to increase the expression
    of the hARE-mediated CAT gene in transformed Hepa 1 cells. The authors
    concluded that binding of Jun and Fos nuclear proteins to the hARE,
    hARE-mediated gene expression and induction of the gene in response to
    phenols were sensitive to alteration by sulfhydryl-modifying agents.
    The protein factors involved in signal transduction from xenobiotics
    to the Jun and Fos proteins were not apparent from the investigations
    but the authors suggested that oxygen species generated during
    reductive metabolism of xenobiotics and GST may be involved (Li &
    Jaiswal, 1994).

         Induction of GST-Ya and NQO gene expression by a variety of
    chemical agents is mediated by regulatory elements EpRE and ARE,
    composed of two adjacent AP-1-like binding sites and activated by
    Fos/Jun heterodimeric complex (AP-1). Investigations were conducted in
    HepG2 cells transfected with an EpRE Ya-cat plasmid constructed by
    ligation of the EpRE from the promoter region of the mouse GST-Ya gene
    to drive the expression of the CAT gene. Intracellular GSH levels were
    depleted by pre-treatment with L-buthionine- S,R-sulfoximine (BSO), a
    specific inhibitor of gamma-glutamylcysteine synthetase. Depletion of
    GSH by pre-treatment with BSO, resulted in a 2.8 fold increase in the
    induction of CAT activity by TBHQ compared to that induced in control
    cells. TBHQ induction of the binding activity of AP-1 was also
    increased following BSO pre-treatment. The authors suggest that the
    chemical induction of AP-1 activity leading to the AP-1-mediated
    transcriptional activation of GST-Ya gene expression occurs at low GSH
    levels created by the generation of intracellular oxidants and
    consumption of GSH (Bergelson  et al., 1994a).

         EpRE and ARE are regulatory elements responsible for the chemical
    induction of GST-Ya expression in the mouse and rat, respectively.
    These elements have been found to be composed of two adjacent
    AP-1-like binding sites and are activated by the Fos/Jun heterodimeric
    complex. HepG2 cells transfected with plasmids containing the EpRE,
    ARE or a single AP-1 consensus sequence site ligated to the GST-Ya
    gene promoter driving expression of the CAT gene were exposed to TBHQ
    or ß-napthoflavone. TBHQ resulted in a 3-, 4- and 1.6-fold induction
    of the EpRE-, ARE- and AP-1-mediated CAT activity, respectively.

    Studies with other chemical inducers provided evidence that these
    agents induce expression of c- fos and c- jun proto-oncogenes and
    enhance synthesis of protein components of AP-1 complex. The authors
    suggested that the increased synthesis of AP-1 complex followed by an
    AP-1-mediated transcriptional activation of GST-Ya and NQO genes may
    provide a molecular mechanism for the chemical induction of these
    enzymes (Bergelson  et al., 1994b).

         Studies were conducted to examine the response of genes in the
    dioxin-inducible  Ah gene battery to three compounds, including TBHQ,
    that protect mouse hepatoma cells (Hepa-1 cells) against menadione
    toxicity. The protective effect of each compound correlated with the
    induction of mRNAs and enzymes of the  Ah gene battery. TBHQ
    significantly enhanced the NM01, ALDH3c and UGT1*06 mRNA levels but
    caused no increases in levels of CYP1A1 mRNA. The studies demonstrated
    that the TBHQ induction processes do not require a functional  Ah 
    receptor, but rather operate via the EpRE, found in the regulatory
    regions of the murine NMO1, ALDH3c and UGT*06 genes (Vasiliou 
     et al., 1995).

         TBHQ was shown to induce the activity of gamma-glutamylcysteine
    synthetase, GSH synthetase, NMO1, ALDH3c, UGT1*06, and GST in mouse
    hepatoma Hepa-1c1c7-wild-type cells. In  Ah receptor nuclear
    translocation-defective Hepa 1 cells, TBHQ induced the activities of
    all the above enzymes except for GST. In both cell lines, TBHQ
    treatment resulted in increased GSH concentrations. The data indicate
    that the enzyme induction resulting from TBHQ treatment is not
    mediated via the  Ah receptor and is not secondary to depletion of
    GSH levels (Shertzer  et al., 1995).

          In vitro studies with hepatocytes from male SD rats were
    conducted to investigate the effects of TBHQ on enzymes in the  Ah 
    gene battery. TBHQ induced the activity of GST-Ya and NQO but had no
    effect on the activity of CYP1A1, UGT*06 or ALDH3c. Co-administration
    of dexamethasone, a synthetic glucocorticoid, potentiated the
    induction of GST-Ya and suppressed the induction of NQO by TBHQ.
    Protein kinase A or C inhibitors had no effect on the induction of
    GST-Ya or NQO enzyme activity by TBHQ either in the presence or
    absence of dexamethasone. The authors concluded that the
    glucocorticoid receptor was involved in the regulation of the GST-Ya
    and NQO enzyme activity by  Ah receptor-independent mechanisms.
    Unlike the action of the  Ah receptor, phosphorylation of the ARE
    transcription factors or the glucocorticoid receptor was not a factor
    in the activation of GST-Ya and NQO genes by electrophilic compounds
    such as TBHQ (Xiao  et al., 1995).

          In vitro studies in HeLa, HepG2 and F9 cells demonstrated that
    the protein that binds to the ARE and mediates the induction of Phase
    II genes by TBHQ is not AP-1. TBHQ was shown to suppress canoylphorbol
    13-acetate-induced AP-1 transcriptional activity. TBHQ induced 
     c-jun, junB, fra-1 and  fra-2, but not  c-fos, and therefore the
    AP-1 complexes induced by TBHQ contain mostly Fra-1. The Fra-1
    proteins are capable of binding with Jun proteins and binding to AP-1

    sites, but are devoid of transcriptional activation function. The
    authors suggest that the anti-tumour-promoting activity of phenolic
    antioxidants results through preferential induction of Fra-containing
    AP-1 complexes (Yoshioka  et al., 1995).

         An assay system was developed using  E. coli stably transfected
    with the  lacZ structural gene fused to a wide variety of  E. coli 
    and  S. typhimurium stress gene promoters in a modified
    ß-galactosidase protocol. The authors suggested that induction of
    specific subsets of promoters could reveal information on the
    mechanism of toxicity of certain xenobiotics. Under the assay
    conditions, TBHQ was shown to induce the  micF, zwf, clpB and  dinD 
    promoters. Since these promoters all belonged to different subsets the
    results provided no clear information about the intracellular effects
    of TBHQ (Orser  et al., 1995).

         Phenolic antioxidants, including TBHQ, were shown to induce
    cytosolic class-3 aldehyde dehydrogenase activity in human breast
    adenocarcinoma MCF-7/0 cell cultures (Sreerame  et al., 1995).

         Studies were conducted to investigate whether the induction of
    AP-1 activity and GST-Ya gene expression by BHA and it metabolite TBHQ
    is due to generation of reactive oxygen species or to an antioxidant
    activity. Exposure of intact HepG2 cells to TBHQ was shown to result
    in generation of *OH radicals and a decrease in intracellular GSH
    levels. The addition of catalase to the culture medium inhibited the
    generation of *OH radicals, the induction of AP-1- and
    NF-kappaB-binding activities and the induction of EpRE Ya- cat 
    expression by TBHQ or BHA. Antioxidants, GSH and N-acetyl cysteine
    were also shown to inhibit the induction of AP-1- and
    NF-kappaB-binding activites by BHA and TBHQ and the induction of
    endogenous GST Ya gene in hepatoma cells or a transfected EpRE-Ya cat
    construct by BHA or TBHQ. The authors concluded that the induction of
    AP-1 and NF-kappaB by TBHQ or BHA is not an antioxidant response but
    secondary to the generation of oxygen radicals (Pinkus  et al., 
    1996).

         The effect of TBHQ on activities and mRNA levels of
    gamma-glutamyl transpeptidase (GGT) and gamma-glutamylcysteine
    synthetase (GCS) was investigated in rat lung epithelial L2 cells.
    TBHQ increased the activities and the mRNAs for GGT and the catalytic
    subunit of GCS but the time and concentration dependencies differed.
    Actinomycin D (an inhibitor of RNA synthesis) abolished the increase
    in GCS-mRNA but not the increase in GGT-mRNA, suggesting a difference
    in regulation of these genes by TBHQ (Liu  et al., 1996a).

         In an  in vitro study using rat lung epithelial L2 cells, TBHQ
    was shown to increase the intracellular GSH concentration. L2 cells
    pretreated with a non-toxic concentration of TBHQ (50 µM) acquired
    resistance to a subsequent challenge with a lethal concentration of
    TBHQ (200 µM). Pretreatment of the cells with an inhibitor of either
    GCS or GGT prevented the TBHQ-induced increase in GSH and markly
    reduced the resistance to 200 µM TBHQ. The authors suggested that the

    elevation of GST and GGT activites participated in the acquired
    resistance to quinone toxicity (Liu  et al., 1996b).

    2.2  Toxicological studies

    2.2.1  Acute toxicity studies

         The results of acute toxicity studies with TBHQ are summarized in
    Table 1.

        Table 1. Acute toxicity studies with TBHQ

                                                                                    
    Animal           Route       LD50                      Reference
                                 (mg/kg bw)
                                                                                    

    Rat, fed         oral        955 (10% in corn oil)     Terhaar et al., 1968a
                                 890 ( 5% in corn oil)     
    Rat, fasted      oral        756 (10% in corn oil)     Terhaar et al., 1968a
                                 802 ( 5% in corn oil)
    Mouse, fasted    oral        1040                      Terhaar et al., 1968a
    Guinea-pig       oral        790                       Terhaar et al., 1968a
    Dog              oral        >4001                     Terhaar et al., 1968a
                                                                                    

    1  At this dose, dogs consistently regurgitated but not until approximately 
    10 hours after dosing.
    
    2.2.2  Short-term toxicity studies

    2.2.2.1  Mice

        In a 13-week toxicity study, groups of 10 B6C3F1 mice/sex/dose
    were fed  ad libitum diets containing 0, 2500, 5000, 10 000, 20 000
    or 40 000 mg/kg TBHQ (99% pure), equal in males/females to 440/500,
    870/1075, 1950/2175, 4000/4630 and 8425/9040 mg/kg bw per day,
    respectively. Animals were observed twice daily, and food consumption
    and body weights recorded weekly. Separate groups of 10 mice/sex/dose
    were used to determine a standard set of haematology and clinical
    chemistry parameters on day 5 and during week 3 and coagulation
    analyses during week 13. Blood was also collected during week 13 from
    core study mice for clinical chemistry and haematology analyses. At
    the end of the study, samples from 0, 2500, 10 000 and 40 000 mg/kg
    mice were collected for reproductive tissue evaluation and estrous
    cycle characterization. A necropsy was performed on all core animals.
    The heart, right kidney, liver, lungs, right testis and thymus were
    weighed. A complete histopathological examination, including
    forestomach, liver and thyroid, was performed on all control and
    high-dose mice. The nose, skin and forestomach of all male and female
    mice were examined.


        Two unscheduled deaths of treated females occurred during the
    study and were considered unrelated to TBHQ. There was a dose-related
    and statistically significant (p<0.05) decrease in final body
    weight and body weight gain in male and female mice in the 10 000,
    20 000 and 40 000 mg/kg dose groups. Although measured feed
    consumption was comparable for all groups, mice in the 10 000, 20 000
    and 40 000 mg/kg groups tended to scatter more feed, suggesting that
    actual food consumption was lower in these groups. TBHQ treatment was
    associated with alopecia and hair discoloration, possibly the result
    of dermal exposure to spilt feed. There was a dose-related decrease in
    blood urea nitrogen observed in both sexes at all time points. There
    were no other toxicologically significant changes in clinical
    chemistry parameters. Haematological examination revealed increases in
    erythrocytes, reticulocytes, platelets, leukocytes and segmented
    neutrophils in the 20 000 and 40 000 mg/kg groups compared to controls
    at all time points. The increases were observed in both sexes, but
    were more often statistically significant in the females. Except for
    the increase in segmented neutrophils, the authors attributed these
    increases to decreased body weight gain and associated dehydration.
    All haematological and clinical chemistry values were within the
    reference ranges provided for B6C3F1 mice in the CRC Handbook of
    Toxicology (1995).

        At the 10 000, 20 000 and 40 000 mg/kg dose level, all measured
    absolute organ weights were lower than those of controls for both
    males and females. However, the respective relative organ weights were
    higher than those of controls in these groups, suggesting that the
    differences were secondary to reduced body weights. The same trend was
    observed in the weight of the left cauda, left testicle and left
    epididymis in the 10 000 and 40 000 mg/kg groups during the
    reproductive tissue examination. The estrous cycle of females exposed
    to 40 000 mg/kg was significantly longer than that of the controls.
    This was considered secondary to the reduced body weights of these
    females. There was a dose-related increase in the incidence and
    severity of mucosal hyperplasia in the forestomach of all TBHQ-exposed
    females and males exposed to 20 000 and 40 000 mg/kg. In the 10 000,
    20 000 and 40 000 mg/kg groups, there was a dose-related increase in
    the incidence of suppurative inflammation in the nose and chronic
    inflammation and epithelial hyperplasia of the skin in both sexes. The
    NOEL was 5000 mg/kg (equal to 870 mg/kg bw per day) based on decreased
    body weight gain and increased incidence of mucosal hyperplasia of the
    forestomach and inflammation of both the nose and skin at higher doses
    (NTP, 1995).

    2.2.2.2  Rats

        Rats were injected (i.p.) with 200 mg/kg bw TBHQ daily for one
    month without mortality but some loss of weight. When 100 mg/kg bw was
    injected (i.p.) daily for one month, there was no loss of weight. No
    histopathological changes occurred in either case. Rats were
    maintained for 22 days on a diet containing 1% TBHQ. Initial rejection
    of food was followed by near normal food intake and growth curve

    paralleling the control group. There were no deaths, nor was there any
    gross or microscopic pathology (Fassett et al., 1968).

        Eight groups, each of 30 SD rats (equally divided by sex), were
    fed four levels of TBHQ in unheated oil and at the same level in
    heated, oil (one hour to raise temperature to 190°C followed by 4
    hours at 190°C) for 6 months. The levels of TBHQ in the oils were 0,
    0.02, 0.1 and 0.5%. Oils were incorporated at a 5% level into a
    standard diet of ground Purina Chow resulting in final dietary levels
    of 0, 10, 50 and 250 mg/kg, respectively. Animals were housed five per
    cage, and water was available at all times. Body weight and gross feed
    consumption were recorded weekly for the first two nights and
    thereafter fortnightly. General appearance and behaviour were observed
    during the test period. Haemograms and urinalysis were conducted on
    the 0.5% unheated and heated and control at 1, 5 and 6 months.
    Haemograms consisted of haemoglobin, haematocrit, WBC and
    differentials and protein determination. Urinalysis consisted of pH,
    SpGr, occult blood, albumin, reducing sugar and microscopic
    examinations of the sediment. AST and AP were done on the high and
    control groups at 3 and 6 months. At autopsy, liver, kidney, heart,
    spleen, lung, brain and testes weights were determined. The following
    tissues were examined microscopically: lung, heart, tongue,
    oesophagus, stomach, small and large intestines, liver, kidney,
    urinary bladder, pituitary gland, adrenal, pancreas, thyroid, gonads,
    spleen, bone marrow, cerebrum, cerebellum and eye. Three deaths
    occurred during the test period, but these were not compound-related.
    Male rats on the 0.5% TBHQ/ unheated fat showed a slight depression in
    weight gain and those in the 0.02% TBHQ/unheated fat a significant
    increase in weight gain over control. These effects were not observed
    in female rats on diets containing heated fats. Female rats in all
    groups showed similar weight gains to controls. Food intake of test
    groups was comparable or better than controls. Haematological tests
    gave similar values for test and control groups with the exception of
    the 0.5% TBHQ/unheated fat male group at three months, where the WBC
    was slightly elevated. This effect was not noted at six months.
    Urinalysis, AST and AP values of test and control groups were
    comparable and within normal limits. Organ/body weight ratios
    indicated a slight increase in ratios for testes and livers of the
    male rats from 0.5% TBHQ/heated oil group and liver weight ratio of
    the female rats of the 0.5% and 0.2% heated oil group. These minor
    differences appeared to be related to heated versus unheated fat
    rather than a compound effect. Histological studies did not reveal any
    compound-related effects (Terhaar & Krasavage, 1968b).

        Following exposure during gestation and lactation in a
    reproductive toxicity study (see section 2.2.4), groups of 10 F344/N
    rats/sex were fed  ad libitum diets containing 0, 2500, 5000 or
    10 000 mg/kg TBHQ (99% pure), equal in males/females to 0/0, 190/190,
    370/360 and 780/750 mg/kg bw per day, for 13 weeks after weaning.
    Clinical findings, body weights and food consumption were recorded
    weekly. Separate groups of 10 rats/sex/dose were used to determine a
    standard set of haematology and clinical chemistry parameters on day 5
    and during week 3 and coagulation analyses during week 13. Blood was

    also collected from the core study rats during week 13 for clinical
    chemistry and haematology analyses. At the end of the study, samples
    from all groups were collected for reproductive tissue evaluation and
    estrous cycle characterization. A necropsy was performed on all core
    study animals. The heart, right kidney, liver, lungs, right testis and
    thymus were weighed. A complete histopathological examination,
    including forestomach and kidney, was performed on all control and
    high-dose rats. In addition, the following tissues were examined: the
    nose of all exposed groups of males and 5000 mg/kg females; the spleen
    of 5000 and 10 000 mg/kg males and all exposed groups of females; the
    mesenteric lymph node of 5000 mg/kg females; and the kidneys of 2500
    and 10 000 mg/kg female rats.

        There were no unscheduled deaths during the study. Mid- and
    high-dose males/females started the study weighing 10/5% and 28/22%
    less than controls, respectively, and ended the study weighing 6/7%
    and 15/12% less than controls, respectively. The differences in body
    weights were statistically significant for high-dose animals at study
    initiation and for mid- and high-dose groups at study termination.
    Weight gains were comparable for all groups except for high-dose
    males, who gained 9% less body weight than controls (statistically
    significant, p<0.01). In males, feed consumption was lower in mid-
    and high-dose groups than controls during week 2 but comparable to
    controls during week 13. High-dose females tended to consume less feed
    than controls throughout the study. The only clinical observation
    attributed to TBHQ was hair discoloration, which was observed in all
    exposed groups except the low-dose females. The mean spermatid count,
    spermatid heads per testis and spermatid heads per gram of testis were
    significantly decreased in males at the 5000 mg/kg dose level but
    unaffected at the 10 000 mg/kg dose level. There was no effect of TBHQ
    on epididymal spermatozoa concentration or motility. There is no
    evidence of a dose-response and therefore the toxicological
    significance of the reduction at the mid-dose is questionable. Estrous
    cycles were significantly longer in the low- and mid-dose females than
    in controls. In the high-dose group, the mean length of the estrous
    cycle was comparable to controls, but it was reported that in 2/10
    high-dose females the estrous cycle was unclear.

        Serum bile acids were significantly increased in the 5000 and
    10 000 mg/kg males and females, at 5 days, 3 weeks and at study
    termination. ALT activity was increased at day 5 in high-dose females.
    At week 3, all exposed females tended to have higher ALT levels than
    controls, but levels remained within normal ranges. ALT levels were
    comparable for all female groups at the study termination. There were
    some changes in absolute and relative organ weights, which appeared to
    be secondary to reduced body weights. The absolute weight of the heart
    of mid- and high-dose males and females was lower than that of
    controls. The absolute lung weight was significantly decreased in all
    exposed males and high-dose females compared to their respective
    controls. At all dose levels, the relative weight of the testis was
    significantly higher than that of controls. The relative liver and
    kidney weights of all exposed males were increased significantly
    (p<0.05) compared to controls while the only absolute weight

    affected was that of the liver in low-dose males (increased). The
    relative liver weights of all exposed females were significantly
    (p<0.05) increased compared to controls, although there were no
    differences in absolute weights. These changes in organ weight were
    not associated with any pathological findings. An increased incidence
    of hyperplasia of the nasal respiratory epithelium was observed in
    mid-dose males and high-dose males and females. Nasal exudate was also
    observed more frequently in high-dose males. There was a dose-related
    increase in the incidence of splenic pigmentation in exposed animals
    of both sexes: 0/10, 1/10, 3/10, 5/10 males and 0/10, 5/10, 8/10 and
    10/10 females in the 0, 2500, 5000 and 10 000 mg/kg groups,
    respectively. In addition, atrophy of the red pulp was observed in
    8/10 and 10/10 mid- and high-dose females, respectively. There was a
    dose-related decrease in the incidence of renal mineralization in
    exposed females. No NOEL could be established for this study because
    an increased incidence of splenic pigmentation was observed in females
    at all dose levels. The LOEL for for this study was 2500 mg/kg in the
    diet, equal to 190 mg/kg bw per day (NTP, 1995).

    2.2.3  Long-term toxicity/carcinogenicity studies

    2.2.3.1  Mice

        Groups of 60 B6C3F1 mice/sex were fed  ad libitum diets
    containing 0, 1250, 2500 or 5000 mg/kg TBHQ (99% pure), equal in
    males/females to 0/0, 130/150, 290/300 and 600/680 mg/kg bw per day,
    respectively, for 2 years. Mice were housed individually and observed
    twice daily, 7 days/week. Individual body weights were recorded weekly
    for the first 13 weeks and once every 4 weeks thereafter. Food
    consumption was measured monthly. An interim sacrifice of 6 to 10
    randomly selected mice/sex/group was conducted at 15 months. At this
    time, complete gross and microscopic evaluations were conducted,
    weights of the right kidney, liver and right testis were recorded, and
    a standard set of haematological parameters was determined. A complete
    necropsy, including gross and microscopic evaluations, was performed
    on all animals at termination and where possible, on mice dying during
    the study.

        Survival rates ranged from 58 to 77% and were unaffected by TBHQ
    treated. The average feed consumption for TBHQ-treatment groups was
    similar to that for controls. Male and female mice in the high-dose
    group tended to weigh less throughout the study, averaging 8 and 11%
    less than controls at termination, respectively (statistics not
    reported). At 15 months, absolute and relative liver weights of
    TBHQ-treated male and female mice tended to be higher than those of
    controls. However, there was no clear dose-response to the trend and
    statistically only the relative liver weight of high-dose females was
    significantly (p<0.01) different from that of the controls. There
    were no other differences in organ weights. Males in the 5000 mg/kg
    group had significantly (p<0.05) higher levels of reticulocytes
    than controls. There were no other differences in haematological
    parameters. There were no treatment-related differences in the

    incidence of neoplasms or non-neoplastic lesions observed at 15 months
    in either sex. At termination, the incidence of hepatocellular adenoma
    and hepatocellular adenoma or carcinoma was significantly lower in
    high-dose males than in controls. The same trend was observed in
    female mice, although the differences were not statistically
    significant. Although females in the 1250 mg/kg group (low-dose) had
    significantly more hepatocellular adenomas and adenomas or carcinomas
    than controls, the incidence was within historical control ranges and
    considered unrelated to treatment. The incidence of follicular cell
    adenoma in the thyroids of 5000 mg/kg females was higher than in
    controls although the differences did not reach statistical
    significance: 1/51, 3/51, 2/50 and 5/54 in the 0, 1250, 2500 and 5000
    mg/kg groups, respectively. This was associated with a dose-related
    increase in follicular cell hyperplasia: 12/51, 19/51, 24/50 and 24/54
    in the 0, 1250, 2500 and 5000 mg/kg groups, respectively. There were
    no thyroid follicular cell carcinomas observed in any female mice on
    study. The incidence of follicular cell adenomas was at the upper
    limit of the historical control range. No historical control data were
    provided on the incidence of follicular cell hyperplasia. There were
    no other treatment-related effects on the incidence of neoplasms or
    non-neoplastic lesions. There was no evidence that TBHQ was
    carcinogenic in the mouse. Because the proliferative effects on the
    follicular cells of the thyroid were noted in female mice at the
    lowest dose tested, a NOEL could not be established. The LOEL for this
    study was 1250 mg/kg diet, equal to 130 mg/kg bw per day (NTP, 1995).

    2.2.3.2  Rats

        Groups of each of 100 albino rats (A & C Farms, Altamont, N.Y.)
    equally divided by sex were maintained on diets containing 0, 0.016,
    0.05, 0.16 or 0.5% TBHQ for 20 months. Animals were observed for
    changes in appearance and behaviour. Body weight was reported at
    approximately 14-day intervals and group food intake at approximately
    23-day intervals. Haematology was carried out at 3, 6, 12 and 20
    months on 10 rats (5/sex) from each of the 0, 0.16 and 0.5% level
    groups. Haematological tests consisted of haemoglobin, PCV, WBC and
    differentials. Clinical chemistry tests and urinalysis were carried
    out on the same groups of animals at 6, 12 and 20 months. The clinical
    chemistry tests consisted of ALT at 6 months, and ALT, ASAT and AP at
    12 and 20 months. Urinalysis consisted of SpGr, albumin, sugar and
    appearance. At 6 and 12 months, approximately 20 rats (10/sex) of each
    group were sacrificed and at 20 months all surviving animals were
    sacrificed and autopsied. Organ weights were determined for liver,
    adrenal, kidney, spleen, heart, brain, lung and testes.
    Histopathological studies were made on trachea, lung, heart,
    oesophagus, stomach, small intestine, large intestine, liver, kidney,
    urinary bladder, adrenal, pancreas, thyroid, ovary or testes, uterus,
    spleen, femoral bone marrow, cerebrum, cerebellum and eye. There were
    no adverse changes in appearance and behaviour of the rats during the
    test period. Mortalities occurred with equal frequency in all groups
    and were particularly heavy during the 12- to 20-month period. Deaths
    did not appear to be compound-related. Growth rate, food intake and
    feed efficiency were comparable for all groups during the experimental

    period. Haematological and biochemical tests and urinalysis of test
    and control groups were similar and within normal limits. Although
    there were some decreases in the absolute organ weights of spleen and
    brain of males of the 0.16 and 0.05% groups at 20 months, these were
    not significantly different from those of control when expressed as
    organ/body weight ratio. No gross or microscopic lesions that could be
    attributed to the test compound were detected (Terhaar et al., 1968b).

        Groups of 68-70 F344/N rats/sex, following  in utero exposure,
    were given  ad libitum access to diets containing 0, 1250, 2500 or
    5000 mg/kg TBHQ (99% pure), equal in males/females to 0/0 50/60,
    110/120, and 225/240 mg/kg bw per day, respectively, for up to 30
    months after weaning. Clinical findings and individual body weights
    were recorded weekly for the first 13 weeks and every 4 weeks
    thereafter. Food consumption was recorded monthly for all exposed
    groups and weekly for the first 26 weeks and then monthly for control
    groups. An interim sacrifice of 9 or 10 randomly selected
    rats/sex/group was conducted at 3 months. At this time, complete gross
    and microscopic evaluations were conducted, the weights of the right
    kidney, liver and right testis were recorded and a standard set of
    haematological parameters was determined. A complete necropsy,
    including gross and microscopic evaluations, was performed where
    possible on rats dying during the study and all animals surviving to
    study termination.

        Survival rates were higher in high-dose rats of both sexes
    compared to respective controls, with differences reaching statistical
    significance (p<0.05) for the females. Throughout the study,
    high-dose males and females weighed on average 7 and 10% less than
    controls, respectively. Exposure to TBHQ had no effect on food
    consumption. The only clinical observation attributed to TBHQ was hair
    discoloration which was observed in all exposed groups. There was no
    evidence that TBHQ affected haematological parameters determined at 3
    months. No neoplasms were observed in either sex at the 3 month
    necropsy. There was a dose-dependent increase in the incidence of
    haemosiderin pigmentation of the spleen in female rats only. There
    were no other treatment-related histopathological findings at the 3
    month sacrifice. At terminal sacrifice, hepatocellular carcinoma was
    observed in 2/60 (3%) of high-dose males compared to none in any other
    dose group. This was probably an incidental finding. The incidence was
    not statistically significant and fell within ranges of historical
    controls reported in the literature (0-6%). No TBHQ-related
    differences in the incidence of hepatocellular adenomas or foci of
    cellular alteration were observed. There was an increased incidence of
    bilateral, interstitial cell adenoma of the testes in mid- and
    high-dose males, which was within historical control ranges reported
    in the literature and was not statistically significant. Although
    there were no differences in the incidence of C-cell or follicular
    cell adenomas of the thyroid, C-cell carcinoma occurred in 2/60 and
    follicular cell carcinoma occurred in 3/60 high-dose males compared to
    none in the control males, but this was not statistically significant.
    In female rats, the incidence of thyroid adenomas or carcinomas was
    similar for all groups. There was no thyroid gland hyperplasia

    reported in any exposed or control males or females. Literature values
    for the incidence of C-cell carcinoma and follicular carcinoma in
    historical controls are reported to be 0.5% (0-2%) and 3.8% (0-12%),
    respectively. Owing to their low incidence and the absence of
    preneoplastic lesions, the occurrence of C-cell and follicular cell
    carcinomas in high-dose males was considered unrelated to treatment.
    TBHQ treatment was associated with a decreased incidence of adenoma in
    the par distalis of the pituitary gland in males, adenoma of the
    adrenal cortex in females, mammary fibroadenoma in both sexes, and
    mammary fibroadenoma, adenoma or carcinoma combined in females. These
    decreases were attributed to the reduced body weights in the mid- and
    high-dose groups. Although in females bile duct hyperplasia occurred
    more frequently in the 5000 mg/kg group than in controls, in males
    there was a dose-related decrease in the incidence of bile duct
    hyperplasia. In females there was a dose-related decrease in the
    incidence of hepatocellular cytoplasmic vacuolization, while the
    incidence was comparable among males. The incidence of haemosiderin
    pigmentation of the spleen increased with TBHQ dose in females at the
    mid- and high-dose levels but was unaffected by TBHQ treatment in
    males. In the absence of any haematological effects, the toxicological
    significance of the pigmentation in the spleen is questionable. The
    incidence of kidney cysts was increased in mid- and high-dose males
    (2/60, 3/60, 7/58, 11/60 in 0, 1250, 2500 and 5000 mg/kg groups,
    respectively) and the incidence of suppurative inflammation was
    increased in high-dose males compared to controls (9/60, 8/60, 9/58
    and 20/60 in the 0, 1250, 2500 and 5000 mg/kg groups, respectively).
    In female rats, the incidence of chronic inflammation was increased in
    the kidneys of mid- and high-dose females (1/60, 1/60, 3/57 and 5/60
    in 0, 1250, 2500 and 5000 mg/kg groups, respectively) while there was
    no change in the incidence of suppurative inflammation. The NOEL for
    the long-term toxicity of TBHQ was 1250 mg/kg diet (equal to 50 mg/kg
    bw per day) based on an increased incidence of cysts and suppurative
    inflammation in the kidneys of males and an increased incidence of
    chronic inflammation of the kidneys of females at higher doses (NTP,
    1995).

    2.2.3.3  Dogs

        Groups of 4 beagle dogs/sex were given  ad libitum access, 1 hour
    per day, 6 days/week, to diets containing 500, 1580 or 5000 mg/kg TBHQ
    (equal to 21/22, 72/73, 260/220 mg/kg bw per day in males/females,
    respectively) for up to 117 weeks. Diets were prepared by adding 6%
    cottonseed oil, containing the appropriate concentration of TBHQ, to a
    commercial diet. An additional control group of 6 dogs/sex received
    the same commercial diet to which 6% cottonseed oil had been added.
    Dogs were individually housed and water was available  ad libitum.

        Dogs were inspected daily for appearance, behaviour, survival and
    physical signs. Body weights and food consumption were determined
    weekly for the first 12 weeks. Thereafter body weights were determined
    bi-weekly and food consumption was determined periodically. Complete
    physical examinations were conducted at various times during the test
    period. Haematological studies, including haemoglobin concentration,

    haematocrit, WBC and differentials, were conducted pre-trial and at
    weeks 0, 12, 26, 52, 78, 104 and 112. In addition, reticulocytes and
    platelets were reported at weeks 99, 104 (105) and 112 weeks and RBCs
    at week 112. At 20 months, peripheral blood smears with Wright and
    Giemsa stains were submitted to a consultant pathologist. Clinical
    biochemistry analyses included serum BUN, LDH and AP on all dogs
    during the pre-trial period and at weeks 0, 12, 26, 52, 78 and 104,
    and serum ALT and bilirubin at week 104 only. In addition, total
    protein, albumin and globulin concentrations were determined in all
    control and high-dose dogs at each time point. Urinalysis, consisting
    of specific gravity, pH, albumin, glucose, ketone bodies and occult
    blood, was conducted pre-trial and during weeks 0, 12, 26, 52, 78 and
    104. Because some abnormalities in the reticulocyte counts were noted
    at week 99, and persisted at week 104, the study was extended to 117
    weeks. At week 115, TBHQ was withdrawn from the diet of one dog/sex of
    the high-dose group. The animals were maintained in metabolism cages
    and 24-hour urine samples were collected daily. Blood samples were
    collected for haematological studies on days 1, 2, 3, 4, 7, 10 and 13
    of the withdrawal period.

        An interim sacrifice of one dog/sex/group was made after one year
    on trial. The remaining animals were sacrificed at week 117. At
    autopsy, animals were examined for gross pathological changes. The
    liver, kidneys, spleen, heart, brain, lungs, gonads, adrenals, thyroid
    and pituitary of all dogs were weighed. Tissues from the following
    organs of all dogs of control and high-dose groups were examined
    microscopically: liver, spleen, gallbladder, stomach, small and large
    intestines, pancreas, kidneys, urinary bladder, adrenals, gonads and
    adnexa, pituitary, thymus, thyroid, salivary glands, lymph nodes,
    heart, lungs, marrow, aorta, skin, muscle, spinal cord and brain. The
    liver, stomach, small and large intestines and kidneys of all dogs on
    low-and mid-level test diets were also examined microscopically. In
    addition, specimens of liver and kidney tissues were prepared for
    electron microscopy.

        No deaths occurred during the test period. Behaviour and
    appearance were normal at all times and physical examinations did not
    reveal any treatment-related problems. Body weights of high-dose
    animals of both sexes tended to be lower than controls but the
    difference was not statistically significant. There were no
    treatment-related effects on food consumption, clinical biochemistry
    or urinalysis. Haemoglobin concentrations were significantly 
    (p < 0.05) lower than those of controls in high-dose males at weeks
    52, 104 and 112 and in high-dose females at weeks 26 and 112. Compared
    to controls, haematocrits were significantly (p < 0.05) lower in
    high-dose males at weeks 52, 104 and 112. Haematocrits were also lower
    in high-dose females at weeks 52, 78 and 112 but the differences were
    not statistically significant. At weeks 99, 104/105 and 112,
    reticulocytes (%) tended to be higher in TBHQ-treated animals of both
    sexes although there was no dose relationship to the response and no
    statistics were provided. RBCs, reported for week 112 only, were
    significantly (p < 0.05) lower in high-dose animals of both sexes
    compared to their respective controls. Peripheral blood smears showed

    an increased incidence of normoblasts and erythrocyte basophilia in
    TBHQ-treated animals: normoblasts were reported in 1 mid- and
    1 high-dose male and 1 low-, 2 mid- and 1 high-dose female and
    erythrocyte basophilia in 1 high-dose male and 1 mid- and 2 high-dose
    females. Mean absolute and relative liver weights of high-dose dogs of
    both sexes tended to be higher than those of controls but only the
    difference in relative liver weight of males was statistically
    significant (p < 0.05). There was a statistically significant trend
    towards increased relative testes weights in TBHQ-treated dogs but the
    high-dose groups did not differ significantly from controls in this
    respect. In females, relative kidney weights tended to be higher in
    treated dogs but only in the mid-dose group were kidneys weight
    significantly (p < 0.05) greater than those of controls. There were
    no other treatment-related effects on organ weights. Gross pathology
    and histopathology failed to reveal any changes that could be
    attributed to TBHQ. Electron microscopy of liver and kidney showed
    normal cellular constituents in test animals. There was no increase in
    the endoplasmic reticulum in liver cells of treated animals. Since the
    organ weights were reported to be within historical control ranges and
    since the observed differences in organ weights were not accompanied
    by any histopathological findings, they may simply reflect differences
    in body weights. Under these study conditions, the NOEL for long-term
    toxicity in the dog was 1580 mg/kg diet (equal to 72 mg/kg bw per day)
    based on decreased haemoglobin and/or haematocrits at various time
    points throughout the study and decreased RBC counts at 112 weeks in
    high-dose dogs of both sexes. Although reticulocyte counts were
    increased and erythrocyte precursors were present at the 500 and 1580
    mg/kg dose levels, the RBC's haematocrit and haemoglobin concentration
    were unaffected at these dose levels (Eastman Chemical Products,
    1968b).

    2.2.4  Reproductive toxicity studies

        Albino Holtzman rats were fed ground Purina chow containing 0 or
    0.5% TBHQ in a two-generation study using one litter per generation.
    After 36 days on treatment, groups of 24 female and 10 male F0 rats
    were mated, 3 females with 1 male, until 10 females were inseminated.
    At 100 days of age, F1 animals were mated, 15 females with 5 males,
    until 8 TBHQ-treated females were inseminated. In both generations,
    body weights and feed consumption were recorded weekly during the
    pre-mating period. With regards to mating, parturition and weaning,
    the following data were recorded: mating index, fertility index,
    gestation index, gestation period, average litter size, number of live
    and dead births, and pup survival to weaning. Pup weights were
    recorded at weaning and at 1 and 2 weeks post-weaning. Adults from the
    F0 and F1 generations were necropsied and liver and kidney weights
    reported. Poor survival in the F1 generation and their offspring was
    attributed to a serious outbreak of pneumonia in the colony. Parental
    animals in both F0 and F1 generations tended to weigh less than
    their respective controls although only in the F1 females was the
    difference statistically significant. In both generations,
    TBHQ-treated animals had lower feed efficiencies than the controls.

    TBHQ had no effect on the mating, fertility or gestation indices,
    average litter size, or the number of live births in either
    generation. In the both generations, the percentage of live pups
    surviving to weaning was comparable for the TBHQ and control groups,
    although for both groups survival of the F2 pups was low (31-41%).
    The low survival was attributed to pneumonia. The pups in the TBHQ
    group weighed less than controls at weaning and at 1 and 2 weeks
    post-weaning. The depression in weight gain was similar in both
    generations. The absolute kidney weights of TBHQ-treated females in
    the F1 generation were significantly lower than those of controls,
    but when kidney weights were expressed relative to body weight they
    were comparable. There were no treatment-related effects on absolute
    or relative liver weights. Histopathological examination revealed no
    cellular changes attributable to TBHQ (Fassett  et al., 1965).

        TBHQ was fed in the diet to groups of 15 male and 15 female SD
    rats for three successive generations at levels of 0 or 0.5%. Pairs of
    rats were mated to produce two litters per generation, with the next
    generation selected from weanlings of the second litter. Data recorded
    in each were as follows: number of inseminations, number of
    pregnancies, gestation period, average litter size, and mortality of
    young from birth to weaning, weaning to one week after weaning, and
    one week and 2 weeks after weaning to sacrifice. The average body
    weight per pup at weaning, one week after weaning, and 2 weeks after
    weaning was also recorded. Tissues were collected from all breeders.
    Animals that were not used as breeders were sacrificed at 7 weeks and
    those selected as possible breeders, but not used, at 14 weeks. All of
    the pups in the b generation were sacrificed and autopsied at
    approximately 7 weeks of age. All animals were examined for gross
    pathology, and micropathology was studied on at least four animals of
    each litter. Organs examined included trachea, lung, heart, tongue,
    oesophagus, stomach, small and large intestines, liver, kidneys,
    urinary bladder, pituitary, adrenal, pancreas, thyroid, parathyroid,
    gonads, uterus, spleen, bone marrow, cerebrum, cerebellum and eye. To
    terminate the study F3b litters were delivered by uterotomy on the
    nineteenth day of gestation. Fetuses were examined for gross
    abnormalities. One-third were stained with alizarin red for skeletal
    defects, while the other two-thirds, after fixation, were sectioned
    freehand and examined for abnormalities. The percentage of
    inseminations and pregnancies, gestation and average litter size
    appeared normal for both matings of each generation. In the F1a and
    F2a litters there were more deaths than in controls in the birth to
    weaning period. The effect was not observed in the subsequent Flb and
    F2b generations. The parent rats (F0 generation) of the test group
    ate less and showed lower weight gain than their respective controls.
    Body weight of pups from treated animals at various time periods from
    weaning were lower than those of the control. Deaths during the
    weaning to 5 weeks period were always more frequent in treated group
    (deaths as a percentage of the total number of animals). Abnormalities
    were reported in 22 fetuses from the F3b generation but 13 of these
    were in the control group. Minor skeletal changes were noted in two
    animals in the test group. No compound-related histological

    abnormalities were observed (Terhaar & Krasavage, 1968a).

        In another study, groups each of 20 male and 20 female SD rats
    were maintained on diets containing TBHQ at concentrations of 0,
    0.015, 0.15 or 0.5% along with Mazola Corn Oil at 5% (w/w)
    incorporated into a basic diet of Purina Chow. Diets were fed for 66
    days prior to breeding. Rats of the same dose level were mated (1:1)
    to produce two groups of first generation litters (F1a and F1b). F1a
    litters were maintained on assigned diets. Data recorded included
    number of inseminations and pregnancies, gestation period, litter
    size, mortality of young at birth, birth to weaning, weaning to one
    week postweaning and one week post-weaning to 2 weeks post-weaning.
    Litters of the F1b generation were treated similarly to F1a litters
    up to the tenth day after birth. At day 10, litters and dams within
    test groups were paired, and one pair was placed on control diet while
    others remained on the test diet. Also half the litters and dams of
    control groups were allowed to remain on the control diet and half on
    the high-level TBHQ diet (0.5%). At 5 weeks of age, pups were
    sacrificed and examined for gross pathology. A number of deaths of
    parent animals occurred during the study (one high-dose male, one
    control group during first 5 days, and later three males, one at the
    low, one at the mid and one at the high-dose level, and one female
    control). None of the deaths appeared to be compound-related. Food
    intake of parent rats was similar to that of controls except at the
    0.5% level, where there was a slight decrease at the commencement of
    the test. Male rats in this group showed a slightly decreased weight
    gain compared to control. There was no apparent effect on gonadal
    function, estrus cycle, mating behaviour, conception rates, gestation
    period, parturition and lactation during the two breedings. The
    average litter size, neonatal viability and growth of the new-born
    pups appeared normal. Autopsy of F1b pups failed to reveal any gross
    pathology. The average litter weights of the F1a test pups were
    similar to those of controls. Pups from the F1b control changed to
    the 0.5% TBHQ showed a somewhat lower body weight than those left on
    control diet for the period of study (up to 2 weeks past weaning).
    This may have been related to rejection of the diet (Krasavage &
    Terhaar, 1970).

        Groups of 16 female F334/N rats (F0 generation) received
     ad libitum access to diets containing 0, 2500, 5000, 10 000, 20 000
    or 40 000 mg/kg TBHQ (99% pure), equivalent to 125, 250, 500, 1000 and
    2000 mg/kg bw per day, from 2 weeks prior to cohabitation until
    weaning of the F1 pups. Dams exposed to 20 000 or 40 000 mg/kg did
    not litter. At the 2500, 5000 and 10 000 mg/kg dose levels, TBHQ did
    not affect gestation length, litter size, number of dams with
    stillborn pups or pup weight on day 4 of lactation. Pup survival was
    significantly (p<0.01) lower in the 10 000 mg/kg group than in
    controls at day 4 of lactation and at weaning (day 28). At 5000 mg/kg,
    pup survival at weaning was lower than for controls but the
    differences were not statistically significant. Rats selected from the
    F1 offspring were used in the 13-week short-term toxicity study
    described in section 2.2.2.2 (NTP, 1995).

    2.2.5  Special studies on teratogenicity

        Groups of 20 female SD rats were fed a basal diet containing 0,
    0.125, 0.25 or 0.5% TBHQ from days 6 to 16 of gestation. During the
    mating period and on all other days of gestation, all treatment groups
    received control diet only. The experiment was terminated on day 20 of
    gestation. Total TBHQ doses of 970, 1880 or 3600 mg/kg bw had no
    effect on the mean body weight gain or feed consumption of the dams.
    The average number of corpora lutea, implantation sites, viable
    fetuses, resorptions, fetal body weights and mortality did not differ
    between the control and treatment groups. A significant number of
    skeletal variations (rudimentary ribs) were seen in all groups, but
    the incidence of these variations was two-fold greater in the control
    group than in any treatment group. It was concluded that TBHQ was not
    teratogenic in rats at any of the doses employed (Krasavage, 1977).

        BHA and its metabolites, including TBHQ and TBQ, were assayed
    using cell culture methods for assessing potential teratogenicity. In
    the rat embryonic cell differentiation assay, TBQ >TBHQ > BHA had
    dose-related inhibitory effects on the differentiation of both
    limb-bud and midbrain cells. TBQ was also the most potent inhibitor in
    the human embryonic palatal mesenchymal cell-growth assay. In this
    assay, TBHQ was a more potent inhibitor than BHA but less potent than
    TBQ (Tsuchiya  et al., 1988).

    2.2.6  Special studies on genotoxicity

        The results of genotoxicity studies with TBHQ are summarized in
    Table 2.

        An alkaline elution assay was conducted with DNA extracted from
    forestomach epithelium of male F344 rats that had received an oral
    dose of 40 mg/ animal of 1.0% BHA (220 mg/kg bw), 0.001% TBHQ (0.22
    mg/kg bw), 0.001% TBQ (0.22 mg/kg bw) or 0.0 0001% TBQ and sacrificed
    3 hours later. The BHA or TBHQ treatments caused no detectable DNA
    damage, but TBQ (an oxidation product of TBHQ) treatment induced
    dose-related DNA damage. The elution rate for 0.001% TBQ was
    significantly higher than for BHA or controls and was similar to that
    induced by 15 mg/kg bw MNNG in rat forestomach epithelial cells. The
    linearity of the elution pattern suggested that DNA damage was not due
    to cell necrosis at a TBQ dose of 0.22 mg/kg bw. TBQ appeared to be
    cytotoxic to the forestomach epithelial cells of rats at doses of 1.0%
    (220 mg/kg bw) (Morimoto  et al., 1991).



        Table 2.  Results of genotoxicity assays on TBHQ

                                                                                                                         

    Test system               Test object                Concentration of            Results          Reference
                                                         TBHQ
                                                                                                                         

    Point mutation

    Ames test1                Salmonella                 < 5.0 µg/ml (TA1535)        Negative         Societé Kemin 
                              typhimurium                < 15 µg/ml (TA1537)                          Europa, 1982a
                              TA98, TA100,               < 50 µg/ml (other 
                              TA1535, TA1537,            strains)
                              TA1538

    Ames test1                S. typhimurium             < 450 µg/plate (-S9)        Negative         Mueller & Lockhart, 
                              TA98, TA100,               < 2700 µg/plate (+S9)                        1983
                              TA1535, TA1537,
                              TA1538

    Ames test1                S. typhimurium             1-1000 µg/plate             Negative         Hageman et al., 
                              TA97, TA100,                                                            1988
                              TA102, TA104

    Ames test2                S. typhimurium             3-3333 µg/plate             Negative         Zeiger et al., 
                              TA97, TA98,                                                             1992
                              TA100, TA102

    Ames test1                S. typhimurium             0.5-10 µg/plate (-S9)       Negative         Matsuoka et al., 
                              TA97, TA98,                0.5-250 µg/plate (+S9)                       1990
                              TA100, TA102

    Reverse mutation          Saccharomyces              100-500 µg/ml (-S9)         Negative         Rogers et al., 
    in yeast1                 cerevisiae D7              50-200 µg/ml (+S9)                           1992

    Table 2.  (continued)

                                                                                                                         

    Test system               Test object                Concentration of            Results          Reference
                                                         TBHQ
                                                                                                                         

    In vitro mammalian        Mouse lymphoma cells       < 31.3 µg/ml                Positive         Litton Bionetics, 
    point mutation assay1     line L5178Y (TK+/-)                                    (+S9 only)       1982a

    In vitro mammalian        CHO cells/HGPRT            < 6 µg/ml (-S9)             Negative         Beilman & Barber, 
    point mutation assay1     locus                      < 250 µg/ml (+S9)                            1985

    In vitro mammalian        Chinese hamster V79        0.17-3.40 µg/ml             Negative4        Rogers et al., 1992
    point mutation assay3     cells, HGPRT locus

    Clastogenic effects and chromosomal aberrations

    In vitro chromosomal      Chinese hamster V79        < 330 µg/ml                 Positive         Societé Kemin Europa, 
    aberration1               cells                                                  (-S9 only)       1982c (N/A for 
                                                                                                      re-evaluation)

    In vitro chromosomal      Chinese hamster ovary      15-62 µmol/litre            Positive5        Phillips et al., 
    aberration1               cells                                                  (-S9 only)       1989

    In vitro chromosomal      Chinese hamster ovary      5-25.2 µg/ml (-S9)          Positive6        NTP, 1995
    aberration1               cells                      100.5-300 µg/ml             (+S9 only)
                                                         (+S9)

    In vitro chromosomal      Chinese hamster lung       12.5-50 µg/ml (-S9)         Positive1,7      Matsuoka et al., 
    aberration1               fibroblast cells           20-40 µg/ml (+S9)           (+S9 only)       1990

    In vivo chromosomal       Mouse bone marrow          < 200 mg/kg bw i.p.         Negative         Litton Bionetics, 
    aberration                                                                                        1985

    In vivo chromosomal       Mouse bone marrow          200 mg/kg bw i.p.           Positive8        Giri et al., 1984
    aberration

    Table 2.  (continued)

                                                                                                                         

    Test system               Test object                Concentration of            Results          Reference
                                                         TBHQ
                                                                                                                         

    In vivo chromosomal       Mouse bone marrow          2 mg/kg bw per day,         Positive9        Giri et al., 1984
    aberration                                           30 days in diet

    Mouse micronucleus        Mouse bone marrow          162, 325 or 650             Negative         Litton Bionetics, 
    assay                                                mg/kg bw, p.o.                               1982b

    Mouse micronucleus        Mouse bone marrow          250 mg/kg bw, p.o.          Positive10       Societé Kemin Europa, 
    assay                                                                                             1982b (N/A for 
                                                                                                      re-evaluation)

    Mouse micronucleus        Mouse bone marrow          9.38-300 mg/kg bw           Negative         NTP, 1995
    assay                                                per day
                                                         3 days, i.p.

    Dominant lethal assay     Sprague-Dawley rats        < 565 mg/kg bw per          Negative         Krasavage & Farber, 
                                                         day, 83 days                                 1983

    DNA interactions

    Mitotic gene              Saccharomyces              100-500 µg/ml (-S9)         Negative         Rogers et al., 1992
    conversion1               cerevisiae D7              50-200 µg/ml (+S9)

    Sister chromatid          Chinese hamster ovary      0.5-16.7 µg/ml (-S9)        Positive         NTP, 1995
    exchange1                 cells                      5-166.7 µg/ml (+S9)         (+S9 only)

    Sister chromatid          Chinese hamster V79        0.17-3.4 µg/ml              Negative4        Rogers et al., 1992
    exchange3                 cells

    Sister chromatid          Mouse bone marrow          0.5-200 mg/kg bw,           Positive         Mukherjee et al., 
    exchange                                             i.p.                                         1989
                                                                                                                         

    Table 2.  (continued)

    1    Both in the absence and presence of a metabolic activation system derived from rat liver S9 fraction.
    2    Both in the absence and presence of a metabolic activation system derived from rat or hamster liver S9 
         fraction.
    3    Cultures with or without added rat or hamster hepatocytes.
    4    The highest dose tested did not achieve a 50% reduction in cloning efficiency.
    5    The number to aberrations and the cytotoxicity of TBHQ was shown to be dependent on the cell density on 
         the plate. Both the cytotoxicity and number of aberrations were reduced by co-administration of catalase. 
    6    The studies with metabolic activation were inadequate because only the lowest dose in one study provided 
         sufficient numbers of cells for scoring. All other doses were cytotoxic, and <25 cells were scored. 
    7    The oxidative metabolites of TBHQ, TBQ and TBQ epoxide-induced chromosomal aberrations in this system 
         in the presence (TBQ and TBQ epoxide) and absence (TBQ epoxide) of S9 metabolic activation.
    8    This study was inadequate as only one dose level was used and this dose appeared to be cytotoxic.
    9    The repeated dosing protocol used was inappropriate for this assay.
    10   Positive at 24-hour harvest only (not at 48 or 72-hour harvests).
    


        BHA and its metabolites, TBHQ, TBQ and  tert-butylquinone oxide
    (BQO) were tested for cytotoxicity and micronucleus induction in
     in vitro assays with Chinese hamster lung (CHL) cells and three CHL
    sub-cell lines. The sub-cell lines included R-8, a H2O2-resistant
    cell line with increased catalase activity, MM-1, a
    2-memadione-resistant cell line with reduced P450 reductase activity,
    and MN-7 with increased superoxide dismutase activity. TBHQ, in the
    presence of S-9, induced micronuclei, and the potency of induction in
    each cell line corresponded its cytotoxicity (Suzuki  et al., 1991,
    abstract only).

        The potential of BHA, TBHQ and TBQ to induce oxidative DNA damage
    were studied by measuring biological inactivation of single-stranded
    bacteriophage PhiX-174 DNA and the formation of
    7-hydro-8-oxo-2'-deoxyguanosine (8-oxodG) from dG by these compounds,
     in vitro, in the absence and presence of peroxidases. TBHQ, but not
    BHA or TBQ, appeared to be a strong inducer of DNA damage as indicated
    by a strong inactivation of phage DNA and a potent induction of 8-
    oxodG formation. This damage was shown to be due to the formation of
    superoxide anion, hydrogen peroxide and hydroxyl radicals. The lack of
    activity of the quinone metabolite was attributed by the authors to a
    lack of cytochrome P-450 reductase  in vitro (Schilderman et al.,
    1993b).

        TBQ was shown to be 7-8 times more cytotoxic to Chinese hamster
    V79 cells than TBHQ and about 100 times more cytotoxic than BHA. At
    dose levels as high as 0.6 µg/ml, TBQ did not increase consistently
    the frequency of sister-chromatid exchanges in V79 cells or the
    frequency of mutation to thioguanine resistance at the  hgprt gene
    locus. In the D7 strain of  Saccharomyces cerevisiae, TBQ tended to
    produce small increases in the frequencies of gene convertants and
    reverse mutations (Rogers  et al., 1993).

        TBHQ induced strand breaks in double-stranded PhiX-174 relaxed
    form I DNA in the presence of micromolar concentrations of copper. The
    induced DNA strand breaks were inhibited by a Cu(I) chelator or by
    catalase, indicating that a CuII/CuI redox cycle and H2O2
    generation were requirements for the observed DNA damage. The authors
    concluded that DNA-associated copper in cells may have the potential
    to activate phenolic compounds, producing reactive oxygen and
    electrophilic phenolic intermediates capable of inducing a spectrum of
    DNA lesions (Li & Trush, 1994).

        The mechanisms by which hydroquinone and TBHQ induce chromosomal
    loss and breakage were investigated in Chinese hamster V79 lung
    fibroblast cells containing prostaglandin H synthetase. A
    cytokinesis-block micronucleus assay was employed using the CREST
    antibody which recognizes a centromeric protein to distinguish
    micronuclei formed by loss of whole chromosomes from those resulting
    from chromosome breakage. TBHQ induced chromosomal loss and breakage
    and these genotoxic effects were shown not to be dependent on
    activation mediated by prostaglandin H synthetase. The genotoxicity of

    TBHQ appeared to be mediated through an oxidation product that could
    not be identified. The oxidation product was not the result of
    autoxidation since it was minimal under the test conditions.
    Introduction of catalase into the media reduced chromosomal breakage
    resulting from TBHQ but had no effect on chromosomal loss. These data
    suggested that oxygen radical-derived species were unlikely to
    contribute to chromosomal loss and were only partly responsible for
    the chromosomal breakage resulting from TBHQ treatment. The authors
    suggested that the chromosomal loss observed following treatment with
    TBHQ might result from the quinone or semiquinone metabolites binding
    to protein critical in microtubule assembly and spindle formation
    (Dobo & Eastmond, 1994).

        The potential of BHA and it primary metabolites TBHQ and TBQ to
    induce oxidative DNA damage and cell proliferation in human
    lymphocytes cultured  in vitro was investigated. Analysis of the
    culture medium and lysed cell fractions indicated that TBHQ was
    actively metabolized in whole blood. No conjugation of TBHQ to
    glucuronic acid or sulfate was observed. Addition of BHA, TBHQ and TBQ
    to lymphocytes resulted in a dose-dependent increase in cytotoxicity.
    TBHQ appeared more cytotoxic than TBQ and TBQ more than BHA. At
    non-cytotoxic doses, TBHQ induced a dose-dependent increase in cell
    proliferation as estimated by incorporation of BrdU. TBHQ was also
    shown to induce the formation of 8-oxodG at non-cytotoxic doses. Of
    the compounds tested, TBHQ appeared to be the best inducer of both
    cell proliferation and 8-oxodG formation. These effects were inhibited
    by co-administration of acetylsalicylic acid, which inhibits
    prostaglandin H synthetase, thus preventing oxidation of TBHQ to
    2- tert-butyl-semiquinone (Schilderman  et al., 1995).

        Oxidative DNA damage caused by BHA and TBHQ was evaluated by
    measuring the formation of 8-oxodG in DNA from calf thymus. TBHQ
    induced the formation of 8-oxodG in a concentration-dependent manner
    and the induction was strongly stimulated by CuCl2. Chelators of CuI
    or CuII and catalase inhibited the TBHQ induction of 8-OHdG formation
    in the presence or absence of CuCl2. The authors suggested that under
    aerobic conditions CuII/CuI redox cycling appears to contribute to
    the formation of 8-oxodG by TBHQ through the generation of oxygen
    radical species (Nagai  et al., 1996).

    2.2.7  Special studies on lung toxicity

        TBHQ was tested for its potential to produce lung damage in mice
    similar to that seen following administration of BHT. Groups of 10
    mice were given single i.p. injections of 63, 125, 250 or 500 mg/kg bw
    TBHQ, or 300, 625, or 1230 mg/kg bw BHT in corn oil. After 5 days, all
    animals were necropsied and the lungs were weighed and examined
    histomorphologically. TBHQ led to mortality at doses of 125 mg/kg bw
    (4/10), 250 mg/kg bw (9/10) and 500 mg/kg bw (10/10). Two of the mice
    that received 1230 mg/kg bw BHT died before the end of the observation
    period. Body weights as well as absolute and relative organ weights
    were comparable in all groups. While BHT produced hyperplasia of

    pulmonary pneumocytes, TBHQ did not lead to any treatment-related lung
    lesions (Krasavage & O'Donoghue, 1984).

    2.2.8  Special studies on the forestomach

    2.2.8.1  Rats

        TBHQ was incorporated into the diet of weanling male Fischer 344
    rats at dose levels of 0.25 or 1%. After nine days on the test diet,
    and one hour before sacrifice, the test animals were injected (i.p.)
    with 0.25 µCi/g methyl-3H-thymidine. The thymidine-labelling index in
    the pre-fundic region of the stomach was determined using a standard
    technique. At the high dose there was a significant increase in the
    labelling index (p<0.002). The increase was about 50% of that
    obtained with rats treated with 0.5% BHA under similar conditions.
    Histopathologically, 1% TBHQ led to hyperplasia of the basal cell
    layer in the forestomach epithelium; no changes were observed at the
    low dose (Nera et al., 1984).

        Groups of 5 to 10 Wistar rats were fed diets containing 2% BHA, 2%
    TBHQ, 1% BHT or other structurally related compounds for 28 days. BHA
    dosing resulted in severe diffuse hyperplasia, acanthosis and
    hyperkeratosis in the forestomach mucosa, which was most pronounced in
    the vicinity of the limiting ridge. Keratinous masses were also
    observed particularly near the glandular stomach. In TBHQ-treated
    animals brownish discolorations and mild hyperplasia of the
    forestomach mucosa with focally increased hyperplasia of basal cells
    were observed. There were no visible forestomach lesions in the
    BHT-treated animals. There did not appear to be any sex-related
    differences in the extent of the forestomach lesions. None of the
    treatments resulted in any changes in the oesophagus or glandular
    stomach. The authors concluded that the induction of forestomach
    lesions is not limited to BHA but may be induced by other structurally
    related compounds. They suggested that the methoxy group in the para
    position was important in the hyperplasiogenic activity (Altmann 
     et al., 1985).

        In a study designed to investigate the hyperplastic activity of
    BHA and related phenols on the rat forestomach, groups of 5-10 male
    and female Wistar rats were fed powdered diets containing BHA, BHT,
    TBHQ or one of 8 structural analogues for 90 days. TBHQ added to the
    diet at a concentration of 2% led to brownish discolorations and mild
    hyperplasia of basal ceils. The local basal cell hyperplasia did not
    tend to differentiate. The hyperplastic activity of TBHQ was however
    considerably lower than that of BHA (Altmann  et al., 1986).

        In a study to examine forestomach cell proliferation, F334 male
    rats received 2% dietary TBHQ alone or in combination with 0.3% sodium
    nitrite (a promoter of forestomach carcinogenesis) in the
    drinking-water, and/or 1% sodium ascorbate in the diet for 4 weeks.
    TBHQ alone, or in combination with sodium ascorbate, had no
    significant effect on mucosal height in the pre-fundic or mid-region

    of the forestomach. Co-administration of sodium nitrite with TBHQ
    resulted in an increase in the height of the forestomach mucosa in
    both regions compared to those receiving TBHQ, sodium nitrite or basal
    diet alone. The mucosal height in the forestomach of rats treated with
    TBHQ, sodium nitrite and sodium ascorbate was significantly greater
    than that in rats receiving TBHQ or basal diet alone but comparable to
    that in rats receiving sodium nitrite and sodium ascorbate combined
    (Yoshida  et al., 1994).

        Groups of male F344 rats received basal diet or basal diet
    supplemented with 2% of various phenolic compounds, one of which was
    TBHQ, with or without 0.3% sodium nitrite in the drinking-water for 4
    weeks. They received single i.p. injections of 20 mg/kg bw BrdU one
    hour before being killed at the end of the study. In the absence of
    sodium nitrite, TBHQ resulted in a significant increase in the
    thickness of the forestomach mucosa in the pre-fundic or mid-regions
    compared controls. Additional treatment with sodium nitrite enhanced
    the forestomach cell proliferation in rats receiving TBHQ by more than
    10 times that observed in rats recieving TBHQ or sodium nitrite alone.
    In the glandular stomach, TBHQ alone significantly increased the
    thickness and labelling indices compared to unsupplemented controls. A
    further increase in both these parameters was observed in rats
    receiving both TBHQ and sodium nitrite. Alone, TBHQ and sodium nitrite
    had no effect on the mucosal thickness of the oesophagus, but in
    combination there was a significant increase in thickness. The
    labeling indices were unaffected by TBHQ or sodium nitrite, alone or
    in combination (Kawabe  et al., 1994).

    2.2.8.2  Hamsters

        Syrian golden hamsters (16 males/group) were fed a powdered basal
    diet containing 0.5% TBHQ (purity >98%) for 20 weeks. This dose is
    approximately one quarter of the LD50. The control group received a
    basal diet only, while 12 other groups received diets containing one
    of 12 other phenolic compounds in concentrations corresponding to one
    quarter of their respective LD50. At the end of the experiment,
    animals were killed and organs were fixed. Five sections each were cut
    from the anterior and posterior walls of the forestomach, two from the
    glandular stomach, and four from the urinary bladder. Tissues were
    processed for histopathology and auto-radiography. Analysis of the
    labelling index was made on 4000 cells of urinary bladder epithelium,
    3000 cells of pyloric gland epithelium (1000 cells each of fundic
    side, middle portion, and pyloric side) and 2000 basal cells of the
    forestomach epithelium. Unlike some of the other phenolic compounds,
    TBHQ did not induce hyperplasia or tumorous lesions of the
    forestomach, the glandular stomach or the urinary bladder.
    Furthermore, it did not increase the labelling index in the tissues
    investigated (Hirose  et al., 1986).

    2.2.9  Special studies on the liver

        Hepatocytes were isolated from male F344 rats and incubated at
    37°C at a concentration of 106 cells/ml with 0.5 mM TBHQ. Incubation
    with TBHQ resulted in 100% cell death between 1 and 2 hours following
    its addition to the medium. A drop in intracellular GSH to
    undetectable levels was observed in the first hour, prior to the
    increase in cell death. Decreases in ATP and reduced protein thiol
    concentrations were observed concurrently with the increase in cell
    death. Although superoxide anion radicals were generated by
    autooxidation of TBHQ, the level of intracellular malondialdehyde was
    not increased during the incubation period (Nakagawa & Moldéus, 1992).

        The cytotoxic effects of BHA and its metabolites, TBHQ and
    3- tert-butyl-4,5-dihydroxyanisole (BHA-OH), were investigated in
    hepatocytes isolated from male F344 rats. At a concentration of 0.5
    mM, TBHQ caused a time-dependent depletion of intracellular levels of
    GSH, protein thiols and adenine nucleotides, which consistently
    preceded cell death. Pre-treatment of the hepatocytes with
     N-acetylcysteine, which facilitates the biosynthesis of GSH,
    ameliorated the toxicity caused by TBHQ. In isolated hepatic
    mitochondria, TBHQ impaired respiration related to oxidative
    phosphorylation. The authors concluded that the depletion of protein
    thiols and ATP and the impairment of mitochondrial respiration were
    the main mechanism in the toxicity of TBHQ (Nakagawa  et al., 1994).

        A database consisting of 100 chemicals tested for the ability to
    enhance the formation of hepatic GST-positive preneoplastic lesions
    was analysed by the CASE structure-activity relational system. TBHQ
    was included in the database and, based on its structure, was not
    predicted to induce GST-positive preneoplastic lesions in rat liver
    (Sakai  et al., 1994).

        The effects of dicoumarol, an inhibitor of DT-diaphorase, on the
    cytotoxicity of TBHQ were studied in hepatocytes isolated from male
    F344 rats. Addition of TBHQ (0.5 mM) to hepatocytes resulted in
    depletion of intracellular ATP, GSH and protein thiols and a
    time-dependent cell death. Treatment of hepatocytes with dicoumarol
    alone (30 µM) did not affect cell viability or cellular levels of ATP,
    GSH or protein thiols. Pretreatment of hepatocytes with dicoumarol
    prior to TBHQ exposure enhanced the cytoxicity of TBHQ. The author
    suggested that dicoumarol enhanced TBHQ-induced cytotoxicity via the
    inhibition of protective redox cycling of the hydroquinone, and the
    DT-diaphorase plays a protective role in the onset of the toxicity
    caused by TBHQ (Nakagawa, 1996).

    2.2.10  Special studies on the kidney and urinary bladder

        The effects of TBHQ on urine composition, bladder epithelial
    morphology and DNA synthesis was studied in comparison with other
    antioxidants and bladder tumour promoters. Groups of 10 male Fischer
    344 rats, 5-weeks old, were given powdered basal diet containing 2%

    TBHQ or one of 11 other compounds, or basal diet only (controls); two
    further groups received two other tumour promoters in their
    drinking-water. Five rats in each group were killed after 4 weeks for
    estimation of DNA synthesis levels and histopathological examination
    by light microscopy. The remaining rats were killed at week 8 for
    light and scanning electron microscopic examination of the urinary
    bladder. During week 4, fresh urine specimens were obtained from rats
    in each group and analysed for pH as well as electrolyte content. TBHQ
    brought about an elevation of DNA synthesis in the urothelium and
    produced morphological surface alterations such as the formation of
    pleomorphic or short, uniform microvilli and ropy or leafy
    microridges. The ability to induce proliferation and cell surface
    alterations was common to all bladder tumour promoters investigated.
    TBHQ also caused an increase in urinary pH, and a decrease in
    potassium and phosphate contents as well as in osmolality (Shibata 
     et al., 1989).

        A study was performed to investigate early proliferation-related
    responses of the renal pelvic epithelium in response to bladder tumour
    promoters. Groups of 10 male F344 rats received control diet or 2%
    TBHQ. At week 4, the DNA-labelling index of the renal pelvic
    epithelial cells was determined from 1000 cells in 5 rats/group. At
    week 8, kidney sections were prepared for scanning electron
    microscopic (SEM) examination. At the end of the study, body weights
    of the treated animals were statistically significantly lower than for
    controls as demonstrated previously by these investigators (Shibata
     et al., 1989; 25% at 4 weeks and 15% at 8 weeks). The mean
    DNA-labelling index in the renal pelvic epithelium after 4 weeks
    treatment was 10-fold higher than in controls, but without statistical
    significance. Slight cell surface alterations were observed by SEM
    after 8 weeks of treatment in some of the treated rats (Shibata 
     et al., 1991).

        The effects of three GSH conjugates of TBHQ, 
    2- tert-butyl-5-glutathion- S-ylhydroquinone (5-GSyl-TBHQ),
    2- tert-butyl-6-glutathion- S-ylhydroquinone (6-GSyl-TBHQ) and
    2- tert-butyl-3,6-bisglutathion- S-ylhydroquinone 
    (3,6-bis-GSyl-TBHQ), on the kidney and bladder of male F344 rats were
    investigated 19 hours after a single treatment. Administration of 400
    µmol/kg (i.v.) of 5-GSyl-TBHQ or 3,6-bis-GSyl-TBHQ resulted in a
    2-fold increase in the urinary excretion of gamma-glutamyl
    transpeptidase and AP. 3,6-bis-GSyl-TBHQ (200 µmol/kg, i.v.) was the
    most potent of the GSH conjugates and produced significant increases
    in the urinary excretion of gamma-glutamyl transpeptidase, AP, LDH and
    glucose (2-, 2-, 22- and 11-fold increases, respectively). The degree
    of nephrotoxicity observed with the light microscope correlated with
    the changes in biochemical parameters. In addition to nephrotoxicity,
    3,6-bis-GSly-TBHQ increased bladder weight and caused severe
    haemorrhaging of the bladder. Additional studies demonstrated that GSH
    conjugation does not impair the redox activity of TBHQ. The authors
    suggested that the cytotoxicity of GSH conjugates of TBHQ to kidney
    and bladder may contribute to the tumour-promoting effect of BHA and
    TBHQ in these tissues (Peters  et al., 1996b).

    2.2.11  Special studies on potentiation and inhibition of cancer

        The effects of TBHQ and seven other antioxidants on
    7,12-dimethylbenz[ a]-anthracene (DMBA)-initiated mammary gland, ear
    duct, and forestomach carcinogenesis were examined in female SD rats.
    Groups of 20 rats were given a single dose of 25 mg/kg bw of DMBA in
    0.5 ml of sesame oil by stomach tube at 50 days of age. Starting one
    week later, rats were given a basal diet containing 0.8% of TBHQ or
    one of the other antioxidants for 51 weeks. Controls received basal
    diet only. Groups of 15 rats served as carcinogen-free controls and
    received the different diets without prior treatment with DMBA. Groups
    receiving antioxidants had reduced body weights at the end of the
    experiment. Histological examinations revealed a reduced rate of
    mammary tumour development in TBHQ-treated and in DMBA-initiated rats
    as compared to DMBA-treated controls. The incidence of ear duct and
    forestomach tumours was not affected by TBHQ treatment (Hirose 
     et al., 1988).

        The effects of dietary TBHQ were tested in a multi-organ
    carcinogenesis model. Groups of 20 male F344 rats were given a single
    intragastric administration of 100 mg/kg bw MNNG, a single
    intragastric administration of 750 mg/kg bw EHEN, two s.c. injections
    of 0.5 mg/kg bw MBN and four s.c. injections of 40 mg/kg bw DMH. At
    the same time, the rats received 0.1% DBN for 4 weeks, followed by
    0.1% DHPN for 2 weeks in the drinking-water for a total carcinogen
    exposure period of 6 weeks. Three days after this regime, the rats
    received 1% TBHQ in the diet or control diet for 36 weeks. Control
    groups of 10 or 11 animals received 1% TBHQ alone or basal diet alone.
    The final body weights of both TBHQ-treated groups were significantly
    lower than those of respective controls (19% for carcinogen-treated
    control and 9% for basal diet control) and this was reflected in
    higher relative liver and kidney weights. Dietary TBHQ following
    carcinogen treatment reduced the incidence and multiplicity of colon
    carcinomas and had slightly reduced the incidence and multiplicity of
    some preneoplastic and neoplastic lesions of the kidney. At the same
    time, this treatment increased the incidence of oesophageal and
    forestomach papillomas and oesophageal papillary or nodular
    hyperplasia compared with controls. The treatment had no effect on the
    tongue, glandular stomach, duodenum, small intestine, liver, lung,
    urinary bladder or thyroid gland (Hirose  et al., 1993).

    2.2.11.1  Liver

        Groups of 10 SD rats were gavaged with sodium nitrite (125 mg/kg
    bw) and dimethylamine (1000 mg/kg bw) followed immediately by 0, 25,
    75 or 225 mg/kg bw of TBHQ. Ascorbate at 200 mg/kg bw was used as a
    positive control. The nitrosamine-forming mixture alone produced
    extensive hepatic necrosis and increases in serum AST, ALT and
    ornithine carbamoyl transferase. Enzyme induction was completely
    suppressed by co-administration of ascorbate. At the highest dose,
    TBHQ gave 60% protection against necrosis and appreciably suppressed
    enzyme activity increases. No significant protective effect was
    observed at doses of 25 or 75 mg/kg bw TBHQ (Astill & Mulligan, 1977).

        Groups of male Fischer rats were initially given a single dose of
    diethylnitrosamine and, starting 2 weeks later, were treated with a
    diet containing 1% TBHQ for 6 weeks. All rats were subjected to
    two-thirds partial hepatectomy at week 3. The number and area of
    preneoplastic GST placental form positive foci in the liver were
    significantly lower in rats treated with 1% TBHQ compared to rats that
    received diethylnitrosamine alone (Ito  et al., 1988).

        The combined effects of low doses of various carcinogens and
    carcinogenesis modifiers on tumour development were investigated in
    the F344 rat. Carcinogens included a group of known hepatocarcinogens
    and a group of nitroso compounds having various target organ
    specificities. A group of antioxidants, BHA, catechol, propyl gallate
    and TBHQ, known to have various target-organ-dependent inhibiting and
    enhancing activities, were used as carcinogenesis modifiers. These
    groups of compounds, alone or in combination, were administered with
    or without prior administration of N-diemethylnitrosamine,
    N-methylnitrosourea and dihydroxy-di-N-propylnitrosamine. The
    hepatocarcinogen group, in combination with various nitroso compounds,
    increased the incidences of liver hyperplastic nodules and
    hepatocellular carcinomas. When antioxidants were administered in
    combination with hepatocarcinogens and/or nitroso compounds, the
    incidence of both hepatic lesions was clearly reduced. The combination
    of nitroso compounds and antioxidants increased the incidence of
    papillomas and carcinomas in the urinary bladder compared to nitroso
    compounds alone. Rats receiving nitroso compounds in combination with
    hepatocarcinogens also showed a tendency for increased tumour
    development. Additive effects on the number of preneoplastic lesions
    in the glandular stomach were produced by combined treatment with
    antioxidants and nitroso compounds. No synergistic effects on tumour
    development were seen in other organs (Fukushima  et al., 1991).

        Two weeks after initiation with diethylnitrosamine, groups of male
    F344 rats received diets containing 1 or 0.25% BHA, 1 or 0.25% TBHQ,
    0.8 or 0.2% catechol or 0.5 or 0.125% sesamol for 6 weeks. Additional
    groups received combined low doses of BHA and TBHQ, catechol and
    sesamol, or BHA, TBHQ, catechol and sesamol. All rats were subjected
    to two-thirds partial hepatectomy at the end of week 3 and killed at
    the end of week 8. Body weights were reduced in the groups receiving
    1% TBHQ and all four antioxidants. All groups receiving TBHQ alone or
    in combination had significantly higher absolute and relative liver
    weights compared to controls. The number and areas of GST placental
    form positive foci in the liver were significantly reduced in groups
    receiving 1% TBHQ, TBHQ and BHA, and all four antioxidants (Hasegawa
     et al., 1992).

        A series of 31 structurally related chemicals, including TBHQ,
    were investigated as  in vitro inhibitors of BP metabolism in hepatic
    microsomes from 3-methyl-cholanthrene-induced mice and DNA-BP adduct
    formation. TBHQ was shown to be a more powerful inhibitor of BP
    metabolism and adduct formation than BHA or BHT. The authors suggested
    that the inhibition of DNA-BP adduct formation by TBHQ was secondary
    to the inhibition of BP metabolism. The quinones were shown to be more

    effective inhibitors of DNA-BP adduct formation than the corresponding
    hydroquinones. The authors postulated that the inhibition observed
    with hydroquinones may result in part from their oxidized forms. The
    study also demonstrated that substituted hydroquinones and quinones
    exhibited a larger inhibitory effect than unsubstituted hydroquinones
    and quinones (Colovai  et al., 1993).

    2.2.11.2  Urinary bladder

        The promoting activity of TBHQ in urinary bladder carcinogenesis
    initiated by N-butyl-N-(4-hydroxybutyl)nitrosamine (BBN) in male
    F344/Dulrj rats was examined and compared with the effect of
    alpha-tocopherol (TP) or propyl gallate (PG). Rats (6-week old) were
    treated with 0.05% BBN in drinking-water for 4 weeks. Groups of 20
    rats received thereafter control diet or diet containing 2.0% TBHQ,
    1.0% PG, or 0.4, 0.75, or 1.5% TP. After 36 weeks, the urinary
    bladders of all animals were examined histologically. The incidence of
    papillary or nodular hyperplasia was significantly higher in rats
    treated with BBN plus TBHQ as compared to rats initiated with BBN but
    receiving control diet. However, there were no significant differences
    for papillomas or cancer. This indicated a weak promoting activity of
    TBHQ in BBN-initiated urinary bladder carcinogenesis. TP and PG were
    inactive in this respect (Tamano  et al., 1987).

        The modifying activities of BHA, BHT and TBHQ and of paired
    combinations of these phenolic antioxidants on bladder carcinogenesis
    in male F344 rats pretreated with BBN were investigated. Groups of 20
    animals (6-weeks old) were given 0.05% BBN in their drinking-water for
    4 weeks, followed by BHA, BHT or TBHQ alone (0.8% each) or in pairs
    (0.4% each) in their diet for 32 weeks. Controls received no further
    treatment after BBN administration. A decrease in body weight gain was
    observed in all antioxidant-treated groups. The incidence of
    preneoplastic papillary or nodular hyperplasia (PN hyperplasia) was
    slightly but significantly higher in the group treated with BHA plus
    TBHQ after BBN than in controls receiving BBN only. The densities of
    PN hyperplasia were also significantly increased in all treated
    groups. However, no synergistic enhancing effects were observed. No
    differences were seen with respect to the incidence and densities of
    papillomas or carcinomas. Thus BHA, BHT and TBHQ all exerted enhancing
    effects in BBN-induced urinary bladder carcinogenesis in rats, but no
    synergism regarding this promotion occurred (Hagiwara  et al., 1989).

        In a review article, studies investigating the promoting activity
    of antioxidants (BHA, BHT, ethoxyquin and TBHQ) in a two-stage bladder
    carcinogenesis model were discussed. TBHQ and ethoxyquin exerted weak
    promoting activites in BBN-initiated male rats while those of BHA and
    BHT were strong. The promoting effects of the antioxidants could not
    be related to urinary components or the potency of their antioxidant
    action. These antioxidants induced scanning electron microscopic
    changes, characteristic of a hyperplastic response, on the surface of
    urinary bladder epithelial cells. A BrdU incorporation study revealed
    that TBHQ and the other antioxidants induced increased DNA synthesis
    in the urinary bladder epithelium in rats (Ito & Fukushima, 1989).

        The effects of 16 weeks of dietary exposure to carcinogens (three
    sodium salts) or promoters (BHA, BHT and TBHQ), alone or in
    combinations, on BBN-initiated rat bladder carcinogenesis were
    examined. Exposure to 0.7% of dietary TBHQ alone significantly
    (P<0.01) lowered final body weight but had no significant effect on
    relative bladder weight compared to BBN-initiated controls. The
    incidence of bladder papillomas and/or carcinomas in the TBHQ
    treatment group was comparable to controls. When rats received a diet
    containing 0.7% TBHQ, 0.7% BHA and 0.3% BHT combined, final body
    weights were significantly lower and relative bladder weight was
    significantly higher than for BBN-initiated controls. This combined
    treatment had an additive effect on the incidence of bladder tumours,
    significantly increasing the incidence of papillomas alone and
    papillomas and carcinomas combined. Treatment with the combination of
    antioxidants was also shown to increase DNA synthesis in the bladder
    although the increase was not statistically significant (Ono 
     et al., 1992).

    3.  COMMENTS

        In studies reviewed at earlier meetings of the Committee, TBHQ was
    shown to be extensively absorbed and rapidly excreted following
    ingestion by rats, dogs and humans. The major urinary metabolites in
    all three species were the O-sulfate and O-glucuronide conjugates,
    with the former predominating. In numerous  in vitro studies, TBHQ
    was shown to induce the activity of phase II enzymes, including
    UDP-glucuronosyl-S-transferase and glutathione-S-transferase, by a
    mechanism independent of the  Ah receptor. Induction of hepatic
    glutathione-S-transferase activity was also demonstrated following
    short-term dietary administration of TBHQ in female mice.

        TBHQ also undergoes redox cycling with the corresponding quinone,
    accompanied by the production of reactive oxygen species.  In a study
    reviewed at the present meeting, three glutathione conjugates of TBHQ
    were identified in the bile of male rats, and sulfur-containing
    metabolites of TBHQ were detected in the urine. In other studies, the
    glutathione conjugates of TBHQ demonstrated increased redox cycling
    activity compared with unconjugated TBHQ and were toxic to the kidney
    and bladder when administered intravenously to male rats.

        In a new thirteen-week feeding study conducted in mice,
    significant treatment-related effects noted in both sexes were
    decreased body weight gain and mucosal epithelial hyperplasia of the
    forestomach. The latter effect was noted only at very high doses of
    TBHQ, i.e. 2% of the diet (equal to 4000-4500 mg/kg bw per day) and
    above. The NOEL for the study was 870 mg/kg bw per day. In a
    thirteen-week feeding study conducted in rats continuously exposed to
    TBHQ, starting  in utero, treatment-related haemosiderin pigmentation
    of the spleen was noted in both sexes. In addition, there was a
    treatment-related increase in atrophy of the red pulp of the spleen in
    female rats receiving the mid and high doses of TBHQ. Bone marrow and
    haematological parameters were not altered at these doses. Forestomach
    hyperplasia was not observed in rats in this study, up to the highest

    dose tested, i.e. 1% of the diet (equal to 800 mg/kg bw per day),
    although it was noted in another study following short-term
    administration in the diet of adult rats at the 2% level. Treatment
    with TBHQ had no effect on the estrous cycle or the histological
    appearance of the reproductive organs. Because pigmentation of the
    spleen was noted in female rats at the lowest dose tested, a NOEL
    could not be established, leaving a lowest-observed-effect level
    (LOEL) of 190 mg/kg bw per day. Irritation and hyperplasia observed in
    the nasal epithelium of both species and the skin of mice were
    considered to be the consequence of direct contact with TBHQ from the
    diet.

        The results from the recently-conducted carcinogenicity studies
    were reviewed. In female mice, TBHQ induced an increase in the
    incidence of thyroid follicular cell hyperplasia, affecting all dosed
    groups. A non-significant increase in follicular cell adenomas was
    reported at the high dose but the incidence was within the range of
    historical control values. No follicular cell carcinomas were
    observed. Decreased body weight gains were also observed at the high
    dose in both sexes. Since the Committee was aware that hydroquinone
    (the unsubstituted parent compound) induced thyrotoxicity in mice, but
    not rats, it considered that the follicular cell hyperplasia observed
    with TBHQ in this study might be a toxicologically significant effect.
    Consequently, it concluded that a NOEL could not be established for
    this study and that the low dose of 150 mg/kg bw per day represented a
    LOEL. In the rat study, toxicologically significant effects were noted
    only at the highest dose tested, which consisted of an increase in the
    incidence of transitional cell hyperplasia and suppurative
    inflammation of the kidneys in male rats and of haemosiderin
    pigmentation of the spleen in females. The Committee considered that
    the mid dose of 2500 mg/kg feed, equal to 110 mg/kg bw per day,
    represented the NOEL and that TBHQ was not carcinogenic in mice or
    rats.

        The 117-week study in dogs on which the temporary ADI had been
    based was re-evaluated in the light of supplementary information
    obtained from the study authors. Based on actual intake data, nominal
    dietary levels of 500, 1580 and 5000 mg/kg feed were equal to doses of
    21, 72 and 260 mg/kg bw per day. Haemoglobin concentrations and
    haematocrits were statistically significantly reduced in high-dose
    dogs of both sexes at several sampling intervals throughout the study,
    although values were within historical control ranges. Red blood cell
    counts were also significantly decreased in male and female dogs in
    the high dose group at week 112, the only time at which this parameter
    was reported. Increases in the reticulocyte count (as % of red blood
    cells) and the presence of immature red blood cell forms in the
    peripheral blood of animals from the treated groups, reported to occur
    late in the study, were not dose-related nor accompanied by changes in
    red blood cell parameters at the low and mid doses. On the basis of
    this re-evaluation, the Committee confirmed that the NOEL for
    long-term toxicity in dogs was the mid dose, equal to 72 mg/kg bw per
    day.

        In view of the conflicting results among the genotoxicity assays
    reviewed at previous meetings of the Committee, many of the studies
    were re-evaluated at the present meeting with respect to the validity
    of the protocol and data interpretation. The conclusions of a number
    of the studies could no longer be supported. The results of the
    well-conducted studies indicated that TBHQ was clastogenic  in vitro 
    in the absence or presence of metabolic activation but did not induce
    the formation of micronuclei  in vivo. In sister chromatid exchange
    assays, TBHQ was positive in mice  in vivo and in an  in vitro 
    system. The results from several studies suggested that DNA damage
    resulting from TBHQ exposure, including chromosome loss and breakage,
    was secondary to the production of reactive oxygen species. In light
    of this information, and the fact that TBHQ was not carcinogenic in
    two rodent species, the Committee concluded that TBHQ was unlikely to
    be genotoxic  in vivo under conditions of use as an antioxidant, and
    that further genotoxicity studies were unnecessary.

        The results of four reproductive toxicity studies in rats were
    evaluated. Taken together, the results of these studies indicated an
    adverse effect of TBHQ on pup survival and/or pup body weight at
    dietary levels of 0.5% or higher. The effect on pup body weight was
    late-occurring in terms of the lactation period and the NOEL was 0.25%
    diet, equivalent to 125 mg/kg bw per day.

    4.  EVALUATION

        The Committee concluded on the basis of the data reviewed at the
    present meeting that TBHQ was not carcinogenic in rats or mice.  In
    reviewing the long-term toxicity studies in mice, rats and dogs and
    the reproductive toxicity studies in rats, the Committee concluded
    that the most sensitive species was the dog. The Committee allocated
    an ADI of 0-0.7 mg/kg bw based on the NOEL of 70 mg/kg bw per day to
    which a 100-fold safety factor was applied.

    5.  REFERENCES

    Altmann, H.-J., Grunow, W., Wester, P.W., & Mohr, U. (1985) Induction
    of forestomach lesions by butylhydroanisole and structurally related
    substances.  Arch. Toxicol., 8(Suppl.): 114-116.

    Altmann, H.-J., Grunow, W., Mohr, U., Richter-Reichhelm, H.B., &
    Wester, P.W. (1986) Effects of BHA and related phenols on the
    forestomach of rats.  Food Chem. Toxicol., 24: 1183-1188.

    Astill, B.D. & Jones, B.E. (1969) Long-term feedings of TBHQ to rats
    and dogs: levels of TBHQ in tissues and organs of autopsied animals.
    Unpublished report from the Biochemistry Laboratory, Eastman Kodak
    (Submitted to WHO by Eastman Chemical Products, Inc.).

    Astill, B.D. & Mulligan, L.T. (1977) Phenolic antioxidants and the
    inhibition of hepatotoxicity from N-dimethylnitrosamine formed 
     in situ in the rat stomach.  Food Cosmet. Toxicol., 15: 167-171.

    Astill, B.D. & Walton, D.A. (1968) Distribution of radioactivity in
    pregnant rats receiving 14C-labelled TBHQ. Unpublished report from
    the Biochemistry Laboratory, Eastman Kodak (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Astill, B.D., Blakeley, R.V., & Cantor, E.E. (1967a) The metabolic
    fate of TBHQ in rats and dogs and of TBHQ-14C in rats. Unpublished
    report of the Biochemical Laboratory, Eastman Kodak (Submitted to WHO
    by Eastman Chemical Products, Inc.).

    Astill, B.D., Cantor, E.E., & Mcewan, D.B. (1967b) Long-term feedings
    of TBHQ to rats and dogs: urinary conjugate excretions, serum TBHQ
    levels and autopsied fat analyses. Unpublished report from the
    Biochemistry Laboratory, Eastman Kodak (Submitted to WHO by Eastman
    Chemical Products, Inc.).

    Astill, B.D., Blakeley, R.V., Cantor, E.E., Tischer, K.S., Walton,
    D.A., McEwan, D.B., Jones, W.H., & Ely, T.S. (1968) Biochemical
    studies on tert-butylhydroquinone (TBHQ): A summary. Unpublished
    report from the Laboratory of Industrial Medicine, Eastman Kodak
    (Submitted to WHO by Eastman Chemical Products, Inc.).

    Astill, B.D., Cantor, E.E., Ely, T.S., Jones, W.H., & Uskavitch, R.J.
    (1967c) The oral ingestion of t-butylhydroquinone (TBHQ) by humans;
    clinical observations and metabolic fate. Unpublished report from the
    Laboratory of Industrial Medicine, Eastman Kodak (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Beilman, J.J. & Barber, E.D. (1985) Evaluation of
    mono-t-butylhydroquine in the CHO/HGPRT forward mutation assay.
    Unpublished report No. 85-0061 from Health and Environment
    Laboratories, Eastman Kodak Co., Rochester, NY, USA (Submitted to WHO
    by Eastman Kodak Co., Kingsport, TN, USA).

    Bergelson, S., Pinkus, R., & Daniel, V. (1994a) Intracellular
    glutathione levels regulate Fos/Jun induction and activation of
    glutathione S-transferase gene expression.  Cancer Res., 54: 36-40.

    Bergelson, S., Pinkus, R., & Daniel, V. (1994b) Induction of AP-1
    (Fos/Jun) by chemical agents mediates activation of glutathionine
    S-transferase and quinone reductase gene expression.  Oncogene, 9:
    565-571.

    Bergmann, B., Dohrmann, J.K., & Kahl, R. (1992) Formation of the
    semiquinone anion radical from  tert-butylquinone and from
     tert-butylhydroquinone in rat liver microsomes.  Toxicology, 74:
    127-133.

    Boehme, M.A. & Branen, A.L. (1977) Effects of food antioxidants on
    prostaglandin biosynthesis.  J. Food Sci., 42: 1243-1246.

    Ciaccio, P.J., Jaiswal, A.K., & Tew, K.D. (1994) Regulation of human
    dihydrodiol dehydrogenase by Michael acceptor xenobiotics.  J. Biol. 
     Chem., 169: 15558-15562.

    Colovai, A.I., Herdan, J., Popovici, C., Mihai, C., Grigoras, C.,
    Voiculetz, N., & Nicolescu-Duvaz, I. (1993)  In vitro inhibition of
    DNA adduct formation by hindered hydroquinols and quinones.
     Carcinogenesis, 14: 1137-1141.

    Dobo, K.L. & Eastmond, D.A. (1994) Role of oxygen radicals in the
    chromosomal loss and breakage induced by the quinone-forming
    compounds, hydroquinone and  tert-butylhydroquinone.  Environ. Mol.
     Mutagen., 24: 293-300.

    Eastman Chemical Products (1968a) Determination of the stability and
    tertiary butyl hydroquinone content of rat and dog fat extracts.
    Unpublished technical report No. 495-F Submitted to WHO by Eastman
    Chemical Products, Inc.).

    Eastman Chemical Products (1968b) Two-year chronic feeding studies
    with tertiary butyl hydroquinone (TBHQ) in dogs. Unpublished report
    from the Food and Drug Research Labs, Inc. (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Eisele, T.A., Sinnhuber, R.O., & Nixon, J.E. (1983) Dietary
    antioxidant effects on the hepatic mixed-function oxidase system of
    rainbow trout  (Salmo gairdneri). Food Chem. Toxicol., 21: 273-277.

    Fassett, D.W., Roudabush, R.L., & Terhaar, C.J. (1965) Reproduction
    study on rats fed monotertiary butyl hydroquinone. Unpublished report
    from the Laboratory of Industrial Medicine, Eastman Kodak (Submitted
    to WHO by Eastman Chemical Products, Inc.).

    Fassett, D.W., Terhaar, C.J., & Astill, B.D. (1968) Summary of the
    safety evaluation of monotertiary butyl hydroquinone. Unpublished
    report from the Laboratory of Industrial Medicine, Eastman Kodak
    (Submitted to WHO by Eastman Chemical Products, Inc.).

    Fukushima, S., Shibata, M.-A., Hirose, M., Kato, T., Tatematsu, M., &
    Ito, N. (1991) Organ-specific modification of tumor development by
    low-dose combinations of agents in a rat wide-spectrum carcinogenesis
    model.  Jpn. J. Cancer Res., 82: 784-792.

    Giri, A.K., Sen, S., Talukder, G., & Sharma, A. (1984) Mutachromosomal
    effects of tert-butylhydroquinone in bone marrow cells of mice. 
     Food Chem. Toxicol., 22: 459-460.

    Hageman, F.J., Verhagen, H., & Kleinjans, J.C.S. (1988) Butylated
    hydroxyanisole, butylated hydroxytoluene and tert-butylhydroquinone
    are not mutagenic in the  Salmonella/microsome assay using new tester
    strains.  Mutat. Res., 208: 207-211.

    Hagiwara, A., Hirose, M., Miyata, Y., Fukushima, S., & Ito, N. (1989)
    Modulation of  N-butyl-(4-hydroxybutyl) nitrosamine-induced rat
    urinary bladder carcinogenesis by post-treatment with combinations of
    three phenolic antioxidants.  J. Toxicol. Pathol., 2: 33-39.

    Hasegawa, R., Tiwawech, D., Hirose, M., Takaba, K., Hoshiya, T.,
    Shirai, T., & Ito, N. (1992) Suppression of
    diethylnitrosamine-initiated preneoplastic foci development in the rat
    liver by combined administration of four antioxidants at low doses.
     Jpn. J. Cancer Res., 83: 431-437.

    Hirose, M., Inoue, T., Asamoto, M., Tagawa, Y., & Ito, N. (1986)
    Comparison of the effects of 13 phenolic compounds in induction of
    proliferative lesions of the forestomach and increase in the labelling
    indices of the glandular stomach and urinary bladder epithelium of
    Syrian golden hamsters.  Carcinogenesis, 7: 1285-1289.

    Hirose, M., Masuda, A., Fukushima, S., & Ito, N. (1988) Effects of
    subsequent antioxidant treatment on 7,12-dimethylbenz[a]anthracene-
    initiated carcinogenesis of the mammary gland, ear duct and
    forestomach in Sprague-Dawley rats.  Carcinogenesis, 9: 101-104.

    Hirose, M., Yada, H., Hakoi, K., Takahashi, S., & Ito, N. (1993)
    Modification of carcinogenesis by alpha-tocopherol,
    t-butylhydroquinone, propyl gallate and butylated hydroxytoluene in a
    rat multi-organ carcinogenesis model.  Carcinogenesis, 14(11):
    2359-2364.

    Ito, N. & Fukushima, S. (1989) Promotion of urinary bladder
    carcinogenesis in experimental animals.  Exp. Pathol., 36: 1-15.

    Ito, N., Tsuda, H., Tatematsu, M., Inoue, T., Tagawa, Y., Aoki, T.,
    Uwagawa, S., Kagawa, M., Ogiso, T., Masui, T., Imaida, K., Fukushima,
    S., & Asamoto, M. (1988) Enhancing effect of various hepatocarcinogens
    on induction of preneoplastic glutathione S-transferase placental form
    positive foci in rats - an approach for a new medium-term bioassay
    system.  Carcinogenesis, 9: 387-394.

    Kahl, R., Weinke, S., & Kappus, H. (1989) Production of reactive
    oxygen species due to metabolic activation of butylated
    hydroxyanisole.  Toxicology, 59: 179-194.

    Kawabe, M., Takaba, K., Yoshida, Y., & Hirose, M. (1994) Effects of
    combined treatment with phenolic compounds and sodium nitrite on
    two-stage carcinogenesis and cell proliferation in the rat stomach.
     Jpn. J. Cancer Res., 85: 17-25.

    Krasavage, W.J. (1973) A study of the teratogenic potential of
    monotertiary butyl hydroquinone in rats. Unpublished report from the
    Health and Safety Laboratory, Eastman Kodak (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Krasavage, W.J. (1977) Evaluation of the teratogenic potential of
    tertiary butylhydroquinone (TBHQ) in the rat.  Teratology, 16: 31-33.

    Krasavage, W.J. (1984) The lack of effect of tertiary
    butylhydroquinone on prothrombin time in male rats.  Drug Chem. 
     Toxicol., 7: 329-334.

    Krasavage, W.J. & Faber, W.D. (1983) Tertiary butylhydroquinone
    (TBHQ): dominant lethal assay in rats. Unpublished report from Health
    and Environment Laboratories, Eastman Kodak Co., Rochester, NY, USA
    (Submitted to WHO by Eastman Kodak Co., Kingsport, TN, USA).

    Krasavage, W.J. & O'Donoghue, J.L. (1984) Lack of lung damage in mice
    following administration of tertiary butylhydroquinone.  Drug Chem. 
     Toxicol., 7: 335-343.

    Krasavage, W.J. & Terhaar, C.J. (1970) The reproductive performance of
    rats fed monotertiary butyl hydroquinone: a single generation study.
    Unpublished report from the Toxicology Laboratory, Eastman Kodak
    (Submitted to WHO by Eastman Chemical Products, Inc.).

    Li, Y. & Jaiswal, A.K. (1994) Human antioxidant-response-element-
    mediated regulation of type 1 NAD(P)H:quinone oxidoreductase gene
    expression: Effect of sulfhydryl modifying agents.  Eur. J. 
     Biochem., 226: 31-39.

    Li, Y. & Trush, M.A. (1994) Reactive oxygen-dependent DNA damage
    resulting from the oxidation of phenolic compounds by a copper-redox
    cycle mechanism.  Cancer Res., 54(Suppl.): 1895s-1898s.

    Litton Bionetics (1982a) Mutagenicity evaluation of EK 81-0318 (TBHQ)
    in the mouse lymphoma forward mutation assay. Unpublished report No.
    20989 from Litton Bionetics Inc. (Submitted to WHO by Eastman Kodak
    Co., Kingsport, TN, USA).

    Litton Bionetics (1982b) Mutagenicity evaluation of EK 81-0318 (TBHQ)
    in the mouse micronucleus test. Unpublished report No. 20996 from
    Litton Bionetics Inc. (Submitted to WHO by Eastman Kodak Co.,
    Kingsport, TN, USA).

    Litton Bionetics (1985) Mutagenicity evaluation of EK 81-0318 (TBHQ)
    in the mouse bone marrow cytogenetic assay. Unpublished report No.
    22202 from Litton Bionetics Inc. (Submitted to WHO by Eastman Kodak
    Co., Kingsport, TN, USA).

    Liu, R.M., Hu, H., Robison, T.W., & Forman, H.J. (1996a) Differential
    enhancement of gamma-glutamyl transpeptidase and gamma
    glutamylcysteine synthetase by  tert-butylhydroquinone in rat lung
    epithelial L2 cells.  Am. J. Respir. Cell Mol. Biol., 14(2): 186-191.

    Liu, R.M., Hu, H., Robison, T.W., & Forman, H.J. (1996b) Increased
    gamma-glutamylcysteine synthetase and gamma glutamyl transpeptidase 
    activities enhance resistance of rat lung epithelial L2 Cells to
    quinone toxicity.  Am. J. Respir. Cell Mol. Biol., 14(2): 192-197.

    Liu, R.M., Vasiliou, V., Zhu, H., Duh, J.-L., Tabor, M., Puga, A.,
    Nebert, D.W., Sainsbury, M., & Shertzer, H.G. (1994) Regulation of
    [Ah] gene battery enzymes and glutathione levels by
    5,10-dihydroindeno[1,2-b]indole in mouse hepatoma cell lines.
     Carcinogenesis, 15: 2347-2352.

    Matsuoka, A., Matsui, M., Miyata, N., Sofuni, T., & Ishidate, M.
    (1990) Mutagenicity of 3-tert-butyl-4-hydroxyanisole (BHA) and its
    metabolites in short-term tests  in vitro.  Mutat Res., 241:
    125-132.

    Morimoto, K., Tsuji, J, Iio, T., Miyata, N., Uchida, A., Osawa, R.,
    Kitsutaka, H., & Takahashi, A. (1991) DNA damage in forestomach
    epithelium from male F344 rats following oral administration of
     tert-butylquinone, one of the forestomach metabolites of 3-BHA.
     Carcinogenesis, 12(4): 703-708.

    Morimoto, K., Takahashi, T., Okudaira, K., Iio, T., Saito, Y., &
    Takahashi, A. (1992) Dose-response study on covalent binding to
    forestomach protein from male F344 rats following oral administration
    of [14C]3-BHA.  Carcinogenesis, 13(9): 1663-1666.

    Mueller, K.R. & Lockhart, H.B. (1983)  In vitro genetic activity
    report: evaluation of mono-tertairy butylhydroquinone in the Ames
     Salmonella/microsome bacterial mutagenesis test. Unpublished report
    from Health, Safety and Human Factors Laboratory, Eastman Kodak Co.,
    Rochester, NY, USA (Submitted to WHO by Eastman Kodak Co., Kingsport,
    TN, USA).

    Mukherjee, A., Talukder, G., & Sharma, A. (1989) Sister chromatid
    exchanges induced by tertiary butyl hydroquinone in bone marrow cells
    of mice.  Environ. Mol. Mutagen., 13: 234-237.

    Nagai, F., Okubo, T., Ushiyama, K., Satoh, K., & Kano, I. (1996)
    Formation of 8-hydroxydeoxyguanosine in calf thymus DNA treated with
     tert-butylhydroquinone, a major metabolite of butylated
    hydroxyanisole.  Toxicol. Lett., 89: 163-167.

    Nakagawa, Y. (1996) Effects of dicoumarol on cytotoxicity caused by
     tert-butylhydroquinone in isolated rat hepatocytes.  Toxicol. 
     Lett., 84: 63-68.

    Nakagawa, Y. & Moldéus, P. (1992) Cytotoxic effects of
    phenyl-hydroquinone and some hydroquinones on isolated rat
    hepatocytes.  Biochem. Pharmacol., 44(6): 1059-1065.

    Nakagawa, Y., Nakajima, K., Moore, G., & Moldeus, P. (1994) On the
    metabolism of 3- tert-butyl-4-hydroxyanisole- and its
    metabolites-induced cytotoxicities in isolated rat hepatocytes. 
     Eur. J. Pharmacol. Environ. Toxicol. Pharmacol. Sect., 270: 341-348.

    Nera, E.A., Lik, E., Iverson, F., Ormsky, E., Karpinsky, K.F, &
    Clayson, D.B. (1984) Short-term proliferative effects of BHA and other
    phenolic antioxidants in the forestomach of Fischer 344 rats.
     Toxicology, 32: 197-213.

    NTP (National Toxicology Program) (1995) Toxicology and carcinogenesis
    studies of t-butylhydroquinone (CAS No. 1948-33-0) in F344/N and
    B6C3F1 mice (feeding studies). NIH, Research Triangle Park, N.C.
    (Draft Technical Report Series No. 459; NIH Publication No. 95-3375).

    Ono, S., Kurata, Y., Shichino, Y., Sano, M., & Fukushima, S. (1992)
    Synergism of environmental carcinogens and promoters on bladder cancer
    development inititated by  N-butyl-N-(4-hydroxybutyl) nitrosamine in
    F344 rats.  Jpn. J. Cancer Res., 83: 955-963.

    Orser, C.S., Foong, F.C.F., Capaldi, S.R., Nalezny, J., Mackay, W.,
    Benjamin, M., & Farr, S.B. (1995) Use of prokaryotic stress promoters
    as indicators of the mechanisms of chemical toxicity.  In vitro 
     Toxicol., 8: 71-85.

    Peters, A.A., Lau, S.S., Dulik, D., Murphy, D., Van Ommen, B., Van
    Bladeren, P.J., & Monks, T.J. (1996a) Metabolism of  tert-
    butylhydroxyquinone to S-substituted conjugates in the male Fischer
    344 rat.  Chem. Res. Toxicol., 9: 133-139.

    Peters, M.M., Rivera, M.I., Jones, T.W., Monks, T.J., & Lau, S.S.
    (1996b) Glutathione conjugates of  tert-butyl-hydroquinone, a
    metabolite of the urinary tract tumor promoter 
    3- tert-butyl-hydroxyanisole, are toxic to kidney and bladder.
     Cancer Res., 56(5): 1006-1011.

    Phillips, B.G., Carroll, P.A., Tee, A.C., & Anderson, D. (1989)
    Microsome-mediated clastogenicity of butylated hydroxyanisole (BHA) in
    cultured Chinese hamster ovary cells: The possible role of reactive
    oxygen species.  Mutat. Res., 214: 105-114.

    Pinkus, R., Weiner, L.M., & Daniel, W. (1996) Role of oxidants and
    antioxidants in the induction of AP-1, Nf-kappaB, and glutathione
    S-transferase gene expression.  J. Biol. Chem., 271(23): 13422-13429.

    Prochaska, H.J. (1987) Mechanism of modulation of carcinogenesis by
    chemoprotective enzyme inducers. Importance of redox liability. 
     Proc. Annu. Meet. Am Assoc. Cancer Res., 28: 127.

    Prochaska, H.J. (1994) Screening strategies for the detection of
    anticarcinogenic enzyme inducers.  J. Nutr. Biochem., 5: 360-368.

    Rahimtula, A.D., Jernstrom, B., Dock, L., & Moldeus, P. (1982) Effects
    of dietary and  in vitro 2(3)-t-butyl-4-hydroxy-anisole and other
    phenols on hepatic enzyme activities in mice.  Br. J. Cancer, 45:
    935-944.

    Rogers, C.G., Boyes, BG., Matula, T.I., & Stapley, R. (1992)
    Evaluation of genotoxicity of  tert-butylhydroquinone in a
    hepatocyte-mediated assay with V79 Chinese hamster lung cells and in
    strain D7 of  Saccharomyces cerevisiae. Mutat. Res., 280: 17-27.

    Rogers, C.G., Boyes, B.G., Matula, T.I., Neville, B., & Stapley, R.
    (1993) Cytotoxic and genotoxic properties of  tert-butyl- p-quinone
    (TBQ) in an  in vitro assay system with Chinese hamster V79 cells and
    in strain D7 of  Saccharomyces cerevisiae. Mutat. Res., 299: 9-18.

    Sakai, T., Klopman, G., & Rosenkranz, H.S. (1994) Structural basis for
    the induction of preneoplastic glutathione S-transferase positive foci
    by hepatocarcinogens.  Teratog. Carcinog. Mutagen., 14: 219-237.

    Schilderman, P.A.E.L., Van Maanen, J.M.S., Ten Vaarwerk, F.J.,
    Lafleur, M.V.M., Westmijze, E.J., Ten Hoor, F., & Kleinjans, J.C.S.
    (1993a) The role of prostaglandin H synthase-mediated metabolism in
    the induction of oxidative DNA damage by BHA metabolites.
     Carcinogenesis, 14(7): 1297-1302.

    Schilderman, P.A.E.L., Van Maanen, J.M.S., Smeets, E.J., Ten Hoor, F.,
    & Kleinjans, J.C.S. (1993b) Oxygen radical formation during
    prostaglandin H synthase-mediated biotransformation of butylated
    hydroxyanisole.  Carcinogenesis, 14(3): 347-353.

    Schilderman, P.A.E.L., Rhijnsburger, E., Zwingmann, I., & Kleinjans,
    J.C.S. (1995) Induction of oxidative DNA damage and enhancement of
    cell proliferation in human lymphocytes  in vitro by butylated
    hydroxyanisole.  Carcinogenesis, 6: 507-512.

    Shertzer, H.G., Vasiliou, V., Liu, R.M., Tabor, M.W., & Nebert, D.W.
    (1995) Enzyme induction by L-buthionine (S,R)-sulfoximine in cultured
    mouse hepatoma cells.  Chem. Res. Toxicol., 8: 431-436.

    Shibata, M.-A., Asakawa, E., Hagiwara, A., Kurata, Y., & Fukushima, S.
    (1991) DNA synthesis and scanning electron microscopic lesions in
    renal pelvic epithelium of rats treated with bladder cancer promoters.
     Toxicol. Lett., 55: 263-272.

    Shibata, M.-A., Yamada, M., Tanaka, H., Kagawa, M., & Fukushima, S.
    (1989) Changes in urine composition, bladder epithelial morphology,
    and DNA synthesis in male F344 rats in response to ingestion of
    bladder tumour promoters.  Toxicol. Appl. Pharmacol., 99: 37-49.

    Société Kemin Europa (1982a) Recherche de mutagénicité sur 
     Salmonella typhimurium HIS selon la technique de B.N. Ames sur le
    produit TBHQ. Unpublished report No. 1PL-R-82044 prepared for Société
    Kemin Europa, S.A.

    Société Kemin Europa (1982b) Recherche de l'éventuelle potentialité
    mutagène de la teriao-butylhydroquinone par la technique du
    micronucleus chez la souris, C.E.R.T.I. Unpublished report No. 678
    prepared for Société Kemin Europa, S.A.

    Société Kemin Europa (1982) Recherche d'aberrations chromosomiques par
    analyse de métaphases sur cellules V79, produit TBHQ, Institut Pasteur
    de Lille. Unpublished report No. 1PL-R-82050 prepared for Société
    Kemin Europa, S.A.

    Sreerame, L., Rekha, G.K., & Sladek, N.E. (1995) Phenolic
    antioxidant-induced overexpression of class-3 aldehyde dehydrogenase
    and oxazaphosphorine-specific resistance.  Biochem. Pharmacol., 49:
    669-675.

    Suzuki, T., Matsuoka, T., Sawada, M., Hayashi, M., Miyata, N., &
    Sofuni, T. (1991) Cytotoxicity and micronucleus induction by 
    3- tert-butyl-4-hydroxyanisole and its metabolies in menadione- and
    H2O2-resistant cells.  Mutat. Res., 253: 278-279.

    Tajima, K., Hashizaki, M., Yamamoto, K., & Mizutani, T. (1991)
    Identification and structure characterization of S-containing
    metabolites of 3- tert-butyl-4-hydroxyanisole in rat urine and liver
    microsomes.  Drug. Metab. Dispos., 19(6): 1028-1033.

    Tajima, K., Hashizaki, M., Yamamoto, K., & Mizutani, T. (1992)
    Metabolism of 3- tert-butyl-4-hydroxyanisole by horseradish
    peroxidase and hydrogen peroxide.   Drug. Metab. Dispos., 20(6):
    816-820.

    Tamano, S., Fukushima, S., Shirai, T., Hirose, M., & Ito, N. (1987)
    Modification by alpha-tocopherol, propyl gallate and tertiary butyl
    hydroquinone of urinary bladder carcinogenesis in Fischer 344 rats
    pretreated with  N-butyl- N-(4-hydroxybutyl)nitrosamine.  Cancer 
     Lett., 35: 39-46.

    Terhaar, C.J. & Krasavage, W.J. (1968a) The reproductive performance
    of rats fed monotertiary butyl hydroquinone. Unpublished report from
    the Toxicology Laboratory, Eastman Kodak (Submitted to WHO by Eastman
    Chemical Products, Inc.).

    Terhaar, C.J. & Krasavage, W.J. (1968b) Monotertiary butyl
    hydroquinone (TBHQ) in heated and unheated cottonseed oil. A six month
    dietary feeding study in rats. Unpublished report from the Toxicology
    Laboratory, Eastman Kodak (Submitted to WHO by Eastman Chemical
    Products, Inc.).

    Terhaar, C.J., Vis, E.A., & Kesel, H.J. (1968a) Acute oral toxicity of
    monotertiary butyl hydroquinone (TBHQ). Unpublished report from the
    Laboratory of Industrial Medicine, Eastman Kodak (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Terhaar, C.J., Krasavage, W.J., Wolf, G.L., & Leonard, W.J. (1968b)
    Study of the tolerance of rats to monotertiary butyl hydroquinone in
    the diet: Final report at 20 months. Unpublished report from the
    Toxicology Laboratory, Eastman Kodak (Submitted to WHO by Eastman
    Chemical Products, Inc.).

    Tischer, K.S. & Walton, D.A. (1968) Dietary feeding of TBHQ and
    related compounds to rats and dogs: the response of liver processing
    enzymes and liver glucose-6-phosphatase activity. Unpublished report
    from the Biochemistry Laboratory, Eastman Kodak (Submitted to WHO by
    Eastman Chemical Products, Inc.).

    Tsuchiya, T., Ishida, A., Miyata, N., Takahashi, A., & Kamiya, S.
    (1988) Effects of 3- tert-butyl-4-hydroxyanisole and its hydroquinone
    and quinone metabolites on rat and human embryonic cells in culture.
     Toxicol. in vitro, 2: 291-296.

    Twerdok, L.E. & Trush, M.A. (1990). Differences in quinone reductase
    activity in primary bone marrow stromal cells derived from C57BL/6 and
    DBA/2 mice.  Res. Commun. Chem. Pathol. Pharmacol., 67(3): 375-381.

    Van Ommen, B., Koster, A., Verhagen, H., & Van Bladeren, P.J. (1992)
    The glutathione conjugates of  tert-butyl hydroquinone as potent
    redox cycling agents and possible reactive agents underlying the
    toxicity of butylated hydroxyanisole.  Biochem. Biophys. Res. 
     Commun., 189(1): 309-314.

    Vasiliou, V., Shertzer, H.G., Liu, R.M., Sainsbury, M., & Nebert, D.W.
    (1995)  Response of [Ah] battery genes to compounds that protect
    against menadione toxicity.  Biochem. Pharmacol., 50(11): 1885-1891.

    Wolf, C.L. & Fassett, D.W. (1968a) Electron microscope study of livers
    of rats fed monotertiary butyl hydroquinone. Unpublished report from
    the Laboratory of Industrial Medicine, Eastman Kodak (Submitted to WHO
    by Eastman Chemical Products, Inc.).

    Wolf, G.L. & Fassett, D.W. (1968b) Electron microscope study of livers
    of dogs fed monotertiary butyl hydroquinone for two years. Unpublished
    report from the Laboratory of Industrial Medicine, Eastman Kodak
    (Submitted to WHO by Eastman Chemical Products, Inc.).

    Wolf, G.L. & Fassett, D.W. (1968c) Electron microscope study of
    kidneys of dogs fed monotertiary butyl hydroquinone for two years.
    Unpublished report from the Laboratory of Industrial Medicine, Eastman
    Kodak (Submitted to WHO by Eastman Chemical Products, Inc.).

    Xiao, G.-H., Pinaire, J.A., Rodrigues, A.D., & Prough, R.A. (1995)
    Regulation of the Ah gene battery via Ah receptor-dependent and
    independent processes in cultured adult rat hepatocytes.  Drug 
     Metab. Dispos., 23: 642-650.

    Yoshida, Y., Hirose, M., Takaba, K., Kimura, J., & Ito, N. (1994)
    Induction and promotion of forestomach tumors by sodium nitrite in
    combination with ascorbic acid or sodium ascorbate in rats with or
    without  N-methyl- N'-nitro-N-nitrosoguanidine pre-treatment. 
     Int. J. Cancer, 56: 124-128.

    Yoshioka, K., Deng, T., Cavigelli, M., & Karin, M. (1995) Antitumor
    promotion by phenolic antioxidants: Inhibition of AP-1 activity
    through induction of Fra expression.  Proc. Natl Acad. Sci. (USA), 
    92: 4972-4976.

    Zeiger, E., Anderson, B., Haworth, S., Lawler, T. & Mortelmans, K.
    (1992).  Salmonella mutagenicity tests: V. Results from the testing
    of 311 chemicals.  Environ. Mol. Mutagen., 19(Suppl. 21): 2-141.

    

    See Also:
       Toxicological Abbreviations
       Butylhydroquinone, tert- (TBHQ) (WHO Food Additives Series 42)