IPCS INCHEM Home

    CHYMOSINS A AND B FROM GENETICALLY
    MODIFIED MICROORGANISMS

    First draft prepared by Dr F.S.D. Lin,
    Division of Toxicological Review and Evaluation
    Center for Food Safety and Applied Nutrition,
    US Food and Drug Administration

    1.  EXPLANATION

         Chymosins A and B derived from genetically modified
    microorganisms have not been previously evaluated by the Joint
    FAO/WHO Expert Committee on Food Additives.

         Chymosin, commonly known as rennin, is the principal milk-
    coagulating enzyme present in rennet.  Rennet, which has a long and
    extensive history of safe use in making cheese and other dairy
    products, is commercially prepared by aqueous extraction of dried
    fourth stomach of unweaned calves.  The aqueous extract contains a
    chymosin precursor, prochymosin, which is subsequently converted to
    enzymatically active chymosin.  Commercial preparations of calf
    rennet contain two forms of chymosin, A and B, usually in the
    proportion of about 40% of A and 60% of B.  Health aspects of rennet
    as a food ingredient were reviewed and evaluated at the fifteenth
    meeting of the Joint FAO/WHO Expert Committee on Food Additives in
    1972 (Annex 1, reference 26).

         Biochemically, chymosin (IUB No. 3.4.4.3) is a protein
    consisting of a single polypeptide chain of 323 amino acids with
    intramolecular disulfide linkages.  Chymosins A and B have been
    shown to differ only by one amino acid in the polypeptide chain; the
    former has an aspartic acid residue at position 286, whereas the
    latter has a glycine residue at the same position.

         Chymosin is produced intracellularly as preprochymosin. 
    Preprochymosin is shortened by 16 amino acids during secretion and
    appears in the stomach as prochymosin, which, in turn, is activated
    to chymosin by cleavage of an additional 42 amino acids.

         As a proteolytic enzyme, chymosin hydrolyses a specific bond in
    kappa-casein of milk, cleaving it into two peptides, para-kappa-
    casein and a macropeptide.  In milk, kappa-casein acts as a micelle
    stabilizer.  After this activity is destroyed by chymosin, milk
    coagulation occurs.  Chymosin A slightly exceeds chymosin B in
    proteolytic activity, whereas chymosin B is more stable at low pH
    (< 3.5) than chymosin A.

         In recent years recombinant DNA technology has made it possible
    to obtain calf chymosin as a fermentation product from
    nontoxicogenic and nonpathogenic strains of bacterium, yeast or

    filamentous fungus, which have been transformed with a plasmid
    vector containing a DNA sequence coding for the chymosin precursor. 
    Available biochemical evidence has established that the transferred
    prochymosin sequence can be expressed correctly in the new host
    organisms.  The prochymosin product has the same molecular weight as
    prochymosin found in calf rennet and it can be cleaved into chymosin
    that has the same chemical, physical and functional (enzymatic)
    properties as its mammalian counterpart.

         The three recombinant chymosins that were reviewed in this
    monograph, as well as their respective production organisms are
    identified below:

         (1)  chymosin A from  Escherichia coli K-12

         (2)  chymosin B from  Kluyveromyces lactis, and

         (3)  chymosin B from  Aspergillus niger var.  awamori.

    1.1  Chymosin A produced from  Escherichia coli K-12 containing
         calf prochymosin A gene

    1.1.2  Construction of production strain

          E. coli K12 JA198 strain was subjected to several genetic
    manipulations to construct the recipient strain for the expression
    plasmid carrying the prochymosin A gene.  The expression plasmid was
    derived from the widely used cloning vector pBR322.  cDNA coding for
    bovine chymosin A was previously cloned and characterized.  The
    prochymosin gene was divided into three sections, each terminated by
    a unique restriction endonuclease recognition site.  Each section
    was assembled from several synthetic oligonucleotides synthesized in
    an automated DNA synthesizer.  Each assembled section was subcloned
    into a pBR322 vector, transformed into  E. coli, and amplified. 
    The correctness of each construct was verified by restriction
    analysis and by sequencing.  The errors found in synthetic
    oligonucleotides were corrected by cassette mutagenesis.  All three
    subcloned sections were assembled together in the correct order to
    reconstruct the prochymosin gene, which was inserted into pBR322
    vector.  The gene was attached to the vector DNA through the
    ribosomal binding site and the  E. coli tryptophan (trp) promoter. 
    The created expression vector was transformed into the recipient
    strain GE81.  The plasmid carries the ampicillin resistance gene as
    a selective marker for bacterial transformants carrying the
    prochymosin gene (Pfizer Central Research, 1988).

    1.1.3  Fermentation

         The production strain is grown in an aqueous solution
    containing carbohydrates, nitrogen, mineral salts and miscellaneous
    inorganic and organic compounds (Pfizer Central Research, 1988).

    1.1.4  Recovery

         The solid prochymosin is liberated from the producing organism
    by cell disruption and harvesting of "inclusion bodies" by
    centrifugation or membrane concentration.  The harvested inclusion
    bodies are washed with phosphate buffer solution containing 1-4 M
    urea.  The residual  E. coli are inactivated by holding at pH less
    than 2.0 for at least one hour.  The inclusion bodies are then
    dissolved by addition of urea to a concentration of 7-9 M and pH
    adjustment to 10.0 - 11.0.  Subsequently, the solution containing
    prochymosin is diluted with a buffer, and the pH is reduced to 8.5-
    9.5, followed by a 2-hour period to allow renaturing of the
    prochymosin, which is subsequently activated to chymosin by
    adjusting the pH to 1.8 - 2.2 and holding for one hour.  Following
    readjustment of the pH to 5.5 - 6.0, the chymosin is purified via
    absorption on a suitable anion-exchange resin followed by elution
    with a buffer containing 1 M sodium chloride (Pfizer Central
    Research, 1988).

    1.1.5  Biochemical characterizations

         A number of side-by-side experiments were performed to
    demonstrate that the recombinant chymosin is identical to natural
    chymosin A, including the following (Pfizer Central Research, 1988):

         1.   Relative mass by SDS-PAGE (Sodium Dodecyl Sulfate-
              Polyacrylamide Gel Electrophoresis)

         2.   Isoelectric Point

         3.   Absorption Spectra

         4.   Anion-Exchange Chromatography

         5.   Amino Acid Analysis

         6.   Amino Terminal Sequencing

         7.   Peptide Mapping

         8.   Composition and Sequence of Selected Internal Peptides

         9.   Kinetic Specificity

         10.  Heat Stability

    2.  BIOLOGICAL DATA

    2.1  Biochemical Aspects

         No available data.

    2.1.1  Absorption, distribution, and excretion

         No available data.

    2.2  Toxicological data

    2.2.1  Acute studies

         No available data.

    2.2.2  Short-term studies

    2.2.2.1  Rat

         Four groups of 15 male and 15 female Long-Evans rats,
    approximately 90 g b.w., were held three weeks, then were given
    chymosin in their diet for 36-39 days at daily dose levels of 0,
    0.5, 2.0, or 5.0  mg/kg b.w.  Treated animals grew normally and
    exhibited normal clinical chemistry, haematology and urinalysis
    findings.  A transient decrease in food consumption and growth was
    observed at week 4 of the study in all treatment groups, male and
    female.  However, normal growth and food consumption continued for
    the remainder of the study.  Organ weights and the incidence of
    gross and microscopic lesions were comparable in control and
    treatment groups (Fisher, 1987b).

    2.2.2.2  Other relevant safety data

         Samples of the chymosin product were tested and found free from
    the production organism and expression plasmid (Pfizer Central
    Research, 1988).

         Chymosin product was tested by Limulus Amoebocycte Lysate
    bacterial endotoxin assay, and the levels of endotoxin found were
    low and comparable to those reported in US drinking water (Pfizer
    Central Research, 1988).

         Using  in vitro cytotoxicity test the production strain was
    shown to produce no detectable levels of "Shiga-like" toxin (SLT). 
    Although the chymosin product was not tested directly for the
    presence of SLT, the absence of SLT was concluded from a feeding
    study in Beagle dogs, in which the animals (6/sex/group) were given
    15 ml/dog of a formulation containing 0.1 (1.5 mg chymosin/kg
    b.w./day) or 0% of the enzyme preparation once daily for 5 days. 

    Treatment was well tolerated and produced no significant effects on
    food consumption or body weight gain.  Results from faecal occult
    blood determinations were negative, and no unusual emesis, diarrhoea
    or other gastrointestinal effects were reported in the treated
    animals (Fisher, 1987a).

    3.  COMMENTS

         The Committee reviewed the available data pertaining to the
    molecular construction of the chymosin expression plasmid;
    characterization of the producing organism; the fermentation
    process; biochemical and enzymatic characterization of the
    recombinant enzyme; and short-term feeding studies on the finished
    enzyme product.

         The Committee noted that a well-documented non-pathogenic
    strain of  E. coli had been used as the host organism for the
    expression plasmid (pPFZ87A) which was derived from a widely used
    cloning vector (pBR322).  The prochymosin A coding sequence, which
    was inserted into pBR322 plasmid, was initially synthesized
    chemically and shown to be identical to that of the natural calf
    gene.  The plasmid construct was introduced into the host organism
    by standard transformation procedures, and the ampicillin-resistance
    gene carried by the plasmid was used as a marker for the selection
    of the transformed cells.  Genetic stability of the transformed
    cells was demonstrated after repeated subculturing.

         The transformed cells were grown under properly controlled
    conditions in media containing ingredients commonly used in the
    production of food-grade substances by fermentation.  Prochymosin A
    was recovered from the producing organisms after disruption and
    separation of the cells.  Residual cells were inactivated by
    acidification, which also served to activate prochymosin A to
    chymosin A.

         After chromatographic purification, the active enzyme was
    formulated with stabilizer and preservatives typically used in
    commercial enzyme preparations.

         The recombinant enzyme was extensively characterized and shown
    to be chemically and functionally identical to the Chymosin A. 

         The results of studies on the microbiological purity of the
    enzyme product indicated the absence of producing organisms, minimal
    transfer of endotoxin from the bacterial cell walls, and the
    presence of insignificant levels of "shiga-like" toxins.  The low-
    levels of residual DNA detected in the product consisted of short
    fragments without any demonstrable genetic activity.

         The Committee noted that in a short-term (one month) study in
    rat to which the enzyme product was administered no adverse effects
    were observed at the highest dose level of 5 mg of chymosin per kg
    of body weight per day.

    4.  EVALUATION

         Taking into account the available safety information and the
    extremely low exposure resulting from its use in food production,
    the Committee established an ADI "not specified" for the recombinant
    chymosin A preparation.

    1.  EXPLANATION

    1.1  Chymosin B produced from  kluyvermyces lactis 
         containing calf prochymosin B Gene

    1.1.2  Construction of production strain

         To obtain donor DNA, messenger RNA (mRNA) for preprochymosin
    was isolated from calf stomach and used to prepare complementary DNA
    (cDNA).  The preprochymosin cDNA was cloned into pBR322 plasmid,
    amplified in  E. coli as an intermediate host, and identified by
    sequencing.  DNA encoding prochymosin was then isolated and used for
    the construction of an expression vector on the basis of pUC18
    plasmid.  The prochymosin-encoding DNA was attached to various DNA
    sequences with regulatory or other functions.  DNA coding for the
    lactase promoter of  K. lactis and the prepro-region of the alpha-
    factor gene from  Saccharomyces cerevisiae was inserted in the
    front of the prochymosin-encoding DNA.  alpha-Factor is a sex
    pheromone of  S. cerevisiae that is excreted to the fermentation
    medium.  The secretion is directed by its prepro-region.  Thus, the
    prepro-region of the alpha-factor provided prochymosin with the
    secretion signal.  The lactase transcription termination signal from
     K. lactis was positioned downstream from the prochymosin-encoding
    DNA and was followed by DNA coding for resistance to G418
    aminoglycoside as a selective marker.  Before transformation of the
    host  K. lactis cells, the expression vector was cleaved in the
    lactase promoter.  Since the lactase promoter of the vector was
    identical to the one in the host organism, the vector was integrated
    by homologous recombination into the lactase promoter of the host. 
    Molecular weights of the chymosin derived from  K. lactis and the
    chymosin from calf rennet, as determined by SDS-Polyacrylamide Gel
    Electrophoresis (SDS-PAGE) were found to be identical (Gist-
    brocades, 1990).

    1.1.3  Fermentation

         The fermentation process consists of three steps:  laboratory
    propagation, inoculum fermentation and primary fermentation.  The
    production of prochymosin occurs during the primary fermentation. 
    During the primary fermentation, the production strain is grown
    aerobically under aseptic conditions in a fermenter equipped with
    suitable devices for pH, oxygen, and temperature controls (Gist-
    brocades, 1990).

    1.1.4  Recovery of chymosin

         After approximately 100 hours of fermentation, the process is
    stopped by the addition of sulfuric acid and sodium benzoate.  The
    yeast cells are killed and prochymosin is activated to chymosin due
    to the low pH (ca. 2).  The cellular debris is then separated by
    filtration.  The product is further purified by several cell

    filtrations and concentrated by ultrafiltration.  The concentrate is
    formulated with sodium chloride and sodium benzoate.  The product
    may also be formulated with glycerol, propyleneglycol or sorbitol. 
    The formulation may contain from 10,000 to 46,000 MCU (Milk Clotting
    Units) per mL of the chymosin enzyme, depending on the market
    requirements, and be either liquid or powdered (Gist-brocades,
    1990).

    1.1.5  Biochemical Characterizations

         A number of side-by-side experiments were performed to
    demonstrate that the recombinant chymosin is identical to natural
    chymosin B, including the following (Gist-brocades, 1990):

         1.   DNA Sequencing

         2.   Chromatographic Fractionation

         3.   Molecular Size

         4.   Molecular Weight

         5.   Immunological Identification

         6.   Sequence and Composition of Amino Acids

    2.  BIOLOGICAL DATA

    2.1  Biochemical aspects

         No available data.

    2.1.1  Absorption, distribution, and excretion

         No available data.

    2.2  Toxicological studies

    2.2.1  Acute studies

                                                               
    Species    Route        LD50         Reference
                        (mg/kg/b.w.)
                                                               

    Rat        oral         5000         van Eeken et al., 1986a

                                                               

    2.2.2  Short-term studies

    2.2.2.1  Rat

         Groups of 9-20 male and 10-20 female Wistar rats, 89-111 g
    b.w., were treated by gavage for 91 days with chymosin at daily dose
    levels of 0, 50, 5000 or 1000 mg/kg b.w.  The report did not
    indicate whether the commercial enzyme preparation or a more
    concentrated or purified form of the enzyme was used in the study. 
    No mortality was observed in experimental animals over the course of
    the study.  Growth, behaviour and external appearance were normal. 
    Haematology measurements in treated animals were comparable to
    controls.  Statistically significant changes were occasionally
    observed in clinical chemistry parameters, but the changes were not
    dose-related.  Absolute and relative organ weights were unaffected
    by treatment, except for liver weights in females, which exhibited a
    dose-related increase, tabulated as follows:

              Increase in Liver Weight
              Absolute       Relative

              14.7%          9.1%
              17.6%          13.6%
              18.6%          15.9%

         However, because the incidence of histopathologic lesions in
    livers of treated females was comparable to or less than control
    incidence, the increase in liver weight in treated females is not

    considered to be toxicologically significant.  No treatment effect
    was observed on incidence of gross or microscopic lesions in any of
    the other tissues that were examined (van Eeken  et al., 1988b).

         Groups of 5 male and 5 female Wistar rats, 92-117 g b.w., were
    maintained for 23 days on a commercial rodent diet, which included a
    daily supplement of 5000 mg cheese (prepared with either
    conventional or recombinant chymosin).  No mortality was observed in
    experimental animals over the course of the study.  Behavior and
    external appearance were normal.  Growth of animals consuming diets
    supplemented with either of the two cheeses was enhanced up to 15%
    compared to controls consuming the standard laboratory diet.  No
    significant changes in organ weight could be attributed to
    treatment.  Serum chemistry and haematology values were not affected
    significantly by treatment.  BUN and glucose were elevated in
    animals consuming diets supplemented with either of the two cheeses,
    compared to controls consuming the standard laboratory diet, but
    these changes were probably related to the increased protein and
    caloric intake occurring in animals consuming diets supplemented
    with cheese.  Incidence of gross and microscopic lesions was not
    increased in rats consuming cheese made with the recombinant
    chymosin (van Eeken, et al., 1986b).

         Groups of 10 male and 10 female Wistar rats, 104-142 g b.w.,
    were maintained for 91 days on a commercial rodent diet, which
    included a daily supplement of 5000 mg cheese (prepared with either
    conventional or recombinant chymosin 0.6-0.9 µg chymosin/5000 mg
    cheese).  No mortality was observed in experimental animals over the
    course of the study.  Behavior and external appearance were normal
    except for overgrowth of incisors observed in 3/10 females fed
    cheese made with recombinant chymosin.  Incisor overgrowth has no
    toxicological implications for humans.

         Rats consuming diets supplemented with cheese (made with either
    conventional or recombinant chymosin) grew more rapidly than rats
    consuming a standard laboratory diet, as seen in the following
    table:  

         Growth Enhancement Relative to Standard Laboratory Diet

         Type of Chymosin         Male      Female

         recombinant              14.6%     19.6%
         conventional             19.7%     28.0%

         This increased growth may have occurred because the cheese-
    supplemented diets tasted better to the animals, leading to
    increased food consumption.  There are no food consumption data to
    substantiate this hypothesis.

         Serum chemistry and haematology results were in the normal
    range except for occasional statistically significant deviations
    that were not dose-related and not considered to be treatment-
    related.  Organ weights of animals were not affected by consumption
    of recombinant chymosin in the cheese supplement.  Increased
    absolute heart weight and decreased mean relative weight of kidney
    and liver was observed in males and females consuming diets
    supplemented with either cheese (made with recombinant or
    conventional chymosin).  The apparent increase in heart weight was a
    reflection of the increased overall body weight of cheese-consuming
    animals, and could not be related to intake of recombinant chymosin
    residues in the cheese.  Relative heart weights of cheese-consuming
    animals were comparable to relative heart weights of animals
    consuming a standard laboratory diet.  The changes in relative
    kidney and liver weight were related to the cheese diet, rather than
    to its content of recombinant chymosin.  Incidence of gross and
    microscopic lesions was not increased in rats consuming cheese made
    with recombinant chymosin (van Eeken, 1988c,d).

    2.2.3  Special study on genotoxicity

         The chymosin preparation was tested for potential mutagenicity
    in the standard Ames test.  Negative results were obtained with and
    without microsomal activation (Hoorn, 1988).

    2.2.4  Special study on immune response

         Groups of 10 male and 10 female young adult guinea pigs (376-
    525 g b.w.) were treated dermally with chymosin preparation for 6
    hr, once/wk for three weeks (the induction phase).  Two weeks after
    the last exposure in the induction phase, test animals were dermally
    challenged with chymosin at a previously untreated site. 
    Concurrently, control animals (5 male, 5 female) never previously
    exposed to the test material were also challenged.  The next day,
    application sites were depilated, rinsed, examined and scored for
    the presence of erythema.

         The incidence and severity of erythema in test animals was
    comparable to that observed in the control group.  No response of
    grade 1 severity was observed in test or control animals, although
    an increased incidence of positive response (7/20) occurred in test
    animals, compared to controls (1/10) (Kreuzmann, 1988).

         Antisera were raised in 10 guinea pigs (Dunkin/Hartley/
    Pirbright, 360-436 g.) by intradermal injection of 0.1 ml chymosin
    (5 µg protein/ml) divided over 4 sites on the back.  Two boosters of
    0.1 ml (50 µg protein/ml) were given at 2 week intervals.  Seven
    control animals received comparable injections of saline.  Sera were
    collected via puncture of the retro-orbital sinus at immunization
    (pre-immune serum), at each booster shot, and two weeks after the
    last booster.

         Non-exposed animals were administered freshly prepared serial
    dilutions of sera in a saline (50 µl volume) by intradermal
    injection along each side of the back.  Each recipient animal
    received 6 test samples in duplicate:  one saline control and 5 sera
    from different donor animals.  Four hours after the injection of
    sera, a solution of rennet or recombinant chymosin (150 µg protein)
    together with a 2% solution of Evan's blue (1 ml/kg b.w.) was
    injected intravenously.  Thirty minutes later, areas of the
    "blueing" reactions at the intradermal injection sites were
    estimated.

         Guinea pigs remained healthy throughout the treatment period,
    and growth was normal.  Preimmune sera and saline controls were
    negative in the PCA assay, as was recombinant chymosin, but
    conventional calf rennet gave a positive allergenic response with
    sera from 7/10 donors.  The PCA titers increased with time.  Sera
    taken after 4 weeks and 6 weeks (2 boosters) were positive, but 6
    week sera were positive at higher dilutions (1:32, 1:128) and with
    larger areas of response than sera taken after 4 weeks (1 booster). 
    Scattered positive reactions were observed for 1:8 dilutions of 2/10
    week-6 chymosin sera, but these responses were considered anomalous. 
    It was concluded that recombinant chymosin poses less of an
    allergenic hazard than conventional chymosin (van Lambalgen,
    undated).

    2.2.5  Other relevant safety data

         The genetic stability of the production strain with respect to
    the inserted DNA was confirmed by a Southern blot analysis after 6,
    50 and 100 generations.

         The ampicillin resistance marker normally present in the pUC18
    vector was not expressed in the production strain.

         The chymosin preparation was tested for the presence of
    recombinant DNA and chymosin production organism.  No recombinant
    DNA was detected above the detection limit of 2 ng/ml product, and
    there were no viable  K. lactis cells present in the final
    preparation (Gist-brocades, 1990).

    3.  COMMENTS

         The Committee reviewed the available data pertaining to the
    molecular construction of the chymosin expression plasmid,
    characterization of the producing organism, the fermentation
    process, biochemical and enzymatic characterization of the
    recombinant enzyme and various toxicological studies of the finished
    enzyme product.

         The host organism for the chymosin expression plasmid was
    originally isolated from dairy products, and is a known source of
    the commercial lactase preparation.  It is neither toxicogenic nor
    pathogenic for humans.

         Plasmid pUC18 was used as the cloning vector for the
    prochymosin B coding sequence.  The sequence was generated by a
    procedure involving several steps, starting with the isolation and
    purification of preprochymosin mRNA from the calf stomach.  A
    complementary DNA (cDNA) sequence to the preprochymosin mRNA was
    prepared, cloned into plasmid pBR322, amplified in  Escherichia coli
    as an intermediate host, and identified by DNA sequencing.  The
    isolated prochymosin sequence was inserted into plasmid pUC18, and
    the expression plasmid was integrated into the host genome by
    standard transformation procedures.  The transformed cells were
    identified by their resistance to G418 aminoglycoside.

         The transformed yeast cells were cultured in medium containing
    food-grade substances commonly used in the fermentation process. 
    During cultivation, prochymosin B was secreted into the medium. 
    Fermentation was stopped by acid treatment, which also served to
    activate prochymosin B to chymosin B.  After cellular debris was
    removed by filtration, the chymosin B-containing filtrate was
    further purified by several cell filtrations and concentrated by
    ultrafiltration.  The enzyme concentrate was formulated with
    materials normally applied in the cheese industry.

         The recombinant enzyme was extensively characterized and shown
    to be biochemically, enzymically and immunologically identical to
    bovine chymosin B.  The purity of the recombinant enzyme was
    reported to be much higher than that of rennet extract.  There was
    no evidence of antimicrobial activity.  In addition, there were no
    other proteases, detectable residues of recombinant DNA or viable
    yeast cells present in the purified enzyme preparation.

         Various standard toxicological tests on the recombinant enzyme
    preparation were reviewed by the Committee.  In an acute toxicity
    study in rats given a single oral dose of 5 g/kg b.w. of the enzyme
    preparation no evidence of toxicity was observed.  In studies in
    rats in which cheese prepared with the recombinant chymosin was
    given in the diet at a level of 5 g per day for 21 or 91 days, no
    treatment-related effects were noted.  In another study

    incorporation of the recombinant enzyme preparation into the diet at
    dose levels up to 1 g/kg b.w./day for 90 days did not cause any
    adverse effects in rats.   In vitro genotoxicity studies indicated
    that recombinant chymosin B is not mutagenic.

    4.  EVALUATION

         On the basis of the available safety information and in view of
    the extremely low exposure from its use in food production, the
    Committee established an ADI "not specified" for the recombinant
    chymosin B preparation.

    1.  EXPLANATION

    1.1  Chymosin B produced from  Aspergillus niger var awamori
         containing calf prochymosin B gene

    1.1.2  Construction of production strain

         The host strain for recombinant DNA was derived from  A.
     awamori NRRL 3112 (also referred to as  A. niger var.  awamori).

         A glucoamylase-overexpressing mutant obtained from the NRRL
    3112 strain was subjected to mutagenesis and selection in order to
    isolate two auxotrophic mutants  pyrG (deficient in orotidine-5'-
    monophosphate decarboxylase; requiring uridine for growth) and  argB
    (deficient in ornithine transcarbamylase; requiring arginine for
    growth).  These strains were crossed and a double mutant  pyrG argB
    was isolated from the progeny and designated as GC12.  The purpose
    of these modifications was to provide selectable markers for the
    subsequent genetic manipulations.

         In the next step, the GC12 strain was modified by replacing its
    gene coding for aspergillopepsin A, an extracellular aspartic
    proteinase that can degrade chymosin and cause off-flavors in
    cheese, with the  A. nidulans argB gene.  A linear DNA fragment
    containing 5' and 3' flanking regions from the aspergillopepsin A
    gene and the  A. nidulans argB gene was used to transform the GC12
    strain.  The transformants were selected for the loss of the
    arginine auxotrophy and screened for an aspergillopepsin-deficient
    phenotype.  A strain designated GCdeltaAP4 was selected for the
    transformation with the expression vector.

         The expression vector designated pGAMpR was constructed on the
    basis of pBR322 plasmid.  The vector contains the  A. awamori
    glucoamylase promoter and glucoamylase coding region fused in frame
    with bovine prochymosin cDNA and followed with the  A. niger
    glucoamylase terminator.  The vector also contains the  pyr4 gene
    from  Neurospora crassa which can complement the  pyrG mutation in
    the host strain and allow for the selection of transformants in
    which the requirement for uridine was lost.  The pBR322 sequences
    retained in the expression vector contain the gene coding for
    ampicillin resistance.  However, this gene is not expressed in the
    production strain.

         The aspergillopepsin-deficient strain GCdeltaAP4 was
    transformed with the expression vector.  A transformant producing
    the highest levels of chymosin (based on an immunoassay) was
    selected for chymosin production and designated as strain 4-1.  The
    analysis of cellular DNA with Southern blotting showed that the
    glucoamylase-prochymosin expression unit has integrated into the
    host genome probably in multiple copies.  The tests for mitotic

    stability of the integrated DNA showed that the production strain is
    capable of losing the glucoamylase-prochymosin sequences at a
    frequency of 10-5.  A sporadic formation of morphological variants
    of the production strain not related to mitotic instability was also
    observed.  These variants retained the capability to produce
    chymosin (Chr. Hansen's, 1989).

    1.1.3  Fermentation and recovery

         The production organism is grown in several stages to build up
    the inoculum for large-scale production.  After inoculation to the
    fermenter, the cells are grown aerobically under the proper
    conditions of pH, temperature, nutrient composition, etc.  After
    chymosin reaches a desired level in the fermentation broth, the
    fermentation is stopped and the fungal cells are inactivated and
    separated from the liquid.  Chymosin is then recovered from the
    broth by one of two methods:  (1) the broth is filtered, followed by
    chromatographic purification and concentration of chymosin; or (2)
    the chromatographic step is preceded by extraction of chymosin from
    the fermentation broth.  Chymosin eluted from the chromatographic
    column is formulated to the commercial strength (Chr. Hansen's,
    1989).

    1.1.4  Biochemical characterizations

         A number of side-by-side experiments were performed to
    demonstrate that the recombinant enzyme is identical to calf
    chymosin B, in the following respects:

         1.   Relative mass by SDS-PAGE (Sodium Dodecyl Sulfate-
              Polyacrylamide Gel Electrophoresis)

         2.   Isoelectric point

         3.   Enzymatic activity as a function of pH and temperature

         4.   Other kinetic parameters

         5.   Amino acid composition

         6.   Amino terminal sequencing

         7.   Immunological characteristics

         8.   Stability during storage

         9.   Peptide map

         The recombinant chymosin B differs from calf chymosin B in that
    10% of the recombinant enzyme is glycosylated (Chr. Hansen's 1989).

    2.  BIOLOGICAL DATA

    2.1  Biochemical aspects

         No available data.

    2.1.1  Absorption, distribution, and excretion

         No available data.

    2.2  Toxicological studies

    2.2.1  Acute studies

         No available data.

    2.2.2  Short-term tests

    2.2.2.1  Mouse

         Mice that were fed a single dose of a preparation of the
    production organism were observed for signs of toxicity for a period
    of 4 weeks before they were sacrificed and necropsied.  No toxic
    effects were noted in these animals, indicating that the production
    organism is non-pathogenic after oral ingestion (David, 1989).

    2.2.2.2  Rat

         Four groups of 20 male and 20 female Charles River CD rats, 43
    days old, were maintained for 13 weeks on a diet containing 0, 10,
    40.5 or 162 ppm of lyophilized chymosin.  Average compound intake
    during the study was 0, 0.7, 2.7, or 10.6 mg/kg b.w./day for male
    groups and 0, 0.8, 3.4 or 13.2 mg/kg b.w./day for female groups. 
    Under conditions of the study, test animals fed recombinant chymosin
    in their diet for 13 weeks suffered no significant adverse effects. 
    No compound-related deaths were reported.  Growth and food
    consumption were normal, as were haematology results.  Mean serum
    cholesterol levels were slightly elevated in treated males, but
    standard deviations were large and the elevation is not considered
    to be biologically significant.  Other serum chemistry parameters
    were unaffected by treatment.  Mean urinalysis volumes were large in
    some treatment groups, including controls, but the effect was not
    dose-related.  In these groups, some individual urine volumes
    approached 50 ml, a volume which would not normally be expected from
    rats during the overnight collection period.  The urinalysis summary
    sheet suggested that water spillage might account for the unusually
    large individual urine volumes.  Some of the other urinalysis
    parameters, such as pH and specific gravity, may have been affected
    by the large urine volumes, which makes the urinalysis results
    difficult to interpret.

         Organ weights and incidence of gross and microscopic lesions
    were not affected by treatment.  In females, the incidence of
    chronic focal inflammation in the liver was increased in high-dose
    females (6/20) compared to controls (2/20), but severity was slight
    and the lesion was not considered to be toxicologically significant. 
    Also in high-dose females, the incidence of kidney lesions was
    increased (5/20) compared to controls (2/20), but 3 of the lesions
    were related to mineralization (compared to one such lesion in
    controls), and were considered nutritional in origin, rather than
    treatment-related.  Incidence of the lesions not involving
    mineralization was low and not considered treatment-related (Keller,
    1988).


        2.2.3  Special studies on genotoxicity

                                                                                            
    Test System         Test Object        Concentration of     Results      Reference
                                           Chymosin (Source)
                                                                                            

    Ames test (1)       S. typhimurium     1-150 µl/plate       Negative     Lawlor, 1989
                        TA98, TA100, 
                        TA1535, TA1537, 
                        TA1538

    Ames test (1)       S typhimurium      1-150 µl/plate       Negative     Valentine, 
                        TA98, TA100,                                         1989
                        TA1535, TA1537, 
                        TA1538

    Chromosomal         Chinese hamster    1250-5000 µg/ml      Negative     Murli, 1989
    aberration (2)      ovary cells

    Forward             Mouse lymphoma     0.5-5.0 µl/ml        Negative     Young, 1989
    Mutation (3)

                                                                                            

    (1)  Assays were conducted with and without rat liver S-9 fraction. 
         Chymosin additions were expressed by volume (µl);
         concentrations were not specified.  Positive controls included
         sodium azide (10 µg/plate), 2-nitrofluorene (10 µg/plate),
         quinacrine mustard (5 µg/plate) and 2-anthramine (2.5 µg plate)

    (2)  Assays were conducted both with and without metabolic
         activation.  Chymosin additions were expressed by volume (µl);
         concentrations were not specified.  Positive controls included
         cyclophamide (20-25 µg/ml, metabolic activation) and mitomycin
         C (0.25-0.50 µg/ml, non-activated system).

    (3)  Assays were conducted both with and without metabolic
         activation.  Chymosin additions were expressed by volume;
         concentrations were not specified.  Positive controls included
         3-methylcholanthrene (2.5-4.0 µg/ml, metabolic activation) and
         ethylmethanesulfonate (0.25-0.40 µg/ml, non-activated system).

    (4)  Addition of 20 µl of Chymosin completely inhibited cell growth
         and cell division, but no morphological changes were seen in
         the nucleus or cytoplasm.  No adverse effects were observed on
         cells at additions of 2 µl or less.  Chymosin additions were
         expressed by volume (µl); concentrations were not specified.
    

    2.2.4  Other relevant safety data

         The enzyme preparation was tested for other enzyme activities
    that might be present (peptidase, lipase, amylase, aspergillopepsin,
    etc.); in all cases, the activities were either not detected or
    detected at very low levels.

         Tests were also conducted on the enzyme preparation to
    demonstrate the absence of antimicrobial activities, PEG residues,
    production organism, and mycotoxins (ochratoxin, aflatoxin, T-2
    toxin, etc.) (Chr. Hansen's, 1989).

    3.  COMMENTS

         The Committee reviewed the available data pertaining to the
    molecular construction of the chymosin expression plasmid;
    characterization of the producing organism; the fermentation
    process; biochemical, immunological and enzymatic characterization
    of the recombinant enzyme and the various toxicological studies of
    the finished enzyme product.

         The host organism, GC delta APr4 strain of  Aspergillus niger
    var.   awamori was derived from the parent strain NRRL3112 after a
    series of genetic manipulations.  The host organism was auxotrophic
    and required uridine and arginine for growth, and those properties
    served as selectable markers for subsequent genetic manipulations. 
    The gene coding for aspergillopepsin A, an extracellular aspartic
    proteinase that degrades chymosin and causes "off" flavour in
    cheese, was also absent.

         The expression plasmid (pGAMpR) was constructed on the basis of
    pBR322 plasmid, a widely used cloning vector, the prochymosin B gene
    sequence was obtained from calf stomach tissue. 

         The constructed expression plasmid was integrated into the host
    genome by standard transformation procedures.  The potential
    pathogenicity of the resultant chymosin-B-producing organism was
    studied in mice for 4 weeks following the administration of a single
    oral dose.  No evidence of pathogenicity was observed.

         The cells were grown under properly controlled conditions in
    media containing ingredients commonly used in the production of
    food-grade substances by fermentation.  The cells secreted chymosin
    B in its mature form, which was recovered from the fermentation
    broth by liquid-liquid extraction with polyethylene glycol, after
    removal of the cells.  The resultant crude enzyme was further
    purified by chromatography.  The purified chromosome B was
    formulated to the commercial strength with ingredients typically
    used in commercial enzyme preparations.

         The recombinant chymosin B was extensively characterized and
    shown to be enzymatically and immunologically identical to calf
    chymosin B.  The recombinant enzyme differed from calf chymosin B
    only in the degree of glycosylation, but was otherwise biochemically
    identical.

         The chymosin B preparation was tested for other enzyme
    activities that could be present; in all cases, the activities were
    either not detected or detected at very low levels.  Results from
    additional studies indicated the absence of mycotoxins, of
    antimicrobial activity, of residues of polyethylene glycol and of
    producing organisms.

         In a subchronic (90-day) feeding study in rats, no adverse
    effects were noted at levels up to 10 mg of chymosin preparation kg
    b.w./day.  Negative results were also obtained in a series of
    standard mutagenicity and clastogenicity tests. 

    4.  EVALUATION

         Taking into account the available safety information and the
    extremely low intake resulting from its use in food production, the
    Committee established an ADI "not specified" for the recombinant
    chymosin B preparation.

    5.  REFERENCES

    ANONYMOUS (1988).  Testing of Maxiren 15 TL, lot no. 151/8/LRE, and
    animal rennet, lot no. FH3, for cytotoxic properties in a cell
    culture test.  Unpublished report from Laboratory of Pharmacology
    and Toxicology, Hamburg, F.R.G.

    CHR. HANSEN'S LABORATORIUM (1989).  Submission to WHO by Chr.
    Hansen's Laboratorium, Denmark.

    DAVID, R.M. (1989).  Acute oral toxicity study of  Aspergillus niger
    var.  awamori in mice.  Unpublished report, by Microbiological
    Associates, Inc.  Submitted to WHO by Chr. Hansen's Laboratorium,
    Denmark.  

    FISHER, D.O. (1987a).  CP-72, 692.  A five day exploratory feeding
    study in Beagle dogs.  Unpublished report No. 86-605-02 from Pfizer
    Central Research.  Submitted to WHO by Pfizer, Inc., Groton, CT,
    USA.

    FISHER, D.O. (1987b).  CP-72, 692.  A one month oral gavage study in
    Long-Evans rats.  Unpublished report No. 86-605-01 from Pfizer
    Central Research.  Submitted to WHO by Pfizer, Inc., Groton, CT,
    USA.

    GIST-BROCADES (1990).  Submission to WHO by Gist-brocades NV, Delft,
    Holland.

    HOORN, A.J.W. (1988).  Mutagenicity test on chymosin batch no. 7005
    in the Ames  Salmonella/microsome reverse mutation assay, p. 532. 
    Unpublished report No.  E-9792-0-401 from Hazleton Biotechnologies,
    Veenendaal, Holland.  Submitted to WHO by Gist-brocades NV, Delft,
    Holland.

    KELLER, K.A. (1988).  Thirteen week dietary toxicity study in rats. 
    Unpublished report No. 540-026 from International Research and
    Development Corporation.  Mattawan, MI, USA.  Submitted to WHO by
    Genencor, Inc., South San Francisco, CA, USA.

    KREUZMANN, J.J. (1988).  Delayed contact hypersensitivity study in
    guinea pigs.  Unpublished report No. 88-3198-21 from Hill Top
    Biolabs, Inc., Cincinnati, OH, USA.  Submitted to WHO by Genencor,
    Inc., South San Francisco, CA, USA.

    LAWLOR, T.E. (1989).  Mutagenicity test on chymosin lot P88032FP in
    the Ames salmonella/microsome reverse mutation assay.  Unpublished
    report No. 20988 from Hazleton Laboratories America, Inc.,
    Kensington, MD, USA. Submitted to WHO by Genencor, Inc., South
    Francisco, CA, USA.

    MURLI, H. (1989).  Mutagenicity test on chymosin lots P88032FP,
    P88035FP and P88040FP in an in vitro cytogenetic assay measuring
    chromosomal aberration frequencies in Chinese hamster ovary (CHO)
    cells.  Unpublished report No. 20990 from Hazleton Laboratories
    America, Inc., Kensington, MD, USA.  Submitted to WHO by Genencor,
    Inc., South San Francisco, CA, USA.

    MYER, J.R. (1988a).  Eye irritation study in rabbits (low volume
    procedure).  Unpublished report No. 540-028 from International
    Research and Development Corporation.  Mattawan, MI, USA.  Submitted
    to WHO by Genencor, Inc., South San Francisco, CA, USA.

    MYER, J.R. (1988b).  Primary dermal irritation test in rabbits. 
    Unpublished report No. 540-027 from International Research and
    Development Corporation.  Mattawan, MI, USA.  Submitted to WHO by
    Genencor, Inc., South San Francisco, CA, USA.

    PFIZER CENTRAL RESEARCH (1988).  International submission on
    chymosin.  Submitted to WHO by Pfizer, Inc., Groton, CT, USA.

    VALENTINE, D.C. & LAWLOR, T.E. (1989).  Mutagenicity  test on
    chymosin lots P88035FP and P88040FP in the Ames  salmonella/
    microsome reverse mutation assay.  Unpublished report No. 20988 from
    Hazleton Laboratories America, Inc., Kensington, MD, USA.  Submitted
    to WHO by Genencor, Inc., South San Francisco, CA, USA.

    Van EEKEN, J.J., ABOULWAFA-Van VELTHOVEN, M.J.E., & BIRTHWISTLE,
    R.D.R. (1986a).  An investigation in the rat of the acute oral
    toxicity of Chymosin.  Unpublished report No. 8605 from Gist-
    brocades NV, Delft, Holland.  Submitted to WHO by Gist-brocades, NV,
    Delft, Holland.

    Van EEKEN, J.J., ABOULWAFA, M.J.E., & BIRTHWISTLE, R.D.R. (1986b). 
    Short term toxicity study in Wistar rats after addition of Chymosin
    prepared cheese to the feed.  Unpublished report No. 8604 from Gist-
    brocades NV, Delft, Holland.  Submitted to WHO by Gist-brocades, NV,
    Delft, Holland.

    Van EEKEN, J.J., ABOULWAFA, M.J.E., & BIRTHWISTLE, R.D.R. (1988c). 
    Ninety-one day toxicity study after oral administration of Chymosin
    to Wistar rats.  Unpublished report No. 8711 from Gist-brocades NV,
    Delft, Holland. Submitted to WHO by Gist-brocades, NV, Delft,
    Holland.

    Van EEKEN, J.J., ABOULWAFA, M.J.E., & BIRTHWISTLE, R.D.R. (1988d). 
    Ninety-one day feeding study in Wistar rats after addition of
    Chymosin prepared cheese to the feed.  Unpublished report No. 8706
    from Gist-brocades NV, Delft, Holland.  Submitted to WHO by Gist-
    brocades, NV, Delft, Holland.

    Van LAMBALGEN, R. (undated).  Allergenicity of chymosin tested by
    passive cutaneous anaphylaxis (PCA) assay.  Unpublished report from
    Gist-brocades NV, Delft, Holland.  Submitted to WHO by Gist-brocades
    NV, Delft, Holland.

    YOUNG, R.R. (1989).  Mutagenicity, test on chymosin lots P88035FP
    and P88040FP in the mouse lymphoma forward mutation assay. 
    Unpublished report No.  20989 from Hazleton Laboratories America,
    Inc., Kensington, MD, USA.  Submitted to WHO by Genencor, Inc.,
    South San Francisco, CA, USA.


    See Also:
       Toxicological Abbreviations