Thaumatin is a mixture of intensely sweet proteins (thaumatins)
    extracted with water from the arils of the fruit of the West African
    perennial plant Thaumatococcus daniellii. The thaumatins have a
    normal complement of amino acids, except that histidine is not
    present. The molecular weights of the thaumatins are approximately
    22,000 and their iso-electric points are in the range of 11.5-12.5.
    There are no unusual side-chains, atypical peptide linkages, or
    end-groups. Extensive disulfide cross-linking confers to thaumatin
    thermal stability, resistance to denaturation, and maintenance of the
    tertiary structure of the polypeptide chain. The maintenance of
    tertiary structure is critical to thaumatin's technical function.
    Cleavage of just one disulfide bridge results in a loss of sweet
    taste. (Iyengar et al., 1979).

         Thaumatin is purified by selective ultrafiltration, but small
    amounts of organic non-protein impurities remain in the commercial
    product. These consist principally of the arabinogalactan and
    arabinoglucuronoxylan polysaccharides, both of which are normal
    constituents of plant gums and mucilages.

         Thaumatin functions primarily as a flavour enhancer and as a
    high-intensity sweetener. The substance was previously evaluated by
    the Committee at its twenty-seventh meeting (Annex 1, reference 62).
    No ADI was allocated at that time, although specifications were


    Biochemical aspects


         In vitro studies with purified mammalian digestive enzymes,
    using a sequential enzyme system simulating the human digestive tract
    and a rapid multienzyme system, resulted in thaumatin being digested
    more rapidly than egg albumin (Hsu et al., 1977; Higginbotham,

         A cross-over nitrogen balance study was performed over two 10-day
    periods using 2 groups of 10 male rats. During the first period, Group
    1 was fed a semi-synthetic diet containing 10% protein from egg
    albumin, while a similar diet, but containing 5% egg albumin and 5%
    thaumatin, was fed during the second period. Animals in Group 2
    received the diets in the reverse order. This procedure was used to
    compensate for the lack of histidine in thaumatin, which is an

    essential amino acid in the rat. The nitrogen digestibility of both
    diets was approximately 90%, but no allowance was made for the loss of
    endogenous nitrogen. The author concluded that the digestibility of
    thaumatin is at least equal to that of egg albumin. However, the
    biological value of thaumatin was lower than that of egg albumin,
    which reflects the lack of histidine (Edwards, 1981).

    Toxicological Studies

    Special studies on mutagenicity

         At doses up to 50 mg/plate, thaumatin was not mutagenic in the
    Ames test using Salmonella typhimurium strains TA98, TA100, TA1535,
    TA1537, or TA1538 or in Escherichia coli strain WP2, either with or
    without S-9 mix (Higginbotham, 1980; Higginbotham et al., 1983).

         In a dominant-lethal assay, groups of 15 male CD1 mice were
    intubated with thaumatin at 200 or 2000 mg/kg/day for 5 days, with
    trimethyl phosphate at 100 mg/kg/day for 5 days (positive control), or
    with the vehicle, distilled water (negative control). Following
    completion of treatment, each male was paired with 3 untreated females
    for a maximum of 7 days each week, and this was repeated for 7
    consecutive weeks. In the positive control group, fertilization was
    unaffected but, during the initial 2 weeks following treatment, a
    marked loss of embryos and foetuses occurred and the normality and
    rate of development of zygotes were adversely affected. On the other
    hand, mating performance and fertility of males, fertilization of ova,
    development and survival of zygotes, and pre- and post-implantation
    losses were unaffected by treatment with thaumatin, as assessed in
    animals sacrificed after 4, 11, or 18 days post coitum. The authors
    concluded that, under the conditions of the study, thaumatin did not
    induce dominant-lethal mutations in the gametes of male mice
    (Tesh et al., 1977a).

    Special study on teratogenicity

         Thaumatin was administered by gavage to groups of 20 pregnant CD
    rats from day 6 to day 15 of gestation, inclusive, at dosages of 0.2,
    0.6, or 2.0 g/kg/day. A control group received the vehicle, distilled
    water, at the same dosage volume of 10 ml/kg throughout the same
    period. On day 21 of gestation, all rats were killed and the number of
    corpora lutea in each ovary and the number, position, and condition of
    implantations were recorded. Viable foetuses were weighed, sexed, and
    examined externally. The thoracic and abdominal cavities of the
    remainder were dissected, examined, and then processed for subsequent
    skeletal examination. No adverse effects were observed in the pregnant
    females nor on litter responses (litter size, foetal weight, or pre-
    and post-implantation losses). No visceral or skeletal abnormalities
    in the foetuses attributable to the treatment were observed
    (Tesh et al., 1977b).

    Special studies on allergenicity

         Studies on ileum preparations from groups of 5 guinea-pigs that
    had been sensitized by i.m. injections of either 50 mg thaumatin or
    50 mg egg albumin (together with incomplete adjuvant, which is
    preferred for evoking an IgE antibody response), revealed that the
    minimum dose of thaumatin capable of eliciting a response was 250 ng.
    This is comparable to the minimum dose of egg albumin needed to evoke
    an anaphylactic response in the gut of control animals sensitized with
    egg albumin. Tests on ileum preparations from guinea-pigs sensitized
    with protein and complete (Freund's) adjuvant gave an essentially
    similar result, although thaumatin was slightly more sensitizing,
    50 ng evoking a response, whereas no effect was elicited with a
    similar dose of egg albumin (Stanworth, 1977).

         Studies in rats that had been sensitized by s.c. injections of
    either 10 g thaumatin or 10 g egg albumin (together with Freund's
    adjuvant) showed that the levels of anaphylactic antibody in sera,
    determined by passive subcutaneous anaphylaxis dilution-titration,
    were uniformly lower with thaumatin than with egg albumin (Stanworth,

         In vitro studies on the ability of thaumatin to initiate
    histamine release non-immunologically showed that approximately 1 mM
    thaumatin was needed to cause a release of 50% of the total available
    histamine from a purified rat-mast cell preparation. Synacthen (ACTH
    1-24polypeptide), the standard histamine-releasing agent used in this
    study, released 50% of the total available histamine at a much lower
    concentration, 2 M. This finding was confirmed in vivo in normal
    baboons; intradermal injection of thaumatin at a concentration of 1 mM
    immediately after an i.v. injection of Evan's blue dye gave only a
    weak "blueing" response, whereas Synacthen gave a measurable response
    at a concentration of 1 M (Stanworth, 1977).

    Acute toxicity

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

    Mouse       oral       > 20,000        Ben-Dyke, 1975

    Rat         oral       > 20,000        Ben-Dyke &
                                           Joseph, 1976

    Short-term studies


         In a preliminary 90-day study, groups of 10 male and 10 female CD
    rats were fed thaumatin at dietary levels of 1.0, 4.0, or 8.0% w/w. A
    control group received a basal diet supplemented with 8.0% casein w/w
    to compensate for the high protein intake.

         No deaths occurred in any group. The body-weight gains of male
    rats fed 4.0 or 8.0% thaumatin were lower (6% and 9%, respectively)
    than those of casein-fed controls. Body-weight gains of female rats
    were not affected by treatment. Food consumption of male and female
    rats receiving 4.0 or 8.0% thaumatin was 5-11% lower than that of
    casein-fed controls. A statistically-significant reduction
    (3.2%, p < 0.05) in haemoglobin content in the blood of female rats
    fed 8.0% thaumatin was observed, although the values were within the
    range normally found for animals of comparable age and weight.
    Otherwise, the cellular and chemical composition of blood and urine
    were unaffected by treatment.

         Tests for hepatic (bromosulfonphthalein retention) and renal
    function (concentrating ability) during week 12 in rats fed 8.0%
    thaumatin revealed responses that were similar to those obtained in
    the casein-control group. Using the Ouchterlony diffusion technique,
    no evidence was obtained for the presence of antibodies to thaumatin
    in the serum of animals treated with thaumatin for 13 weeks. Apart
    from a statistically-significant increase (p < 0.05) in both the
    absolute (14%) and relative (17%) liver weights of female rats
    receiving 8.0% thaumatin, detailed macroscopic and microscopic
    examination of a wide range of tissues revealed no changes which could
    be attributed to treatment (Ben-Dyke et al., 1976).

         In a second 90-day study, groups of 20 male and 20 female CD rats
    received thaumatin at dietary levels of 0, 0.3, 1.0, or 3.0% w/w.
    During the treatment period, body weights and food consumption were
    recorded weekly. Additionally, the liver, kidneys, heart, lungs,
    spleen, and brain from all animals fed 0.3 or 1.0% thaumatin were
    examined microscopically.

         No visible signs of reaction to treatment were observed. No
    animals died as a result of treatment with thaumatin. Body-weight
    values for males receiving 3.0% thaumatin were 6% higher during the
    study period than those of the control group and females receiving
    1.0% thaumatin showed a 6% reduction in body weight from 5 weeks
    onward. Food consumption of females receiving 3.0% thaumatin was 7%
    lower than that of the controls throughout the treatment period. No
    treatment-related effects were observed on water intake or on
    ophthalmoscopic examination. Blood and serum analyses showed
    significant increases of 8 and 13% in haematocrits at week 12 in male

    rats fed 1.0 or 3.0% thaumatin, respectively, and significant
    reductions of 10 and 12% in haematocrits in female rats in the 0.3 and
    1.0% groups, respectively. A dose-related reduction in triglyceride
    levels of up to 60% was observed in females during weeks 4 and 12 of
    treatment. Urinalysis was unaffected by treatment.

         No treatment-related changes were seen during a detailed
    macroscopic examination of all animals at necropsy. An 8% increase in
    the kidney weights of treated females was observed; when related to
    body weight, the increase was 13%. Thyroid weights were significantly
    higher in all male treated groups than in controls and significantly
    lower in all female treated-groups than in controls, both in absolute
    weight and when related to body weight. The authors did not consider
    the changes in thyroid weights to be of toxicological significance,
    since the mean absolute thyroid weights of the control males were
    abnormally low compared with historical control values; the mean
    values for treated males were within the historical control values
    cited by the author. On the other hand, the mean absolute thyroid
    weights of control females were within the expected range when
    compared to historical controls, giving an apparent treatment-related
    decrease in absolute and body-weight-related thyroid weights. No
    treatment-related changes were seen in the tissue sections taken from
    the control and 3.0% groups, which were examined microscopically. A
    subsequent microscopic examination of thyroid tissues from all animals
    in all groups revealed no treatment-related changes (Hiscox et al.,
    1981; Wood, 1984).

         Confirmation of the lack of hypo- or hyperthyroid effects was
    obtained from an additional study in CD rats. Two groups, 10 males and
    10 females in each, were fed 3% thaumatin or 3% egg albumin in the
    diet for 4 weeks. Blood samples taken from the animals after the
    treatment period were analysed for thyroxine (T4) and
    triiodothyronine (T3). No statistically-significant differences in
    thyroid hormone levels were observed between animals fed thaumatin and
    those fed egg albumin. The authors concluded that thaumatin had no
    effect on thyroid function in rats at a dietary level of 3% (Danks
    et al., 1984).


         Groups of 4 male and 4 female beagle dogs were fed thaumatin in
    the diet for a minimum of 90 days at levels of 0, 0.3, 1.0, or
    3.0% w/w. During the treatment period, body weights were recorded
    weekly and food consumption daily. At the end of the treatment period,
    a detailed macroscopic examination was performed and weights of the
    adrenals, brain, heart, kidneys, liver, lungs, ovaries, pituitary,
    prostate, spleen, testes, thyroid, and uterus were recorded.
    Microscopic examinations were performed on these and a wide range of
    other tissues from all animals.

         No deaths occurred and there were no overt signs of reaction to
    treatment. Males receiving thaumatin showed slightly increased body
    weights relative to controls. Food consumption and water intake were
    unaffected by treatment. Ophthalmoscopic examination did not reveal
    any changes that could be related to treatment. Haematological
    examination of blood samples during weeks 4 and 12 of treatment
    revealed slight decreases in haemoglobin concentrations, erythrocyte
    counts, and haematocrits in males fed 3.0% thaumatin. However, the
    values were within the range determined from historical controls.
    Biochemical examination of blood samples did not reveal any treatment-
    related effects and urinalysis was unaffected by treatment.
    Macroscopic pathology at termination did not reveal any treatment-
    related changes. An increase in absolute liver weight (20%) was
    observed in males fed 3.0% thaumatin. When related to body weight,
    organ weights showed no treatment-related variations. Microscopic
    examination revealed no changes that the authors considered to be
    related to the administration of thaumatin (Barker et al., 1981).

    Long-term studies

         No information available.

    Observations in man

         The intense sweetness of the fruit of Thaumatococcus daniellii
    was first described by a British surgeon in the Pharmaceutical
    Journal (Daniell, 1855).

         The long history of human use of the fruit as a sweetener, now
    largely displaced by sugar in urban areas, and the absence of unusual
    or toxic effects following its ingestion, is attested by numerous
    affidavits from village elders in Ghana and the Ivory Coast
    (Higginbotham & Stephens, 1984).

         Thaumatin was assessed for oral allergenicity in humans by giving
    100 mg/day thaumatin or lactose in gelatin capsules to 4 women and 6
    men for a period of 14 days using a double-blind cross-over design.
    The volunteers were randomly assigned to 2 groups of 5 each and given
    either the test substance or lactose. All volunteers were prick-tested
    for common allergens and with a solution of thaumatin before the
    study. Seven volunteers were tested with thaumatin a second time
    before the study commenced to determine if sensitization to thaumatin
    could result from prick-testing itself. No sensitization was observed.
    At the completion of the study a further prick test was performed to
    determine if sensitization had occurred following thaumatin ingestion.
    No sensitization was detected. Blood was taken from the volunteers at
    the commencement and again after the 28-day study period. Examination
    of the sera for antibodies to thaumatin by the passive subcutaneous

    anaphylaxis technique in baboons and rhesus monkeys showed no
    reactions when challenged s.c. or orally with thaumatin. Clinical
    assessment showed no treatment-related allergic effects (Eaton et
    al., 1981).

         Thaumatin was assessed for oral sensitivity and irritation in
    humans. Chewing gum containing 150 ppm thaumatin was administered to
    25 volunteers, who each chewed five 5.3-gram gum sticks per day, each
    stick for 15 minutes, over a period of 28 days. A similarly-
    constituted group of 25 volunteers received untreated gum. Allocation
    of volunteers to test and control groups was random and under a
    double-blind code. No weal or flare reactions were noted in any
    volunteers after prick-testing either before or after the treatment
    period, nor were any visible signs of irritant or allergic responses
    detected on the oral mucosa after chewing either treated or untreated
    gum. The authors concluded that, under these conditions, thaumatin did
    not cause irritation of the oral mucosa or any allergic responses
    (McLeod et al., 1981).

         A clinical study was conducted to determine the effects of
    thaumatin on haematological and blood biochemistry parameters.
    Eighteen male and 12 female volunteers were randomly assigned to 2
    groups. Each participant was given a weekly supply of capsules, each
    containing either 280 mg thaumatin or 210 mg egg albumin, and asked to
    ingest one capsule each morning at 9 o'clock. This procedure was
    followed for 13 consecutive weeks. The capsules were coded and their
    composition was known only to the physician and consultant
    pathologist. Blood was collected and analysed during the week
    preceding the start of the trial and subsequently after 4, 8, and 12
    weeks. No treatment-related changes in either the chemical or cellular
    composition of the blood were observed in volunteers consuming
    thaumatin when compared to the control group. The cumulative intake of
    thaumatin by these subjects was 25 g, which is some 140 times the
    estimated maximum consumer intake over this period (Tompkins &
    Enticknap, 1984).

         Prick-testing of laboratory personnel who had inhaled thaumatin
    intermittently for periods up to 7 years showed that about one-half
    (67/140) responded to common inhalant allergens. A positive response
    to thaumatin was observed in 13 subjects, all except one of whom were
    atopic or allergic (Higginbotham et al., 1983).


         There is no evidence that thaumatin is treated differently than
    other proteins with respect to hydrolysis or digestion. No antibodies
    to thaumatin were detected in either rats or humans after prolonged
    oral administration of quantities of thaumatin that substantially
    exceed the anticipated human exposure, thus indicating that the intact
    protein is not absorbed, and confirming the digestibility of

    thaumatin. The possibility that hormonally-active polypeptides are
    present in digests of thaumatin, and that these may be absorbed intact
    and retain their activity, is unlikely because endocrine disturbances
    were not observed in toxicological studies.

         Thaumatin showed no mutagenic or teratogenic effects and no
    allergenic effects were noted.

         Variations in thyroid weights in a 90-day rat study (increases in
    males and decreases in females) revealed no treatment-related
    histological abnormalities; hypo- or hyperthyroid effects were not
    observed in a follow-up study in which statistically-significant
    differences in thyroid hormone levels (T3 and T4) were not observed.

         Slight changes in haemoglobin concentrations, red blood cell
    counts, and packed-cell volumes observed in rats and dogs fed up to
    3.0% thaumatin were not observed in a 13-week clinical study in human
    volunteers ingesting levels of thaumatin on the order of 140 times
    higher than the anticipated maximum daily intake, which has been
    calculated to be 1-2 mg/person/day.

         The lack of toxicity, combined with its ready digestion to normal
    food components, indicate that thaumatin's only dietary effect is to
    make an insignificant contribution to the normal protein intake.


    Estimate of acceptable daily intake for man

    ADI "not specified".


    Barker, J.D., Hiscox, D.N., & Wood, C.M. (1981). Talin protein: 90-day
         toxicity study in the dog by dietary admixture. Unpublished
         report No. TAL/2/81 from Toxicol Laboratories, Ltd., Ledbury,
         England. Submitted to WHO by Tate & Lyle PLC.

    Ben-Dyke, R. (1975). Talin: Acute oral toxicity in mice. Unpublished
         report No. 75/TYL2/058 from Life Science Research, Stock,
         England. Submitted to WHO by Tate & Lyle PLC.

    Ben-Dyke, R. & Joseph, E.C. (1976). Talin: Acute oral toxicity in
         rats. Unpublished report No. 76/TYL5/131 from Life Science
         Research, Stock, England. Submitted to WHO by Tate & Lyle PLC.

    Ben-Dyke, R., Ashby, R., & Newman, A.J. (1976). Talin: Toxicity in
         dietary administration to rats for thirteen weeks. Unpublished
         report No. 76/TYL4/188 from Life Science Research, Stock,
         England. Submitted to WHO by Tate & Lyle PLC.

    Daniell, W.F. (1855). Katemfe, or the miraculous fruit of
         Soudan. Pharm. J., 14, 158.

    Danks, A., Hooks, W., Ashby, R., & Whitney, J.C. (1984). Talin: Four-
         week dietary study in rats to investigate thyroid function.
         Unpublished report No. TYL/073/TAL from Life Science Research,
         Stock, England. Submitted to WHO by Tate & Lyle PLC.

    Eaton, K.K., Daniel, J.W., Snodin, D.J., Higginbotham, J.D.,
         Stanworth, D.R., & Al-Mosawie, T. (1981). Talin protein:
         Assessment in man for oral allergenicity on challenge testing.
         Unpublished report submitted to WHO by Tate & Lyle PLC.

    Edwards, D.G. (1981). Talin (Thaumatin): Nitrogen digestibility in the
         rat. Unpublished report No. B.128 from RHM Research Ltd., High
         Wycombe, England. Submitted to WHO by Tate & Lyle PLC.

    Higginbotham, J.D. (1978). The digestibility of Talin protein in
         vitro. Unpublished report from Tate & Lyle PLC, Reading,
         England. Submitted to WHO by Tate & Lyle PLC.

    Higginbotham, J.D. (1980). Mutagenicity testing of Talin protein
         sweetener in vitro. Unpublished report from Tate & Lyle PLC,
         Reading, England. Submitted to WHO by Tate & Lyle PLC.

    Higginbotham, J.D., Snodin, D.J., Eaton, K.K., & Daniel, J.W. (1983).
         Safety Evaluation of Thaumatin (Talin protein). Fd. Chem.
         Toxicol., 21, 815-823.

    Higginbotham, J.D., & Stephens, J.P. (1984). Food uses of
         Thaumatococcus daniellii in West Africa. Unpublished report
         from Tate & Lyle PLC, Reading, England. Submitted to WHO by Tate
         & Lyle PLC.

    Hiscox, D.N., Hill, R.E., & Wood, C.M. (1981). Talin protein: 90-day
         toxicity study in the rat by dietary admixture. Unpublished
         report No. TAL/1/81 from Toxicol Laboratories Ltd., Ledbury,
         England. Submitted to WHO by Tate & Lyle PLC.

    Hsu, H.W., Vavak, D.L., Satterlee, L.D., & Miller, G.A. (1977). A
         multienzyme technique for estimating protein digestibility.
         J. Food Sci., 42, 1269-1273.

    Iyengar, R.B., Smits, P., van der Ouderaa, F., van der Wel, H., van
         Brouwershaven, J., Ravestein, P., Richters, G., & van Wassenaar,
         P. (1979). The complete amino-acid sequence of the sweet protein
         thaumatin. I. Eur. J. Biochem., 96, 193-204.

    MacLeod, G.L., Eaton, K.K., Daniel, J.W., Snodin, D.J., Higginbotham,
         J.D., & Waite, D. (1981). Assessment of oral sensitisation and
         irritation when formulated in peppermint chewing gum. Unpublished
         report submitted to WHO by Tate & Lyle PLC.

    Stanworth, D.R. (1977). Preliminary assessment of the potential
         allergenicity of the sweet protein, Talin. Unpublished research
         report from the University of Birmingham, England. Submitted to
         WHO by Tate & Lyle PLC.

    Tesh, J.M., Davidson, E.J., & Willoughby, C.R. (1977a). Talin: Test
         for dominant lethality in the male mouse. Unpublished report No.
         77/TYL11/096 from Life Science Research, Stock, England.
         Submitted to WHO by Tate & Lyle PLC.

    Tesh, J.M., Earthy, M., Tesh, S.A., & Willoughby, C.R. (1977b). Talin:
         Effects of oral administration upon pregnancy in the rat.
         Unpublished report No. 77/TYL10/179 from Life Science Research,
         Stock, England. Submitted to WHO by Tate & Lyle PLC.

    Tompkins, G.D. & Enticknap, J.B. (1984). A comparison of the effects
         on the chemical and cellular composition of blood following the
         administration of thaumatin and egg albumin to human subjects for
         13 weeks. Unpublished report submitted to WHO by Tate & Lyle PLC.

    Wood, C.M. (1984). 90-day toxicity study in the rat by dietary
         admixture; further microscopic examination of thyroids from rats
         in study TAL/1/81. Unpublished report from Toxicol Laboratories
         Ltd., Ledbury, England. Submitted to WHO by Tate & Lyle PLC.

    See Also:
       Toxicological Abbreviations
       THAUMATIN (JECFA Evaluation)