IPCS INCHEM Home


    INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY

    WORLD HEALTH ORGANIZATION





    SAFETY EVALUATION OF CERTAIN FOOD
    ADDITIVES AND CONTAMINANTS



    WHO FOOD ADDITIVES SERIES: 44





    Prepared by the Fifty-third meeting of the Joint FAO/WHO
    Expert Committee on Food Additives (JECFA)





    World Health Organization, Geneva, 2000
    IPCS - International Programme on Chemical Safety

    THICKENER



    CURDLAN

    First draft prepared by M.E. von Apeldoorn & G.J.A. Speijers

    Section on Public Health, Centre for Substances and Risk Assessment,
    National Institute of Public Health and the Environment, Bilthoven,
    Netherlands

    Explanation
    Biological data
         Biochemical aspects
              Absorption, distribution and excretion
              Mineral bioavailability in rats
              Nutritional aspects
    Toxicological studies
         Acute toxicity
         Short-term studies of toxicity
         Long-term studies of toxicity and carcinogenicity
         Genotoxicity
         Reproductive toxicity
         Special studies
              Immunogenicity
              Pathogenicity of  Alcaligenes faecalis var.  myxogenes NTK-u
              strain
              Cytotoxicity of  Alcaligenes faecalis var.  myxogenes NTK-u
              strain to HeLa cells
    Observations in humans
         Sensitization
         Studies of adverse side-effects
    Comments
    Evaluation
    References

    1.  EXPLANATION

         Curdlan is a linear polymer consisting of -(1 -> 3)-linked
    glucose residues. It is derived by fermentation from the bacterium
     Alcaligenes faecalis var.  myxogenes. The Committee considered its
    use in food as a formulation aid, processing aid, stabilizer and
    thickener, or texturizer. Curdlan has not been evaluated previously by
    the Committee.

         Information was available on the current per-capita intake of
    curdlan in Japan and from a scenario based on levels of use of the
    additive and food consumption data in the United States. This
    information was inadequate to assess intake completely because data on
    maximum use and the distribution of intake of foodstuffs that might
    contain the additive in various regions of the world were not
    provided.

    2.  BIOLOGICAL DATA

    2.1  Biochemical aspects

    2.1.1  Absorption, distribution, and excretion

         Rats

         Two male Sprague-Dawley rats, seven weeks old, received an oral
    dose of 20 mg/kg bw of [U-14C]curdlan prepared from [U-14C]glucose
    with  Alcaligenes faecalis var.  myxogenes strain NTK-u suspended in
    water. Expired carbon dioxide, urine, and faeces were collected for 72
    h and the excreted radiolabel was determined. Considerable amounts of
    radiolabel were excreted in the expired carbon dioxide (Table 1;
    Matsuo & Suzuoki, 1972).

         In a comparison of the metabolism of curdlan and cellulose, two
    groups of three male Wistar rats, six weeks of age, received
    [14C]curdlan or [14C]cellulose at a dose of 20 mg/kg bw suspended in
    water. Expired carbon dioxide, urine, and faeces were collected for 48
    h and the excreted radiolabel was determined. After dosing with
    [14C]curdlan, excretion of [14C]carbon dioxide was low during the
    first 3 h but increased linearly up to 12 h and reached a plateau of
    39% of the administered radiolabel. After dosing with [14C]cellulose,
    33% was excreted as [14C]carbon dioxide after 24 h. In both cases,
    excretion as [14C]carbon dioxide was initially delayed (an effect not
    seen in studies with starch or glucose). Less than 2% of the curdlan
    and cellulose was excreted in the urine within 48 h, and faecal
    excretion within this period was 38% for curdlan and 54% for cellulose
    (Table 2; Matsuo & Suzuoki, 1972).

         When groups of three male Wistar rats aged six weeks received
    tetracycline in the drinking-water at a dose of 5 mg/ml for five days
    before and two days after oral dosing with [14C]curdlan, excretion as
    [14C]carbon dioxide was decreased to about one-third, whereas faecal
    excretion was increased by 2.7 over that in male Wistar rats receiving
    [14C]curdlan only (Table 3). The results indicate that intestinal
    microflora are partly responsible for the metabolism of curdlan to
    carbon dioxide (Matsuo & Suzuoki, 1972).

         Three male Wistar rats, five weeks old, received an oral dose of
    2.3, 23, or 230 mg/kg bw [14C]curdlan suspended in water. Excretion
    as [14C]carbon dioxide decreased, with an accompanying increase in
    fecal excretion, as the dose was increased from 23 mg/kg bw,
    indicating limited metabolism at higher doses (Matsuo & Suzuoki,
    1972).

         Groups of 10 male SD/Ta rats received 0, 3, 10, or 30% curdlan or
    30% agar powder in their diet for 30 days. The 24-h faeces from rats
    at 30% curdlan, analysed for curdlan after 14 days, contained 4.3 g.
    As the average food consumption during week 2 was 27 g and the
    ingested food contained 0.3  27 g, or 8.1 g curdlan, the recovery of
    curdlan in the faeces was 54%. The faeces also contained 2.5%
    water-soluble sugar. Curdlan recovered from the faeces had lost the
    ability to gel on heating (Yokotani et al., 1969).

        Table 1. Recovery of radiolabel after oral administration of [14C]curdlan
                                                                            

    Collection period        Recovery of radiolabel as % of administered dose
                                                                            
                             Carbon dioxide   Urine     Faeces      Total
                                                                            

    0-24 h                   77               2.6        7.7         88
    0-48 h                   85               3.1       12          100
    0-72 h                   89               3.3       12          100
                                                                            

    Table 2. Recovery of radiolabel after oral administration of [14C]curdlan
             or [14C]cellulose
                                                                             

    Collection period       Recovery of radiolabel as % of administered dose
                                                                           
                            Carbon dioxide   Urine     Faeces     Total
                                                                             

    Curdlan
    0-24 h                  39               1.3       34         74
    0-48 h                  40               1.4       38         80

    Cellulose
    0-24 h                  33               1.6       40         75
    0-48 h                  36               2.0       54         92
                                                                             
    
         Two male Sprague-Dawley rats, seven weeks of age, received 20
    mg/kg bw [14C]curdlan suspended in water intraperitoneally and were
    killed 0.5, 3, 6, and 24 h later, the animals excreted extremely
    little radiolabel (Table 4), indicating that most of the curdlan
    remained in the body. Whole-body radioautography showed that the
    radiolabel was distributed in the intestinal fluids (Matsuo & Suzuoki,
    1972).

        Table 3. Recovery of radiolabel after oral administration of [14C]curdlan
             with and without tetracycline
                                                                                   

    Collection period               Recovery of radiolabel as % of administered dose
                                                                                   
                                    Carbon dioxide   Urine      Faeces     Total
                                                                                   

    0-24 h
      Curdlan + tetracycline        18               0.68       63         82
      Curdlan                       54               2.0        24         80

    0-48h
      Curdlan + tetracycline        20               0.98       69         90
      Curdlan                       56               2.3        26         86
                                                                                   

    Table 4. Recovery of radiolabel after intraperitoneal administration
             of [14C]curdlan
                                                                           

    Collection period     Recovery of radiolabel as % of administered dose
                                                                           
                          Carbon dioxide   Urine      Faeces      Total
                                                                           
    0-24 h                1.8              3.5        0.05        5.4
    0-48 h                4.1              4.1        0.12        6.4
                                                                           
    
         Humans

         Four healthy male volunteers, two of whom were pretreated with an
    antibiotic for three days, received 20 Ci of [14C]curdlan (location
    of radiolabel not given) as a single oral dose, and samples of blood,
    urine, faeces, and carbon dioxide were monitored for radiolabel for 96
    h. The antibiotic treatment suppressed the bacterial flora and in one
    case appeared to have totally eliminated the faecal bacteria. Although
    the extent to which curdlan is metabolized in the human
    gastrointestinal tract appears to reflect the action of the gut
    bacteria, even the subject with no apparent bacteria at the start of
    the study had very limited carbon dioxide production. Only one subject
    who had not received pretreatment had detectable radiolabel in blood
    after 24 h. The presence of significant amounts (1.6% of the
    administered dose) in the urine of this person and of a small but
    significant amount in the other man who had not received an antibiotic
    (0.014% of the administered dose) is a further indication of the

    bacterial metabolism of curdlan in the gastrointestinal tract. The
    difference between these two men can be explained by the predominance
    of enterococcus,  Escherichia coli, and  Bacillus species in the
    stool samples of the first and essentially only  E. coli in the
    second; their anaerobic bacteria also differed. The large amount of
    radiolabel in the expired carbon dioxide of the first man (31% of the
    administered dose) and the presence of a peak in expired radiolabel in
    the second also indicate the importance of the gut bacteria. The men
    pretreated with antibiotics expired the least radiolabel.

         The results suggest that people with a full complement of
    intestinal bacteria handle curdlan differently from individuals with
    few or no intestinal bacteria. The nature of the faecal bacteria may
    also play a role. The main pathway for excretion appears to be the
    faeces, except for a portion that appears to be fully metabolized to
    carbon dioxide. The wide discrepancy in the total recovery of the
    radiolabel in three of the four men is illustrated by the high
    percentages of unrecovered radiolabel (43, 83, and 58% respectively).
    The authors attributed this low total recovery to incomplete stool
    collection. The value of this study is therefore limited (Wazeter et
    al., 1975a).

    2.1.2  Mineral bioavailability in rats

         Groups of 20 four-week-old male Sprague-Dawley rats weighing
    67-86 g received 0, 1, 5, or 15% curdlan (purity assumed to be 100%)
    or 15% cellulose in their diet for eight weeks and were then killed.
    The study was performed according to GLP. All rats were observed twice
    a day for deaths, morbidity, and overt signs of toxicity. Detailed
    clinical examinations were performed weekly. Body weights were
    determined weekly and food consumption daily. The faeces of each
    animal were collected daily and weighed, pooled on a weekly basis, and
    stored frozen until analysis for calcium, magnesium, iron, zinc,
    copper, and magnesium. At termination of the study, blood samples and
    the liver were taken from each animal and frozen until analysis for
    the same minerals.

         No deaths or signs of overt toxicity were observed. The body
    weights of animals given 15% curdlan were statistically significantly
    decreased, and the food consumption in this group was slightly
    decreased, whereas that of animals given 15% cellulose was
    statistically significantly increased. The faecal weights of animals
    at 15% curdlan or cellulose were increased, the latter being most
    pronounced (no statistical analysis was used for faecal weights). The
    ratio of faecal mineral excretion to consumption in animals given 15%
    curdlan was statistically significantly increased for all minerals
    except zinc in week 1. Thereafter, a revised procedure was used to
    avoid contamination of faeces with food particles, and the ratios of
    faecal mineral excretion to consumption in all treated groups were
    similar or lower than control ratios for the remainder of the study.

    Statistically significantly increased ratios for all of the minerals
    were observed for the group given 15% cellulose throughout most of the
    study, except for zinc which was statistically significantly increased
    only during weeks 1 and 6. At the end of the study, the blood
    concentrations of the six minerals in curdlan-treated animals were
    comparable to those of controls, whereas in animals given 15%
    cellulose the iron concentration in blood was statistically
    significantly increased. The manganese concentration in the liver was
    statistically significantly increased in animals at 5 and 15% curdlan,
    while the copper concentration in the liver was statistically
    significantly decreased in those at 15% curdlan. In animals given 15%
    cellulose, the concentrations of all six minerals in the liver were
    comparable to those of controls (Goldenthal, 1994).

    2.1.3  Nutritional aspects

         Groups of five male Sprague-Dawley rats, 21 days old, received
    3.44 g of carbohydrate-free basal diet (52% casein, 29% cottonseed
    oil, 112% salt mixture, 4.4% cellulose powder, 0.9% D,L-methionine,
    0.3% choline, and 1.5% vitamin mixture) or the same diet supplemented
    with 1, 2, 3, or 6.56 g of starch or of curdlan daily for eight days.
    The addition of starch was reported by the authors to provide a
    nutritionally complete diet. After treatment, the rats were weighed
    and killed and the full and empty caecal weights were determined. The
    body-weight gain of the rats increased with increasing amounts of
    starch, but addition of curdlan did not cause dose-dependent growth,
    indicating that it was not used as a source of carbohydrate. Curdlan
    but not starch supplementation caused enlargement of the caecum and
    increased contents, and rats given curdlan excreted three to five
    times more faeces than controls, with a moisture content of 70%
    compared with 41-50% for rats fed carbohydrate-free diet (Matsuo &
    Suzuoki, 1974a).

         Groups of five male Sprague-Dawley rats, 26 days old, received
    3.44 g of carbohydrate-free diet supplemented with 0, 2, 2.5, 3, 3.5,
    4, or 6.56 g starch or carbohydrate-free diet plus 2 g starch
    supplemented with 0.5, 1, or 1.5 g of curdlan daily for seven days.
    The body weights of rats given diets supplemented with starch plus
    curdlan were the same as those of rats receiving only starch,
    indicating that curdlan was not used as a source of carbohydrate
    (Matsuo & Suzuoki, 1974a).

         Groups of two to three Sprague-Dawley rats were fasted for 20 h
    and then given 4-5 g/kg bw of glucose solution or curdlan in aqueous
    suspension. A control group was fed  ad libitum. The blood glucose
    and liver glycogen contents of the fasted animals were raised when
    they were given glucose but not in those given curdlan, indicating
    that curdlan was not used as a source of carbohydrate (Matsuo &
    Suzuoki, 1974a).

         Groups of male Sprague-Dawley rats, 26 days old, received 5 g of
    basal diet (44% casein, 13% sucrose, 3.3% cellulose powder, 6.6% salt
    mixture, 30% corn starch, 1.3% vitamin mixture, and 1.4% cottonseed
    oil) daily for five days. They were then weighed, divided into groups
    of five, and continued on the basal diet or given the basal diet
    supplemented with 1, 2, or 4 g of corn starch, sucrose, or curdlan
    daily for seven days, after which time they were weighed and killed
    and the caecum and its contents were removed for weighing. Body weight
    decreased gradually in the group fed 5 g/day basal diet, whereas that
    in the groups given basal diet supplemented with starch or sucrose
    increased with dose; however, no dose-dependent increase in body
    weight was seen with curdlan but remarkably enlarged caeca with
    increased contents were observed. The net increases in body weight,
    minus the weight of the caecal contents, were slightly lower in groups
    supplemented with starch than in those supplemented with sucrose, but
    the groups supplemented with curdlan showed a dose-dependent decrease
    in net body weight. The authors concluded that curdlan is
    nutritionally inert in rats (Matsuo & Suzuoki, 1974b).

    2.2  Toxicological studies

    2.2.1  Acute toxicity

         Studies of the toxicity of single doses of curdlan are summarized
    in Table 5. No abnormalities or deaths were observed in mice or rats
    after oral administration. Transient diarrhoea occurred after
    intraperitoneal injection, and macroscopic examination showed an
    agar-like residuum of unabsorbed test material in the abdominal
    cavity. Massive amounts of the compound were detected in rats, which
    received larger quantities. Rats showed adhesion of curdlan to liver
    and spleen, which were covered with a white, fibrin-like coating, but
    the parenchymal portion of the organs showed no lesions (Aomori &
    Tanida, 1968).

    2.2.2  Short-term studies of toxicity

         Mice

         In an eight-week study of possible side-effects (not performed
    according to GLP or a current guideline), groups of 10 male (body
    weight, 24-34 g) and 10 female (body weight, 19-29 g) Charles River
    CD-1 mice were fed curdlan in their diet at concentrations of 0, 1, 5,
    10, 20, or 30%, equal to 0, 1.4, 7.1, 14, 29, and 43 g/kg bw,
    respectively. Feed and water were available  ad libitum. The animals
    were observed daily for changes in behaviour, appearance, and
    excretion. Individual body weights and the food consumption of animals
    of each sex in each group were measured weekly. All animals were
    examined macroscopically.


        Table 5. LD50 values for curdlan
                                                                                                

    Species      Sex        Route                LD50             Remarks
                                                 (mg/kg bw)
                                                                                                

    Mouse        M & F      Oral                 > 10 000         10% aqueous suspension
    Rat          M          Oral                 > 10 000         10% aqueous suspension
    Mouse        M          Intraperitoneal        2 750           5% suspension in saline
    Mouse        F          Intraperitoneal        2 500           5% suspension in saline
    Rat          M          Intraperitoneal        2 750           5% suspension in saline
                                                                                                

    From Aomori & Tanida (1968). As the origin of the test compound was not given its purity
    is unknown. The tests were not performed according to GLP or a current guideline. Only a
    summary report was submitted, and the numbers of treated groups per test were not given.
    Six animals per group were used (no further details) and observed for seven days.
    

         One female mouse at 30% curdlan died of undescribed causes during
    the study, and one to two mice at 1, 5, and 10% and half of those at
    20 and 30% had large stools. Soft stools were seen in a few mice at 20
    and 30% curdlan. The body-weight gain of male mice at 30% was
    decreased by 10% in comparison with male controls. The authors
    reported that the total food consumption (basal diet plus curdlan)
    generally increased with dose, but the Committee concluded that the
    total food consumption of males at all doses was increased by > 10%
    over that of controls, with no dose-response relationship, and over
    that of females at 20 and 30% on the basis of mg/kg bw feed
    consumption. The actual food consumption (basal diet only) did not
    differ from that of the control group. At the end of treatment, the
    weights of the full caeca of males at 10, 20, and 30% curdlan and of
    females at 20 and 30% and the weights of the empty caeca of males and
    females at 30% and of males at 10% were statistically significantly
    increased. No abnormalities were observed macroscopically in any of
    the mice, including the female at 30% which died during the study.
    Microscopy was not performed. The NOEL was 5% curdlan in the diet on
    the basis of the increase in full caecal weights and large stools at
    higher doses (Wazeter et al., 1973).

         Rats

         Five groups of five male Sprague-Dawley Ta rats aged five weeks
    and weighing 110-158 g received a basal diet to which 3, 10, or 30%
    curdlan (purity not stated) had been added (providing approximately 0,
    2.5, 8.5, and 30 g/kg bw) or a basal diet to which 30% agar powder had
    been added (approximately 33 g/kg bw) for one month. The study was not
    carried out according to GLP or current guidelines. Data on individual
    animals were not submitted.

         Animals at 10 and 30% curdlan and 30% agar excreted faeces that
    were two to five times larger than normal, had a slight, peculiar
    odour, and were brown and grey-white and poorly adhesive. The groups
    given 10 and 30% curdlan had diarrhoea after day 17, whereas those
    given 30% agar did not have soft or diarrhoeal stools. The faeces of
    rats given 3% curdlan were slightly greyish but of normal form. No
    significant differences in body-weight gain were seen in comparison
    with controls. The total food consumption (basal diet plus curdlan or
    agar) of animals at 3, 10, and 30% curdlan or 30% agar was 7.8, 9.8,
    26, and 40% higher than that of controls. On the basis of basal diet
    only, the food consumption differed from that of the control group by
    4.6, -1.1, -12, and -1.6%, respectively. Food efficiency in rats given
    3, 10, and 30% curdlan or 30% agar was 98, 92, 78, and 69%,
    respectively, that of the control group; when calculated on the basis
    of basal diet, the values were 100, 100, 112, and 98%, respectively.
    The average water consumption of these groups was 100, 96, 140, and
    130% that of the control group. The differential leukocyte count
    showed dose-related increases in neutrophils in groups treated with
    curdlan, which were significant at 10%. Occult blood was present in
    the urine of four animals given 10% curdlan, two at 30% curdlan, and
    one at 30% agar. No treatment-related changes were found in other

    haematological (reticulocyte count, haemoglobin, haematocrit, mean
    corpuscular volume, mean corpuscular haemoglobin, mean corpuscular
    haemoglobin content, or leukocyte count), biochemical (serum alkaline
    phosphatase, alanine aminotransferase, or aspartate aminotransferase
    activity, glucose, protein, urea, or cholesterol concentration, or
    Bromsulphalein retention), or urinary (colour, pH, proteins, glucose,
    ketone bodies, or urobilinogen) parameters. Statistically significant,
    dose-related decreases in relative kidney weights were observed in all
    treated groups, including that given 30% agar. Statistically
    significant decreases in relative liver weight were seen in rats given
    30% curdlan or agar. The relative adrenal weights were statistically
    significantly decreased with 30% agar, and the relative pituitary
    weights were statistically significantly increase with 30% curdlan.
    All groups given curdlan had dose-related, statistically significant
    increases in the relative weights of the empty caecum, and macroscopic
    examination revealed enlarged caeca in all treated groups including
    those given agar. Histopathology showed no treatment-related changes.
    The NOEL was 3% curdlan in the diet on the basis of large faeces and
    diarrhoea at higher doses (Yokotani et al., 1969).

         Groups of 10 male and 10 female Sprague-Dawley rats aged five
    weeks (the males weighing 138-160 g and the females 116-130 g)
    received 0, 5, 10, or 20% curdlan (purity not given; sterilized by dry
    heating for 20 min at 120 C) in their diet for three months, equal to
    0, 4.4, 9, and 19 g/kg bw for males and 0, 5.5, 12, and 24 g/kg bw for
    females. The study was not conducted according to GLP or current
    guidelines. The general condition of the animals was observed daily,
    and body weight and food consumption were recorded twice a week. Food
    efficiency was calculated. At the end of the treatment period,
    haemato-logical (haemoglobin, haematocrit, erythrocyte and leukocyte
    counts, differential leukocyte count, and platelets) and biochemical
    (serum alkaline phosphatase, alanine aminotransferase, aspartate
    aminotransferase, and lactic dehydroge-nase activity, Bromsulphalein
    retention, cholesterol, glucose, blood urea nitrogen, creatinine,
    total bilirubin, total proteins, albumin, globulin, the albumin:
    globulin ratio, and calcium) parameters were measured in all animals.
    After 6 and 11.5 weeks, urine from all animals was analysed for
    colour, pH, proteins, glucose, ketone bodies, occult blood, and
    sedimentation rate. The absolute and relative weights of 14 organs
    were determined. Macroscopy and microscopy were performed, but the
    numbers of animals and groups examined were not reported.

         Animals at doses > 10% in the diet had large, slightly
    discoloured, soft, deformed faeces. Body-weight gain decreased with
    increasing dose in males but statistically significantly only at 20%
    in the diet, while the total food intake (basal diet plus curdlan) of
    males at all doses and of females at 10 and 20% increased. As a
    consequence, food efficiency was decreased in males at all doses and
    in females at the two higher doses. A dose-dependent decrease in
    platelet count was observed in males which was statistically

    significant at 10 and 20% curdlan, and statistically significant
    decreases in total protein and globulin concentrations were observed
    in male and female animals at 10 and 20% in the diet, leading to
    increased albumin:globulin ratios which were statistically significant
    only at 20%. The calcium and cholesterol concentrations in males at
    20% in the diet were statistically significantly decreased and serum
    alkaline phosphatase activity was significantly increased in males at
    10 and 20%. Decreased urinary volume was reported at 20% (data not
    given in the table), and a slight increase in urinary pH in treated
    males and females was seen after both 6 and 11.5 weeks, with no
    dose-response relationship. The absolute carcass weight was
    statistically significant increased in males at 10 and 20% curdlan in
    the diet, but decreases were observed in the absolute weight of the
    liver in males at 20% (statistically significant), the absolute
    weights of the kidney in males at 10 and 20% (statistically
    significant), the absolute and relative weights of the ovary at 20%,
    and the absolute and relative weights of the pituitary in females at
    all doses (absolute weight statistically significant at 20% in diet;
    relative weight statistically significant at 5 and 20% in the diet).
    The relative uterine weight increased with dose (statistically
    significant at 20%), and the relative weights of the adrenal and
    submaxillary glands were increased in males at 10 and 20%
    (statistically significant). The absolute and relative weights of the
    full and empty caeca increased with dose in males and females at all
    doses, but the increase was more pronounced in males and was
    statistically significant at 10 and 20%. The wall of the caecum did
    not become thinner. Macroscopic examination showed a dose-dependent
    decrease in deposition of adipose tissue in the abdominal cavity in
    males at all doses and in females at 20% in the diet. The large
    intestine (full) of males and females at 20% was two to three times
    larger than that of controls, and moderate enlargement was seen in
    some male animals at 5 and 10% curdlan and in females at 10%. No
    histopathological effects of treatment were seen. The NOEL was 5% in
    the diet mainly on the basis of changes in faecal size, diarrhoea, the
    weights of the full and empty caeca, and enlarged large intestines
    (Nakaguchi et al., 1972) 

         Groups of 25 male and 25 female Charles River CD rats were paired
    by litter to receive 15% curdlan in the diet or basal diet only at an
    amount equal to that which the paired treated rat had consumed the
    previous day. The behaviour and appearance of the rats were not
    affected, but males showed a statistically significantly increase in
    growth throughout treatment and females showed statistically
    significantly increased growth during the first two weeks only. No
    other parameters were examined (Wazeter et al., 1976a).

         Dogs

         In a study reported as a summary and that was not performed
    according to GLP or current guidelines, groups of two beagles (assumed
    to consist of one male and one female per group), aged eight to nine
    months, were fed coagulated or uncoagulated curdlan (purity not

    achieve a final concentration of 8 or 16%. Growth was normal although
    two bitches receiving uncoagulated curdlan ate only one-third to
    three-quarters of their daily diet. Stool consistency and urinary
    excretion remained normal, and the haematological parameters were
    normal except for increased eosinophil counts in animals receiving
    uncoagulated curdlan. No biochemical abnormalities or changes in organ
    weights were observed. Histopathological examination showed rare
    pigmented macrophages and slight to moderate glycogen depletion in the
    liver and inflammatory changes in the small intestine (dose not
    stated). There was no NOEL (Woodard & Imming, 1971).

         In another study that was not performed acccording to GLP or
    current guidelines, groups of four male and four female beagles
    weighing 7.8-12 kg (males) and 7.1-11 kg (females) received 0, 1, 5,
    or 15% curdlan or 40% gelled curdlan (prepared by heating a 10%
    suspension of curdlan in distilled water to 90 C for 15 min and then
    cooling it to room temperature) in the diet for one year. Appearance
    and behaviour were monitored daily, and body weight and food
    consumption were recorded weekly. Physical examinations were performed
    before treatment, at monthly intervals during the first six months,
    and at 9 and 12 months. Ophthalmoscopy was carried out in all animals
    before treatment and after 3, 6, and 12 months. Haematological
    (haemoglobin, haematocrit, reticulocyte, leukocyte, and differential
    leukocyte counts, reticulocyte sedimentation rate, and platelet
    count), serum biochemical (serum alkaline phosphatase, alanine and
    aspartate aminotransferases, blood urea nitrogen, glucose, total
    proteins, albumin, bilirubin, cholesterol, creatinine, sodium,
    potassium, chlorine, and calcium) and urinary (appearance, colour,
    volume, pH, specific gravity, albumin, glucose, bilirubin, occult
    blood, and sedimentation rate) parameters were measured in all animals
    before treatment and after 3, 6, and 12 months. At the end of the
    study, the absolute and relative weights of the liver, kidneys, heart,
    spleen, brain, caecum (full and empty), adrenals, thyroid/parathyroid,
    testes/ovaries, and pituitary were determined. Macroscopy and
    microscopy were performed on 28 tissues from all animals.

         One male at 15% curdlan died at week 37 of the study, probably
    because of intercurrent infection; excessive salivation and blood in
    the refuse pan were seen on the day before death. Soft stools were
    seen more frequently in animals at 15% curdlan and at 40% gelled
    curdlan during the first six months of the study than in other groups,
    and thereafter only at 15% curdlan. During the first two months,
    mucoid stools and/or blood-tinged mucoid stools were observed rarely
    in animals at 1%, occasionally in those at 5%, and frequently in those
    at 15% curdlan and 40% gelled curdlan. Thereafter, the incidences
    decreased at 1 and 5% curdlan and at 40% gelled curdlan but not at 15%
    curdlan. Males at 1, 5, and 15% curdlan and 40% gelled curdlan had
    slightly (< 10%), nonsignificantly decreased growth with no
    dose-response relationship. No treatment-related changes were found in
    physical appearance, food consump-tion, or haematological,
    biochemical, or urinary parameters. The group mean absolute and
    relative weights of the full and empty caeca were statistically

    significantly increased at 15% curdlan, but no other significant
    changes in absolute or relative organ weights were seen (no group mean
    weights given). Macroscopic examination showed petechiation and
    ecchymosis of the small intestinal mucosa in one animal at 1% curdlan,
    two at 15% curdlan, and five at 40% gelled curdlan; the male at 1%
    curdlan and two of the five at 40% gelled curdlan also shallow mucosal
    erosion. Mucosal ecchymosis of the small intestine was also observed
    in one control dog. Microscopy revealed small areas of recent mucosal
    bleeding in the small intestine of one animal at 1% curdlan, two at
    15%, and two at 40% gelled curdlan. The NOEL was 5% curdlan on the
    basis of soft stools or diarrhoea, mucoid stools and/or blood-tinged
    mucoid stools, and increased full and empty caecal weights at 15%
    curdlan (Wazeter et al., 1975b).

    2.2.3  Long-term studies of toxicity and carcinogenicity

         Mice

         In a study that was not performed according to GLP or current
    guidelines, groups of 100 male and 100 female Charles River CD-1 mice
    weighing 17-35 g (males) and 18-29 g (females) received diets
    containing 0, 1, 5, or 15% curdlan or 40% gelled curdlan (prepared as
    described above) daily until the individual groups had been reduced by
    approximately 80% by sacrifices and deaths during treatment. Nine
    males and five females at 1% curdlan and five males and five females
    from all other groups were killed at 12 months, five males and five
    females were killed after 18 months, and the remainder at weeks 99-114
    when the survival rate in the groups was approximately 20%. Changes in
    behaviour and appearance were checked daily, and body weights and food
    consumption were recorded weekly. All mice were examined
    macroscopically, and approximately 20 tissues from five males and five
    females per group killed after 12 and 18 months, from all mice that
    died or were killed in extremis during the study, and from all
    remaining mice were examined microscopically. All gross lesions
    suspected of being neoplastic were also examined microscopically.

         No treatment related differences in survival or body-weight gain
    were found between treated and control groups. The food consumption
    (grams per mouse per day) of males at 40% gelled curdlan was increased
    by about 19%, whereas that of females at 15% curdlan was decreased by
    about 13% (no statistical analysis was performed). No
    treatment-related macroscopic or microscopic abnormalities and no
    treatment-related changes in tumour incidence were observed. The NOEL
    was 5% curdlan in the diet on the basis of decreased food consumption
    at 15% curdlan (Wazeter et al., 1976b).

         Rats

         In a study not performed according to GLP or current guidelines,
    groups of 60 male and 60 female Charles River CD rats weighing 64-110
    g (males) and 71-120 g (females) received 0, 1, 5, or 15% curdlan or
    40% gelled curdlan (prepared as described above) in their diet for two

    years. Their appearance and behaviour were checked daily, and body
    weight and food consumption were recorded weekly. Ophthalmoscopy was
    carried out in all rats before treatment and at 3, 6, 12, 18, and 24
    months. After those same times, haematological (haemoglobin,
    haematocrit, erythrocyte, leukocyte, differential and platelet
    counts), serum biochemical (serum alkaline phosphatase, alanine and
    aspartate aminotransferases, blood urea nitrogen, glucose, total
    proteins, albumin, albumin:globulin ratio, bilirubin, cholesterol,
    creatinine, sodium, potassium, and chlorine), and urinary (appearance,
    colour, volume, pH, specific gravity, albumin, glucose, bilirubin,
    occult blood, and sedimentation rate) parameters were measured in five
    male and five female rats per group. Five male and five female rats
    per group were killed after 12 months; three male and three female
    rats at 5 and 15% curdlan and five male and five female rats from all
    other groups were killed after 18 months; and the remaining animals
    were killed after two years. The absolute and relative weights of the
    liver, kidneys, heart, spleen, brain, caecum (full and empty),
    adrenals, thyroid/parathyroid, testes/ovaries, and pituitary were
    determined in all rats, and the animals were examined macroscopically.
    Approximately 25 tissues from all animals killed after 12 and 18
    months and from 10 males and 10 females per group killed after 24
    months were examined microscopically. 

         No changes in mortality, behaviour, appearance, or ophthalmic
    parameters were observed. At 15% curdlan, males gained about 11% less
    and females about 10% less weight than controls, although the
    decreases were not significant, and males consumed about 22% less and
    females about 16% less basal diet (no statistical analysis performed).
    No treatment-related changes in haematological, biochemical or urinary
    parameters were observed. The absolute and relative weights of the
    full and empty caeca were increased in animals at 15% curdlan at all
    times and were statistically significant after 18 and 24 months. No
    significant treatment-related abnormalities were seen histologically.
    The NOEL was 5% curdlan in the diet on the basis of decreased
    body-weight gain and food consumption and increased full and empty
    caecal weights at 15% in the diet (Wazeter et al., 1976c).

         In a study not performed according to GLP or current guidelines,
    groups of 60 male and 60 female Charles River CD rats of the F1a
    generation of a multigeneration study of reproductive toxicity,
    weighing 38-110 g (males) and 33-10) g (females), received 0, 1, 5, or
    15% curdlan or 40% gelled curdlan (prepared as described above) in
    their diet daily until weeks 124-127 when the survival rate in each
    group was 20%. Appearance and behaviour were checked daily, and body
    weights and food consumption were recorded weekly. After 3, 6, 12, 18,
    and 24 months and before termination, ophthalmoscopy was carried out
    on all rats, and haematological (haemoglobin, haematocrit,
    erythrocyte, leukocyte, and differential leukocyte and platelet
    counts), serum biochemical (serum alkaline phosphatase, alanine and
    aspartate aminotrans-ferases, blood urea nitrogen, glucose, total

    proteins, albumin, albumin:globulin ratio, bilirubin, cholesterol,
    creatinine, sodium, potassium, and chlorine), and urinary (appearance,
    colour, volume, pH, specific gravity, albumin, glucose, bilirubin,
    occult blood, and sedimentation rate) parameters were measured in 10
    male and 10 female rats in each group. Five males and five females in
    each group were killed after 12 and 18 months, and the remaining rats
    were killed in weeks 124-127 when the survival rate in each group was
    approximately 20%. The absolute and relative weights of the liver,
    kidneys, heart, spleen, brain, caecum (full and empty), adrenals,
    thyroid/parathyroid, testes/ovaries, and pituitary were determined in
    all rats, which were examined macroscopically. About 25 tissues from
    all rats were examined microscopically. 

         No changes in mortality, behaviour, appearance, or ophthalmic
    end-points were observed. The body weights of animals at 15% curdlan
    were statistically significantly decreased; whereas those of males at
    5% curdlan were statistically significantly decreased until week 65 of
    treatment and then approached those of controls. The body weights of
    males and occasionally of females given 40% gelled curdlan showed
    statistically significant decreases until week 78 of treatment and
    then became similar to those of controls. The consumption of basal
    diet was decreased by 8.3% in males and 4.4% in females at 15% curdlan
    and increased by 5% in males and 13% in females at 40% gelled curdlan.
    Food efficiency was slightly decreased for females at 15% curdlan and
    for male and female rats at 40% gelled curdlan. No treatment-related
    changes in haematological or urinary parameters were seen.
    Statistically significantly increased aspartate aminotransferase
    activity was seen in males and females at 15% curdlan after 12 and 29
    months and in animals given 40% gelled curdlan group at some
    measurement times. Statistically significant increases in serum
    alkaline phosphatase activity were observed in males and females at
    15% curdlan or 40% gelled curdlan up to 12 months. The absolute and
    relative weights of the full caecum were statistically significantly
    increased in all treated groups at terminal sacrifice, and the
    absolute and relative weights of the empty caecum were increased in
    males at 5% curdlan, in females at 15% curdlan, and in females at 40%
    gelled curdlan at this time. At the other times, increased full and,
    less frequently, increased empty caecal weights were seen in one or
    more groups given curdlan. Macroscopic and microscopic examination
    revealed a statistically significant increased incidence of uterine
    polyps in females at 15% curdlan that were killed at the end of or
    during the study, with incidences of 0/450 in controls, 3/50 with 1%
    curdlan, 4/51 with 5% curdlan, 7/50 with 15% curdlan, and 2/50 with
    40% gelled curdlan and thus possibly related to treatment. The NOEL
    was 1% curdlan on the basis of increased empty caecal weights at 5 and
    15% in the diet and decreased growth and food consumption and an
    increased incidence of uterine polyps at 15% curdlan (Wazeter et al.,
    1976d).

    2.2.4  Genotoxicity

         The results of studies on the genotoxicity of curdlan are
    summarized in Table 6.

    2.2.5  Reproductive toxicity

         Rats

         In a three-generation study of reproductive toxicity and
    teratogenicity that was not performed according to GLP or current
    guidelines (mating periods lasted 15 days instead of three weeks as
    prescribed in OECD guideline 416; the parameters examined resembled
    those in OECD guideline 416 [for reproductive toxicity] and 414 [for
    teratogenicity]), groups of 20 male and 40 female Charles river CD
    rats, weighing 64-120 g (males) and 59-110 g (females), received 0, 1,
    5, or 15% curdlan or 40% gelled curdlan (prepared as described above)
    in their diet until they were 100 days of age. The F0 parents were
    mated twice. After the first litter had been weaned, the number of F0
    parents was reduced to 10 males and 20 females per group. The F1
    parents were mated three times and the F2 parents twice. The pups
    from the first mating in each generation were examined for external
    abnormalities and killed and discarded after weaning (21 days of age),
    although some animals from each group of the F1a generation were
    selected for use in the long-term study described above. At the age of
    four group from the second matings of the F0 and F1 generations (the
    F1b and F2b litters) were selected to produce the next generation.
    After the third mating of F1 parents, which producing the F2c
    litters, 50% of the F1 females were killed on day 13 of gestation,
    laparatomies were performed, and their uteri and ovaries were
    examined. The remaining 50% of F1 females were killed on day 20 of
    gestation and their fetuses were removed. The parental rats were
    observed daily for changes in behaviour and appearance, and individual
    body weights and food consumption were recorded weekly. During the
    reproduction phase, fertility, litter size, numbers of male and female
    pups, the viability of the newborns, the growth of the pups, and
    survival of pups to weaning were observed. After weaning of the F3b
    litters, the absolute and relative weights of the liver, kidneys,
    heart, spleen, full and empty caeca, and testes/ovaries were
    determined in five male and five female F3b pups per group and five
    male and five female F2 parental rats per group, and the animals were
    examined macroscopically. About 15 tissues from five male and five
    female F3b pups and five male and five female F2 parental rats from
    the control, 15% curdlan, and 40% gelled curdlan groups were examined
    microscopically. In the evaluation of teratogenicity, the numbers of
    corpora lutea, implantations, and resorptions and the number,
    distribution, and location of viable and dead fetuses were counted in
    F2c dams killed on day 13 or 20 of gestation. All fetuses from F2c
    dams killed on day 20 of gestation were examined for external
    abnormalities, sexed, and weighed. Two-thirds of the fetuses in each
    litter were examined for visceral abnormalities and one-third for
    skeletal abnormalities.


        Table 6. Results of studies of the genotoxicity of curdlan
                                                                                                                     

    End-point      Test object       Concentration         Purity           Result                         Reference
                                                                                                                     

    Reverse        S. typhimurium    15-5000 g/plate      Not reported     Negative in presence and       Dillon 
    mutationa      TA98, TA100,      as suspension in      but assumed      absence of S9; slight          (1994)
                   TA1537, TA1538    sterile ultra-pure    to be 86%        toxicity only in absence 
                   TA1535,           water                                  of S9 at 5000 g/plate in 
                                                                            all strains except TA1538;
                                                                            no precipitation

    Gene           tk locus in       12.5-5000 g/ml       86%              Negative in presence and       Riach &
    mutationb      mouse lymphoma    as suspension in                       absence of S9; toxicity at     Willington
                   L518Y cells       tissue culture                         2500 and 5000 g/ml            (1994)
                                     medium                                 (relative growth, 69 and
                                                                            31%, respectively, in
                                                                            absence of S9 and 66 and
                                                                            33%, respectively, in
                                                                            presence of S9)

    Chromosomal    Chinese           625 (1250)-5000       86%              Negative in presence and       Leddy
    aberrationc    hamster           g/ml as suspension                    absence of S9; no toxicity;    (1994)
                   ovary cells       in tissue culture                      treatment time, 22 h
                                     medium                                 in absence of S9, 6 h in
                                                                            presence of S9; harvesting
                                                                            time, 24 and 48 h after
                                                                            start of treatment

    Micronucleus   Male and          0, 500, 1000, and     86%              No micronuclei and no          Holmstrom
    formationd     female CD-1       2000e mg/kg bw                         overt mortality or             & Innes
                   mice              dissolved in water                     adverse reactions; ratio       (1994)
                                     by gavage at 24-h                      of polycromatic to normal
                                     intervals;                             erythrocytes unchanged
                                     bone-marrow samples
                                     taken 24 h after
                                     last dose
                                                                                                                     

    a  Performed according to GLP and OECD Guideline No. 471
    b  Performed according to GLP and test protocol identical to OECD Guideline No. 476
    c  Performed according to GLP and OECD Guideline No. 473
    d  Performed according to GLP and OECD Guideline No. 474
    e  Maximum dose attainable
    

         In the evaluation of reproductive toxicity, male parental rats of
    the F0 and F1 generations at 15% curdlan and female parental rats of
    the F0 generation given 40% gelled curdlan showed slightly decreased
    mean growth. The mean consumption of basal diet (in grams per rat per
    day) by male and female parental rats of the F0 and F1 generations
    at 15% curdlan was slightly decreased. F2 parental females given 15%
    curdlan or 40% gelled curdlan showed increased absolute and relative
    weights of the full caecum, which were reported by the authors to be
    statistically significant ( p < 0.05), but the results of the
    statistical analyses were not given in the table. The Committee noted
    that the absolute and relative weights of the empty caecum in females
    were also increased at these doses. No macroscopic or microscopic
    treatment-related changes were observed in the F2 parents examined.
    Although no treatment-related changes in fertility, gestation, or
    viability indices were observed, the weights of the pups in nearly all
    litters in all generations at 15% curdlan were significantly decreased
    during lactation, in the F1a and F1b litters at 14 and 21 days of
    age; in the F2a and F3a litters at 4, 7, 10, 14, 17, and 21 days of
    age; and in F3b litters at 21 days of age. Variations in the weights
    of pups in the three generations at 1 and 5% curdlan and 40% gelled
    curdlan were observed occasionally during lactation but were not
    consistent and were considered not to be biologically relevant. The
    authors suggested that the reductions in pup weight reflected the
    inexperience of primiparous dams and were due to the fact that from
    the age of 7-10 days the young started to consume some of their
    mothers' feed. They stated that the weights of the pups at various
    times  post partum were within the normal range of values for rats of
    this strain, but data on historical controls were not submitted. F3b
    weanlings at 5 and 15% curdlan and at 40% gelled curdlan had decreased
    full caecal weights, and those at 1% curdlan and 40% gelled curdlan
    had increased empty caecal weights. The authors reported that these
    changes were statistically significant ( p < 0.05), but the results
    of the statistical analyses were not shown. No macroscopic or
    microscopic treatment-related changes were observed in the F3b pups
    at 15% curdlan or 40% gelled curdlan.

         In the evaluation of teratogenicity, the numbers of gravid
    animals, corpora lutea, implantation sites, and viable and dead
    embryos on days 13 and 20 of gestation were normal. The fetuses of
    dams killed on day 20 of gestation had a normal sex ratio and showed
    no increased incidence of visceral or skeletal abnormalities. The mean
    live fetal weights in all treated groups were statistically
    significantly increased in comparison with controls, but no
    dose-response relationship was seen.

         The NOEL for parental animals was 5% curdlan on the basis of
    decreased growth and increased full and empty caecal weights at 15%.
    The NOEL for embryotoxicity was also 5% on the basis of decreased
    weight gain during lactation at 15% (Wazeter et al., 1976e).

         A single-generation study with two litters was perfomed to
    determine whether the offspring of treated females would grow normally
    if they were nursed by untreated females and whether the offspring of
    untreated females would grow normally if they were nursed by treated
    dams. In this study, which did not conform to GLP or any current
    guideline, a control group of 40 male and 80 female Charles River CD
    rats and three test groups of 20 male and 40 female rats received 0,
    5, or 15% curdlan or 40% gelled curdlan (prepared as described above)
    in their diet continuously throughout the study. When they were 100
    days old, a period of 15 days was allowed for mating to produce an
    F1a generation. The F1a litters were reduced to 10 pups per litter on
    day 0. Then, 20 dams in each group nursed their own offspring through
    lactation while the remaining 20 dams nursed the litters of 60 control
    dams, which in turn nursed the pups of the litters of the other 20
    dams in each of the three treated groups. The exchange of pups between
    control and treated groups took place within 24 h  post partum. After
    weaning, the F1a offspring were killed and the parental females
    rested for 10 days and then bred a second time to produce F1b
    litters. After weaning of the F1b litters, all parental animals and
    offspring were killed. The individual body weights and food
    consumption of the parents were recorded weekly, and the fertility
    index, length of gestation, litter size and viability, numbers of live
    and dead births, and the lactation index were determined. Pups were
    examined for gross abnormalities at birth and throughout lactation,
    and the individual body weights and survival indices of the pups were
    recorded at 0, 4, 7, 10, 14, 17, and 21 days  post partum.

         No changes in mortality, behaviour, or appearance were observed.
    Male parents at 15% curdlan grew slightly less (< 10%) than control
    males, and the food consumption of male and female parents at this
    dose was lower, whereas dams at 40% gelled curdlan had increased feed
    consumption. No changes in fertility or lactation indices were
    observed when the F1a and F1b litters were born, and the pups showed
    no gross abnormalities. The survival of F1a pups at 5 and 40% curdlan
    was statistically significantly lower than in the control group, but
    that of F1b pups was normal. The F1a litters of all treated dams that
    nursed their own young showed statistically significant decreases in
    weight gain, the greatest decrease being observed with 15% curdlan,
    but F1b litters showed reduced weight gain only at 15% curdlan. When
    pups born to treated dams were transferred to control dams,
    statistically significant decreases in weight gain were still observed
    during lactation, but the effect was reduced. The decreases in F1a
    litters appeared later and for a shorter time (days 17 and 21 with 5%
    curdlan, days 10 and 21 with 15% curdlan, and day 21 with 40% gelled
    curdlan), whereas F1b litters at 5 and 15% curdlan showed
    statistically significant decreases in weight gain on days 4 and 14 of
    lactation, respectively. When control pups were nursed by dams given
    15% or 40% curdlan, statistically significant decreases in weight gain
    were seen in F1a control pups on days 7-21 and day 10 of lactation,
    respectively. F1b control pups nursed by dams given 5 or 15% curdlan

    showed statistically significant decreases in growth on days 4 and
    14-17 of lactation, respectively. The authors concluded that the young
    from treated dams gained weight at a normal or nearly normal rate when
    nursed by control dams but had considerably reduced weights when
    nursed by treated dams. They ascribed the effect on pup weight to
    consumption by the pups of their mothers' diets which contained
    non-nutritive material. There was no NOEL (Wazeter et al., 1976f).

         In a similar study, cellulose was fed in the diet at
    concentrations of 0, 5, or 15% from 80 days before mating throughout
    gestation, parturition, and lactation until weaning of two litters. No
    effect on the appearance, behaviour, or survival of the parents or
    pups was observed, and the body weights and feed consumption of the
    parents showed no significant abnormality. Pups in the F1a and F1b
    litters at 15% cellulose in the diet showed decreased weight gain
    during lactation, which was ascribed by the authors to consumption by
    the pups of their mothers' diet (Wazeter et al., 1976g).

         A further single-generation study was performed to determine
    whether withdrawal of curdlan from the diet of dams during lactation
    would still result in reduced weight gain in the pups. In this study,
    which was not performed according to GLP or a current guideline, four
    groups of 20 male and 40 female weanling Charles River CD rats
    received 0, 5, or 15% curdlan or 40% gelled curdlan (prepared as
    described above) in their diet. Males received the diet continuously
    throughout the study after a seven-day conditioning period. Females
    received the diet after a seven-day conditioning period; when they
    were 100 days old they were allowed a period of 15 days for mating to
    produce an F1a generation, and the test diets were fed until day 20
    of gestation and removed during the lactation period. The F1a litters
    were reduced to 10 pups per litter on day 4 post partum, and all pups
    were killed after weaning. The dams then received the test diets again
    during a 10-day rest period and during the subsequent 15-day mating
    period to produce the F1b generation, until day 20 of gestation. The
    F1b litters were reduced to 10 per litter at day 4  post partum, and
    all animals were killed after the weaning period. The appearance,
    behaviour, individual body weights, and food consumption of the
    parents were recorded weekly, and the fertility index, pup survival,
    litter size, viability, numbers of live and dead births, and the
    lactation index were determined. Pups were examined for gross
    abnormalities at birth and during lactation, and the individual body
    weights of the pups were determined on days 0, 4, 7, 10, 14, 17,
    and 21.

         Females at 40% gelled curdlan consistently consumed more food
    than other groups, probably as compensation for the 40% of
    non-nutritional material in their diet. The reproductive performance
    of the parents was not affected, and no adverse effects on pups were
    observed. The weight gain of the pups during lactation was not
    affected even though the authors reported that the pups could have
    eaten their mothers' diet (basal diet) from about day 10  post partum.

    The authors considered that the lack of effect on the weight gain of
    the pups supported their suggestion that the reduced weight gain of
    pups in the previous studies had been due to their consumption of
    curdlan in their mothers' diet. The NOEL for both parental toxicity
    and embryotoxicity was 15% curdlan and 40% gelled curdlan (Wazeter et
    al., 1976h).

         Rabbits

         In a study of teratogenicity, which was not performed according
    to GLP or a current guideline, groups of 15-20 pregnant Dutch Belted
    rabbits received curdlan at 0, 1, 2, or 5 g/kg bw per day orally in
    gelatine capsules delivered from a syringe daily during days 6-18 of
    gestation An additional group received 20 g/kg bw per day gelled
    curdlan (prepared by heating a 10% suspension of curdlan in distilled
    water to 90 C for 15 min, cooling it to room temperature, maintaining
    it in a refrigerator until use, and then fragmenting it in a mortar)
    by the same route. The doses of 5 g/kg bw per day curdlan and 20 g/kg
    bw per day gelled curdlan were given in two divided doses. The control
    group received two empty capsules per day. The rabbits were killed on
    day 28 of gestation and their fetuses were removed. The weights of the
    dams were determined on days 6, 12, 18, and 28 of gestation, and the
    uteri of all dams were examined, and the numbers and locations of live
    and dead fetuses, empty implantation sites, early and late
    resorptions, and corpora lutea were recorded. The fetuses were weighed
    and examined for external, visceral, and skeletal abnormalities.

         None of the controls died, but three at 1 g/kg bw per day, one at
    2 g/kg bw per day, and three at 5 g/kg bw per day curdlan and 16 at 20
    g/kg bw per day gelled curdlan died, the last effect being due to
    mechanical occlusion of the pharynx by the gel. Eleven resorptions
    were seen with 5 g/kg bw per day curdlan compared with four in
    controls, six at 1 g/kg bw per day curdlan, five at 2 g/kg bw per day
    curdlan, and two at 20 g/kg bw per day gelled curdlan; however, the
    numbers of females that had resorptions were similar in all groups (4,
    4, 3, 2, and 2 females at 0, 1, 2, 5, and 20 g/kg bw per day (gelled)
    curdlan, respectively). No teratogenic effects were observed. The NOEL
    for both maternal and embryotoxicity was 5 g/kg bw per day, the
    highest dose tested. Owing to the high mortality seen with 20 g/kg bw
    per day gelled curdlan, no conclusion can be drawn about its maternal
    toxicity, embryotoxicity, or teratogenicity (Wazeter et al., 1974).

    2.2.6  Special studies

    2.2.6.1  Immunogenicity

         Mice

         Groups of four CH3 mice were given a single intravenous injection
    of 10 g curdlan, 10 g dextran 2000, or 0.1 mg alum-precipitated
    bovine serum albumin. Although the last two compounds induced a strong
    passive haemagglutination reaction, curdlan did not induce any
    reactive antibodies (Arakawa et al., 1974).

         Rats

         Groups of five male Wistar rats, five weeks old, were fed diets
    containing curdlan at 0, 0.1, 1, or 5 g/kg bw per day for 30 days. No
    antibodies against curdlan were induced, as demonstrated by passive
    haemagglutination and Ouchterlony precipitation tests. An
    intraperitoneal injection of 10 g of curdlan or dextran 2000
    (suspended in physiological saline) to groups of five male Wistar rats
    twice a week for four weeks did not induce a specific antibody
    response, whereas an intravenous injection of 1 mg of
    alum-precipitated bovine serum albumin elicited strong antibody
    responses (Arakawa et al., 1974).

    2.2.6.2  Pathogenicity of Alcaligenes faecalis var. myxogenes NTK-u
             strain

         Mice

         Suspensions of dead cells of the curdlan-producing strain of
     A. faecalis var.  myxogenes NTK-u and of  Xanthomonas campestris
    in 5% aqueous gum arabic solution were administered orally to groups
    of eight four-week-old male ICR-JCL mice at doses of 10 or 20 g/kg bw,
    and the animals were observed for seven days. No deaths occurred and
    no abnormalities were found at autopsy on day 8 (Yokotani, 1969).

         Groups of 9-10 four-week-old male ddY-SLC mice, groups of 10
    fasted male ddY-SLC mice, and groups of 9-10 suckling male and female
    ddY-SLC mice aged 14-18 days (believed to be relatively susceptible to
    intestinal infection) received 1010 viable organisms of  A. faecalis
    var.  myxogenes NTK-u orally for two days or, for comparison, the
    non-pathogenic strains  Alcaligeneses faecalis, Escherichia coli
    Umezawa,  Xanthomonas campestris (which produces xanthan gum) and/or
    the pathogenic strains  E. coli 0-78 and  Salmonella enteritidis.
    The animals were autopsied after seven days of observation and the
    respective strains were measured in mesenteric lymph nodes. The only
    abnormal findings or deaths occurred in mice given  S. enteritidis,
    and that was the only strain found in mesenteric lymph nodes.

         The pathogenicity of  A. faecalis var.  myxogenes NTK-u was
    also studied after intravenous, intraperitoneal, and intracerebral
    injection into four-week-old male ddY-SLC mice and compared with the
    pathogenicity of the same non-pathogenic and pathogenic strains of
    bacteria listed above. The LD50 values for  A. faecalis var.
     myxogenes NTK-u were similar to or higher than those for the other
    bacterial strains (Imai, 1972).

    2.2.6.3  Cytotoxicity of Alcaligenes faecalis var. myxogenes NTK-u
             strain to HeLa cells

         Neither the filtrate of a culture medium in which  A. faecalis
    var.  myxogenes NTK-u had been grown, nor the bacterium or its
    components was cytotoxic to HeLa cells at concentrations up to 100
    mg/ml (Ito, 1971).

    2.3  Observations in humans

    2.3.1  Sensitization

         A modified Draize 'multiple insult' patch test, which was not
    performed according to GLP or a current guideline, was used on 213
    healthy male prisoners (73% white, 21% black, 4% Mexican, 2% other;
    average age, 35 years) who received 10 applications of curdlan (dose
    not reported) as an aqueous paste on the skin under occlusion, the
    test material being applied fresh every other day. A 48-h challenge
    application of the test material (concentration and amount not
    reported) was given 10-14 days after the last dose on a different
    site, which was examined after 48, 96, and 144 h. During the induction
    phase, trace irritation was seen in a few men but was not significant.
    No sensitization reactions were seen in the challenge phase (Wazeter
    et al., 1975c). The Committee noted that these results are of limited
    value owing to deficiencies of the tests.

    2.3.2  Studies of adverse side-effects

         In a study of the possible adverse side-effects of curdlan,
    groups of six young men received milk-shakes with or without curdlan.
    Neither the subjects nor the physician conducting the study were aware
    of who received curdlan. The men were given 6 g/day for five days, 35
    g/day for two weeks, and then 50 g/day from day 21 until the end of
    the treatment on day 28. Prior to the study, the past illnesses,
    allergies, and gastrointestinal function of each subject were
    determined, and before and at the end of the study each subject was
    given a complete physical examination, with laboratory tests for blood
    sugar, urea nitrogen, uric acid, cholesterol, calcium, total
    bilirubin, alkaline phosphatase, total protein, albumin and globulin,
    phosphorus, alanine aminotransferase, lactic dehydrogenase, complete
    blood count, haemoglobin, haematocrit, urinary creatinine (24 h), and
    an electrocardiogram. They were also questioned at weekly intervals
    about any problems or adverse effects, with particular emphasis on
    gastrointestinal symptoms. Body weight, blood pressure, and pulse rate
    were recorded.

         No evidence of toxicity was observed during or after the study.
    Three of the six men receiving curdlan reported increased flatulence
    and intestinal pain due to gas, and two subjects reported some
    diarrhoeal movements, in one subject particularly towards the end of
    the study. In contrast, another subject receiving curdlan complained

    of constipation. These transient effects did not cause the withdrawal
    of any subject from the study. No other effects attributable to
    administration of curdlan were reported (Wazeter et al., 1975d).

    4.  COMMENTS

         In two studies, rats given 14C-curdlan at a dose of 20 mg/kg bw
    orally excreted about 80% and 40% of the administered radiolabel,
    respectively, as 14C-carbon dioxide within 24 h. In these studies,
    excretion in the urine represented about 3% and 1.5% of the dose and
    excretion in faeces about 8% and 34%, respectively. After 48 h, 100%
    and 80% were recovered from carbon dioxide, urine, and faeces
    combined. When tetracycline was given concomitantly in the
    drinking-water, excretion as carbon dioxide decreased by one-third,
    whereas excretion in faeces was increased, indicating that intestinal
    microflora may be responsible for the metabolism of this compound.
    Excretion of the radiolabel as carbon dioxide also decreased with
    increasing dose of curdlan, indicating more limited metabolism at
    higher doses. In humans, the faeces appeared to be the main pathway
    for excretion, except for a portion that was fully metabolized to
    carbon dioxide. The extent of metabolism to carbon dioxide in humans
    also appeared to reflect the action of intestinal bacteria: when the
    bacterial microflora were suppressed by pretreatment with antibiotics,
    very limited production of 14C-carbon dioxide was seen.

         Curdlan given to rats at concentrations of 1, 5, or 15% in the
    diet had no effect on the bioavailability of calcium, magnesium, iron,
    zinc, copper, or manganese. The LD50 value in mice and rats treated
    orally was > 10 000 mg/kg bw, and no abnormalities were seen at
    autopsy.

         In short-term and long-term studies in experimental animals, the
    only effects of orally administered curdlan were soft stools and/or
    laxation, reduced body-weight gain, and increased weights of full and
    empty caeca due to the presence of high concentrations of
    'indigestible' curdlan. In an eight-week study in mice and a four-week
    study in rats given curdlan at concentrations up to 300 g/kg of diet,
    the only effects were large faecal pellets, soft stools and/or
    laxation, and increased weights of full and empty caeca.

         In a three-month study in rats, the lowest dose of 50 g curdlan
    per kg of diet was the NOEL. Growth inhibition was seen at the highest
    dose of 200 g/kg of diet, even though food intake was increased. Soft
    stools, enlarged large intestines when full, and increased weights of
    full and empty caeca appeared to be the major effects at 100 and 200
    g/kg of diet. Dose-dependent decreases in platelet counts and protein
    and globulin concentrations and increased serum alkaline phosphatase
    activity, absolute carcass weight, and the relative weights of the
    adrenal and submaxillary glands were seen in males at 100 and 200 g/kg
    of diet. In addition, males at the highest dose had decreased serum
    calcium and cholesterol concentrations, and females at this dose had

    decreased relative pituitary weights and increased relative uterine
    weights. At necropsy, decreased deposition of adipose tissue was seen
    in the abdominal cavities of females at all doses and in males at the
    highest dose.

         In a one-year study in dogs treated in the diet, blood-tinged,
    mucoid, soft stools were seen with curdlan at 150 g/kg of diet and
    with gelled curdlan at 400 g/kg of diet (containing curdlan at 100
    g/kg, providing a final concentration of 40 g/kg of diet). Increased
    full and empty caecal weights were observed with curdlan at a dose of
    150 g/kg of feed. The petaechial haemorrhages and mucosal ecchymosis
    occasionally observed in the small intestinal mucosa of dogs at all
    doses were considered to be unrelated to treatment.

         In a lifetime study of carcinogenicity in mice, addition of
    gelled curdlan at 400 g/kg of diet or curdlan at dietary levels of up
    to 150 g/kg did not cause significant abnormalities, although
    decreased food consumption was seen at the highest dose of curdlan and
    increased food consumption with the gelled curdlan. No changes in
    tumour incidence were observed. In a two-year study in rats, the
    highest dose of curdlan (150 g/kg of diet) decreased growth and food
    consumption and increased the weights of full and empty caeca. The
    gelled curdlan had no effect.

         In another two-year study, rats were exposed  in utero. The
    growth of those exposed to curdlan at 150 g/kg of diet was inhibited
    and they showed a slight decrease in food consumption. Increased
    emptied caecal weights were seen in males given curdlan at 50 g/kg of
    diet, in females given 150 g/kg of diet, and in females given gelled
    curdlan at 400 g/kg of diet. Clinical chemical analyses during
    treatment showed increased aspartate aminotransferase and serum
    alkaline phosphatase activities in animals given the highest dose of
    curdlan or gelled curdlan. Gross and microscopic examination revealed
    a significantly increased incidence of benign uterine polyps in rats
    exposed to curdlan at 150 g/kg of diet, with incidences of 0/50 in
    controls, 3/50 in rats given curdlan at 10 g/kg of diet, 4/47 at 50
    g/kg of diet, 7/50 at 150 g/kg of diet (significant), and 2/50 with
    the gelled curdlan. The authors reported that benign uterine polyps
    were seen infrequently in control animals; the incidences in
    historical controls were not available.

         In a three-generation study of reproductive toxicity in rats,
    with two litters per generation, no effect was seen on fertility,
    gestation, or the viability of the pups. Parents given curdlan at 150
    g/kg of diet or gelled curdlan, providing 400 g/kg of diet, showed
    slight growth inhibition. Food consumption was slightly decreased in
    parents at the highest dose of curdlan. Furthermore, female F2
    parents given the high dose of curdlan or gelled curdlan in the diet
    had increased full and empty caecal weights. The weights of the pups
    in nearly all litters of dams at the high dose of curdlan were

    significantly decreased during lactation: in F1a and F1b litters at
    14 and 21 days of age; in F2a and F3a litters at 4, 7, 10, 14, 17,
    and 21 days of age; in F2b litters only at day 4 of age; and in F3b
    litters only at day 21 of age. The NOEL for both the maternal and
    embryonal toxicity of curdlan was 50 g/kg of diet. Although the
    authors suggested that the decrease in pup weight gain during
    lactation at the highest dose of curdlan was due to consumption by the
    pups of their mothers' diet, it might also have been a
    treatment-related effect or a combination of consumption of the
    mothers' diet and an effect through the milk. In order to investigate
    these suggestions, a number of single-generation studies (two litters
    per generation) were performed in which the offspring of treated
    females were nursed by untreated female rats and the offspring of
    untreated females were nursed by treated dams. A further
    single-generation study was conducted with cellulose at 50 or 150 g/kg
    of diet. In these studies, curdlan or cellulose at 150 g/kg of diet
    significantly decreased pup weight gain during lactation, and transfer
    of pups from treated to control dams during lactation decreased this
    effect. When treatment of the dams with curdlan was withdrawn during
    lactation, the weights of the pups were comparable to those of control
    pups at this time.

         The three-generation study of reproductive toxicity included a
    study of teratogenicity in the F2c litters. No embryotoxic or
    teratogenic effects were observed at any dose of curdlan up to 150
    g/kg of feed or with gelled curdlan at 400 g/kg. In a study of
    teratogenicity in rabbits treated orally by gavage, no effects were
    seen.

         Curdlan was inactive in assays for gene mutation  in vitro in
    bacteria and in mouse lymphoma cells and did not induce chromosomal
    aberrations in hamster ovary cells. It did not induce micronucleus
    formation in mice  in vivo.

         No pathogenicity was observed in mice that received oral doses of
    live or dead cells of the curdlan-producing strain,  Alcaligenes
     faecalis var.  myxogenes NTK-u, or in mice that received
    intravenous, intraperitoneal, or intracerebral injection of live
    organisms. The strain was not cytotoxic to HeLa cells.

         Curdlan was not immunotoxic in mice or rats. It did not induce
    skin sensitization in humans, although this study was of limited
    value.

         In a four-week study in which six volunteers consumed up to 50 g
    of curdlan daily, increased flatulence was observed. One subject who
    consumed 50 g of curdlan per day had some diarrhoea. No evidence of
    toxicity was seen.

    5.  EVALUATION

         Curdlan did not induce genotoxic, carcinogenic, or teratogenic
    effects or effects on reproduction. At high doses, curdlan decreased
    growth and/or food consumption and increased the weights of full
    and/or empty caeca. These effects are commonly observed after the
    consumption of large amounts of 'indigestible' bulking materials.

         The Committee noted the significant increase in the incidence of
    benign uterine polyps in rats exposed to curdlan in utero at the high
    dose of 150 g/kg of diet in the long-term study. The effect appeared
    to be dose-related; however, uterine polyps were not observed in the
    long-term study in mice or in a long-term study in rats of the same
    strain and from the same laboratory that did not involve exposure  in
     utero. These benign growths are known to occur naturally in older
    rats at incidences of 1-20%, depending on the study and strain. In
    view of the lack of genotoxicity and the structure and metabolism of
    curdlan, the Committee allocated a temporary ADI 'not specified'1 for
    use of curdlan as a food additive, pending the provision of
    information on its use and intake. Information necessary for an
    assessment of the intake of curdlan, including its use, the maximum
    and typical expected levels in the food categories in which curdlan is
    proposed for use, and the consumption of foodstuffs that might contain
    curdlan in different regions of the world, is required for evaluation
    in 2001.

    6.  REFERENCES

    Aomori, T. & Tanida, M. (1968) Preliminary acute toxicity of
    polysaccharide 13140 in mice and rats. Unpublished report from
    Biological Research Laboratories, Takeda Chemical Industries Ltd.
    Submitted to WHO by Takeda Chemical Industries Ltd.

    Arakawa, M., Takaoki, M. & Kawaji, H. (1974) Immunogenicity of
    polysaccharide 13140 in rats and mice. Unpublished report from Central
    Research Laboratories, Takeda Chemical Industries Ltd. Submitted to
    WHO by Takeda Chemical Industries Ltd.

                  

    1 ADI 'not specified' is a term applicable to a food component of
    very low toxicity which, on the basis of the available chemical,
    biological, toxicological, and other data, the total dietary intake of
    the substance arising from its use at the levels necessary to achieve
    the desired effect and from its acceptable background in food, does
    not, in the opinion of the Committee, represent a hazard to health.
    For this reason and for those stated in the evaluation, the
    establishment of an ADI expressed in numerical form is deemed
    unnecessary.

    Dillon, D.M. (1994) Curdlan (lot No. FU11S) testing for mutagenic
    activity with Salmonella typhimurium TA 1535, TA 1537, TA 1538, TA 98
    and TA 100 (IRI Project No. 754834). Unpublished report No. 10239 from
    Inveresk Research International Ltd. Submitted to WHO by Takeda
    Chemical Industries Ltd.

    Goldenthal, E.I. (1994) 8-Week mineral bioavailability study in rats.
    Unpublished report No. 295-154 from International Research and
    Development Corporation dated 8 September 1994. Submitted to WHO by
    Takeda Chemical Industries Ltd.

    Holmstrom, L.M. & Innes, D.C. (1994) Curdlan (lot No. FU11S)
    micronucleus test in bone marrow of CD-1 mice (IRI Project No.
    754860). Unpublished report No. 10210 from Inveresk Research
    International Ltd. Submitted to WHO by Takeda Chemical Industries Ltd.

    Imai, A. (1972) Pathogenicity of  Alcaligenes faecalis var.
     myxogenes NTK-u strain in mice. Unpublished report from Biological
    Research Laboratories, Takeda Chemical Industries, Ltd. Submitted to
    WHO by Takeda Chemical Industries Ltd.

    Ito, H. (1971) Cytotoxic effects of culture medium of  Alcaligenes
     faecalis var.  myxogenes NTK-u; IFO 13140, and its cell components
    on the Hela cell. Unpublished report from Biological Research
    Laboratories, Takeda Chemical Industries, Ltd. Submitted to WHO by
    Takeda Chemical Industries Ltd.

    Leddy, I.A. (1994) Curdlan (lot No. FU11S) chromosomal aberrations
    assay with Chinese hamster ovary cells in vitro (OECD Protocol) (IRI
    Project No. 754855). Unpublished report No. 10337 from Inveresk
    Research International Ltd. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Matsuo, T. & Suzuoki, Z. (1972) Metabolism of polysaccharide 13140 in
    the rat. Unpublished report from Biological Research Laboratories,
    Takeda Chemical Industries, Ltd. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Matsuo, T. & Suzuoki, Z. (1974a) Nutritional study on polysaccharide
    13140 in the rat. Unpublished report from Biological Research
    Laboratories, Takeda Chemical Industries, Ltd. Submitted to WHO by
    Takeda Chemical Industries Ltd.

    Matsuo, T. & Suzuoki, Z. (1974b) Caloric inertness of polysaccharide
    13140 in the rat. Unpublished report from Biological Research
    Laboratories, Takeda Chemical Industries, Ltd. Submitted to WHO by
    Takeda Chemical Industries Ltd.

    Nakaguchi, T., Hosokawa, S., Aono, M., Orita, S. & Aramaki, Y. (1972)
    Oral three month toxicity study of polysaccharide 13140 in rats.
    Unpublished report from Biological Research Laboratories, Takeda
    Chemical Industries Ltd. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Riach, C.G. & Willington, S.E. (1994) Curdlan (FU11S) mouse lymphoma
    assay (IRI Project No. 754876). Unpublished report No. 1026 from
    Inveresk Research International Ltd. Submitted to WHO by Takeda
    Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I. & Geil, R.G. (1973) Compound: PS13140.
    Subject: Eight week tolerance study in mice dated 4 December 1974.
    Unpublished report No.295-009 from International Research and
    Development Corporation. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I. & Harris, S.B. (1974) Compound: PS
    13140. Subject: Teratology study in rabbits dated 16 September 1974.
    Unpublished report No. 295-013 from International Research and
    Development Corporation. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Pollycove, M. & Freeman, L. (1975a)
    Metabolism study of PS 13140 in humans. Unpublished report No. 295-023
    from International Research and Development Corporation. Submitted to
    WHO by Takeda Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Cookson, K.M., Geil, R.G., Blanchard,
    G.L. & Keller, W.F. (1975b) One year feeding study in dogs dated 8
    January 1975. Unpublished report No. 295-008 from International
    Research and Development Corporation. Submitted to WHO by Takeda
    Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I. & Epstein, W.L. (1975c) Compound: PS
    13140. Subject: Modified Draize multiple insult patch test dated 22
    August 1975. Unpublished report No. 295-022 from International
    Research and Development Corporation. Submitted to WHO by Takeda
    Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Batterman, R.C. & Freeman, L. (1975d)
    Human tolerance study of PS 13140. Unpublished report No. 295-024 from
    International Research and Development Corporation. Submitted to WHO
    by Takeda Chemical Industries Ltd.

    Wazeter, F.X, Goldenthal, E.I. & Benson, B.W. (1976a) Compound: PS
    13140. Subject: Ninety day paired feeding study in rats dated 5
    February 1976. Unpublished report No. 295-018 from International
    Research and Development Corporation. Submitted to WHO by Takeda
    Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Geil, R.G., Jessup, D.C., Rac, V.S. &
    Benson, B.W. (1976b) Compound: PS 13140. Subject: Lifetime
    carcinogenic study in mice dated 5 August 1976. Unpublished report No.
    295-010 from International Research and Development Corporation.
    Submitted to WHO by Takeda Chemical Industries Ltd.

    Wazeter, F.X, Goldenthal, E.I., Benson, B.W., Geil, R.G., Keller, W.G.
    & Blanchard, G.L. (1976c) Compound: PS 13140. Subject: Two year
    feeding study in rats dated 18 August 1976. Unpublished report No.
    295-014 from International Research and Development Corporation.
    Submitted to WHO by Takeda Chemical Industries Ltd.

    Wazeter, F.X, Goldenthal, E.I., Geil, R.G., Jessup, D.C., Benson,
    B.W., Keller, W.G. & Blanchard, G.L. (1976d) Compound: PS 13140.
    Subject: Lifetime feeding study in rats dated 30 December 1976.
    Unpublished report No. 295-011 from International Research and
    Development Corporation. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Geil, R.G., Jessup, D.C. & Harris,
    S.B. (1976e) Compound: PS 13140. Subject: Multigeneration reproduction
    study in rats dated 4 May 1976. Unpublished report No. 295-012 from
    International Research and Development Corporation. Submitted to WHO
    by Takeda Chemical Industries Ltd.

    Wazeter, F.X, Goldenthal, E.I., Jessup, D.C. & Harris, S.B. (1976f)
    Compound: PS 13140. Subject: Single generation reproduction study in
    rats dated 20 August 1976. Unpublished report No. 295-019 from
    International Research and Development Corporation. Submitted to WHO
    by Takeda Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Jessup, D.C. & Harris, S.B. (1976g)
    Compound: Cellulose. Subject: Single generation reproduction study in
    rats dated 17 May 1976. Unpublished report no. 295-021 from
    International Research and Development Corporation. Submitted to WHO
    by Takeda Chemical Industries Ltd.

    Wazeter, F.X., Goldenthal, E.I., Jessup, D.C. & Harris, S.B. (1976h)
    Compound: PS 13140. Subject: Single generation reproduction study in
    rats dated 12 April 1976. Unpublished report No. 295-020 from
    International Research and Development Corporation. Submitted to WHO
    by Takeda Chemical Industries Ltd.

    Woodard, M.W. & Imming, M. (1971) Dose range finding study of
    polysaccharide 13140 by dietary feeding to dogs. Unpublished report
    from Woodard Research Corporation. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Yokotani, H. (1969) Acute toxicity of the dead cells of  Alcaligenes
     faecalis var.  myxogenes strain NTK-u; IFO, 13140 in mice.
    Unpublished report from Biological Research Laboratories, Takeda
    Chemical Industries, Ltd. Submitted to WHO by Takeda Chemical
    Industries Ltd.

    Yokotani, H., Ogihara, S., Yamada, H., Orita, S., Nomura, M. &
    Matsukura, A. (1969) Oral one month toxicity study of polysaccharide
    13140 in rats. Unpublished report from Takeda Chemical Industries Ltd.
    Submitted to WHO by Takeda Chemical Industries Ltd.
    


    See Also:
       Toxicological Abbreviations