3.1. Microbial gene mutation assays

Folpet and captan generally produce positive results in microbial gene mutation assays, which include assays in Salmonella, Esherichia coli, Bacillus subtilis, and also yeast mitotic recombination assays (see Bridges, 1975, for a review of the early studies). In one of the first studies utilizing a standard Ames plate incorporation method (Ames et al., 1975), folpet and captan were positive in strains TA1535 and TA100, responsive to base pair mutation, whereas strains TA1537 and TA1538, sensitive to frameshift mutation, were either negative or equivocal (Simmon et al., 1977). This specificity for base pair mutation was consistent with the observations in the earlier work (Ficsor and Nii, 1972) and was borne out in subsequent investigations (Hour et al., 1998; Ohta et al., 2002). Although captan and folpet generally show greater activity in base pair–sensitive strains, they are positive in the frameshift strains as well (Clarke, 1971; De Flora et al., 1984; Moriya et al., 1983; Nakajima, 1998; Ruiz and Marzin, 1997), particularly when tested in the more sensitive assays and strains. The DNA damage induced by captan is subject to excision repair as demonstrated by negative responses in the exrA⁻ strains and positive responses in strains containing mutations in the uvrA⁻ gene (Bridges et al., 1972, 1973). Captan and folpet were later shown to be mutagenic in strain TA104, lacking excision repair, but negative or less mutagenic in strain TA102, which has an intact excision repair system (Barrueco and de la Pena, 1988). Many other bacterial studies have also demonstrated positive results (Carere et al., 1978; Nagy et al., 1975; Shiau et al., 1981; Shirasu et al., 1976). Nonbacterial microbial systems, Saccharomyces and Aspergillus, using bacterial methodology similar to the Ames test, produced positive mutagenicity results for captan and folpet (Bignami et al., 1977; Simmon et al., 1977).

Rat and human blood and to a lesser extent plasma were shown to inactivate captan in the early S. typhimurium tester strain TA1950 (Ficsor et al., 1977). The involvement of thiols was reported in a short communication showing mutagenic activity could be significantly reduced or eliminated when folpet and captan were preincubated with rat liver S9, rat blood, or cysteine (Moriya et al., 1978). Dose-response studies (Carver et al., 1985) evaluating the addition of cysteine or glutathione to captan or folpet further linked increased thiol-rich additives to decreased mutagenic activity. The reduction of the mutagenic activity of folpet by the addition of thiol-rich S9 was confirmed using a preincubation assay (Nakajima, 1998). The mitigation of mutagenicity is also affected by the length of preincubation (May, 1993). Thus, mutagenic activity was greatest in absence of S9 followed by the addition of S9 without preincubation and decreasing as periods of preincubation increased.

Consistent with inactivation of captan and folpet by blood and liver homogenates, host-mediated tests in which Salmonella tester strains are injected in rodents were negative (Ficsor et al., 1977; Kennedy et al., 1975).

Folpet and captan are mutagenic in most microbial test systems; however, both are rapidly degraded when allowed to react with thiols added to the test system with resultant decreases in mutagenic activity.

3.2. Mammalian in vitro gene mutation assays

The mammalian in vitro gene mutation assay systems generated results consistent with the microbial gene mutation assays, although they were generally less responsive. Mutagenic activity was also abated in the presence of thiol-rich components. There are several different mammalian mutation assay systems commonly used. They fall into two general categories, those performed in cell suspension like the mouse lymphoma assay and those treated in monolayers like Chinese hamster ovary (CHO) or hamster V79 cell assays. All employ reverse mutation for detection of gene mutation, similar to the S. typhimurium test systems. The mouse lymphoma system has the capability of detecting chromosomal damage in addition to gene mutations when small colonies are scored.

Folpet was tested in the mouse lymphoma assay (soft agar) as part of a large program to evaluate 20 pesticides for mutagenic activity (Jotz et al., 1980 as cited by Clock, 1997). Folpet was found to increase mutation frequency when dosed from 0.3 to 0.51 μg/ml without S9, but was 10 times less active in the presence of S9 (US EPA, 1997). Captan was evaluated in the mouse lymphoma assay (soft agar) at doses from 0.01 to 0.1 µg/ml (Oberly et al., 1984). There was a dose-related increase in mutation frequency (mutants per 10⁵ surviving cells) from 0.06 to 0.1 µg/ml. Toxicity was evident at 0.1 µg/ml resulting in 39% survival (expressed as relative total growth [RTG]). The mutation frequency at 0.06, 0.08, and 0.1 µg/ml was 57, 63, and 72 mutants per 10⁶ surviving cells, respectively. However, the control mutation frequency of 21 mutants per 10⁶ surviving cells was relatively low compared to the currently acceptable range for negative-control frequencies of 35–140 mutants per 10⁶ surviving cells (Moore et al., 2006). The unusually low background mutation frequency confounds the assessment.

A mouse lymphoma assay was performed with captan both with and without S9 using the soft agar method (Edgar et al., 1985). No mutagenic activity was observed in the presence of S9. There was a dose-related increase in mutation frequency for two trials in the absence of S9. In the first trial at 0.2 and 0.3 µg/ml, there were 119 and 235 mutants per 10⁶ surviving cells, respectively, compared to 57 mutants per 10⁶ surviving cells in the control cultures. Survival for these doses was 32% and 6% relative total growth (RTG), respectively. In the second trial, only the 0.2 µg/ml dose yielded a mutation frequency (237 mutants per 10⁶ surviving cells) above background levels, but there was only 5% RTG at this concentration. Current criteria set a 10% RTG limit on cytotoxicity for mutation analysis (Moore et al., 2006); thus, the cultures showing 6% RTG (trial 1) and 5% RTG (trial 2) would have been excluded from analysis if they were assessed today. The early mouse lymphoma studies are difficult to interpret. Assay methods and conditions were in flux at the time of their performance, as were the data evaluation criteria.

The in vitro mammalian mutation assays utilizing the CHO or Chinese hamster V79 gene mutation assays are more informative. Folpet and captan were evaluated in a CHO/hypoxanthine-guanine phosphoribosyltransferase (HGPRT) assay that was pioneered in the conducting laboratory (O’Neill et al., 1981). A positive response was observed in three independent trials in the absence of S9 at a dose concentration of 0.25 µg/ml and above. In that study, tissue culture flasks were sealed to prevent volatilization and incubation of the cells with the chemical (5 hours) was done in the absence of fetal calf serum. These conditions optimized the potential for detection of mutations by capping of the flasks, which did not allow dissipation of thiophosgene, and with the lack of calf serum in the flasks, which minimized the reaction of folpet and thiophosgene with serum thiols. No trials were done in the presence of S9. Folpet was negative in Chinese hamster V79 cells (HGPRT locus) with and without S9 at two expression times (Bootman et al., 1986). Concentrations ranged from 0.25 to 2 µg/ml without S9 and 3.125 to 50 µg/ml with S9.

Overall, folpet and captan displayed a positive but weak mutagenic response in the in vitro mammalian gene mutation assays. In these test systems the mutagenicity response was much less than the relatively potent responses observed in the microbial test systems without the presence of S9.

3.3. Mammalian in vitro chromosomal assays

As opposed to mutation assays that detect specific gene defects, the chromosomal assays evaluate the structure of the whole chromosome. As the cell goes into mitosis, it is possible to observe the individual metaphase chromosomes. The chromosomes are stained, identified, and microscopically evaluated for structural changes such as breaks, deletions, and other more complex forms of structural damage such as quadiradials. A wide variety of cell culture systems can be used including primary human lymphocytes.

Three studies of folpet and two studies of captan have evaluated the induction of chromosomal damage in mammalian tissue culture. As in the gene mutation assays, negative, equivocal, as well as positive results were obtained using these in vitro assays. In a study evaluating chromosomal aberrations, CHO cells were exposed to various concentrations of folpet for either 10 or 20 hours in the absence and presence of S9 (Loveday, 1989). Dose ranges from 0.08 to 2.5 µg/ml in trials without S9 and from 0.8 to 75 µg/ml with S9 were used. Positive results were obtained in the CHO cells following both 10 and 20 hours of exposure in the absence of S9 but at one concentration only, 0.75 µg/ml, the highest scorable dose based on excessive cytotoxicity. Aberrant chromosomes were primarily chromatid breaks with some triradials and quadradials (18% cells with aberrations). Similar results were obtained in trials using S9. The 20-hour assay displayed a clearer dose-response. A positive response was observed at 25 µg/ml in the presence of reduced mitotic activity with higher doses unscorable. Concentrations of folpet necessary to induce a response (27% cells with aberrations) at the highest scorable dose were 10- to 30-fold greater in the presence of S9 than in its absence.

The effects on chromosomal structure following exposure to folpet were investigated in primary human lymphocytes taken from whole blood (Bootman et al., 1987). Tests were conducted with triplicate cultures and in two independent trials, both with and without S9. In the first trial, cells were incubated with folpet for 2 hours with S9 and 24 hours without S9. For the second trial, cultures both with and without S9 were exposed for 2 hours. In the first trial without S9, there was evidence of weak activity (single-chromatid breaks and fragments) at concentrations of 3 µg/ml. When retested in the second trial, no clastogenicity was observed at doses of 3 or 5 µg/ml. No clastogenic activity was observed for folpet in either trial with S9.

Folpet was also evaluated in a chromosomal assay in a human lymphoid cell line and a Burkitt lymphoma cell line at 0.5–4 µg/ml (Sirianni and Huang, 1978). No treatments were performed with the addition of S9. The authors reported chromosome or chromatid gaps and breaks, and severe cell growth inhibition in both cultured human lymphoid cell lines after short periods of exposure. Following longer treatment periods, fewer breaks were noted but despiralized or pulverized chromosomes were increased likely due to increased cytotoxicity.

The potential of captan to induce chromosomal aberrations in diploid human fibroblast cultures was evaluated at doses of 0 through 4 µg/ml (Tezuka et al., 1978). Significant mitotic inhibition was observed but no chromosomal aberrations were induced at cytotoxic dosages of 3 and 4 µg/ml captan. Captan induced chromosomal aberrations in Chinese hamster V79 cells in vitro at doses of 45 µM (13.5 µg/ml) and higher (Tezuka et al., 1980). No cytotoxicity was observed even at the highest dose level in the V79 cell culture. These divergent results are likely due to the exposure to significantly higher concentration of captan used in the V79 cultures and the different cytotoxicity sensitivities of the two test systems.

Folpet and captan are weak clastogens in some mammalian in vitro cell cultures, resulting primarily in chromosome and chromatid breaks. Analogous to the results in the bacterial and in vitro mammalian gene mutation assays, the addition of a rat liver homogenate mitigated the clastogenic activity of folpet.

3.4. In vitro ancillary tests

Other assays have been developed to assess the ability of a compound to cause DNA or chromosomal damage. These tend to be sentinel assays that indicate the potential for mutation but do not demonstrate the compound is actually a mutagen. For example, unscheduled DNA synthesis (UDS) detects DNA damage that is repaired by a long patch excision repair system. The fact that the damage is repaired is itself evidence that, although the cell sustained DNA damage, it has the ability to recover before mutation occurs. Likewise, sister-chromatid exchange (SCE) assays monitor the exchange of sister chromatids within a chromosome. This recombinational event is a natural process that can repair damaged DNA and is increased following damage to DNA.

Unscheduled DNA synthesis (UDS) was monitored in human fibroblast cultures following treatment with folpet or captan both with and without S9. Captan and folpet were assessed as positive (2-fold background) in the presence of S9 but was negative in the absence of S9 (Simmon et al., 1977). A second trial was subsequently performed by the same investigators and the results of both trials were then reassessed as negative (Jones et al., 1984). UDS was not detected in primary human lymphocyte and rat thymocyte cultures exposed to folpet or captan (Rocchi et al., 1980). Both these studies measured incorporated radiolabeled thymidine in DNA extracts and did not employ the current UDS methodology based on auto radiographic grain counts.

Sister-chromatid exchanges (SCE) were not induced in V79 cells following exposure to folpet (Sirianni, 1978). Another in vitro SCE study was performed in Chinese hamster V79 cultures with captan at higher doses (Tezuka et al., 1980). This study was positive, as was the Chinese hamster V79 chromosomal aberration study performed in the same laboratory. Captan was marginally positive (18.7 compared to 15.7 SCE/cell for captan and control, respectively) at one dose in cultured human lymphocytes (Vigfusson and Vyse, 1980). None of these SCE assays included S9. As a sentinel test, SCE induction would be expected if the test material causes chromatid or chromosomal breaks.

It was reported that captan causes chromosomal breaks in Chinese hamster V79 cells as evidenced by alkaline elution of the DNA fragments extracted from these cells; the effect was prevented in the presence of rat S9 (Swenberg et al., 1976). Captan induced DNA strand breaks in in vitro cultured human diploid fibroblast cultures at 99.0 µM (Snyder, 1992). Addition of the excision repair inhibitor 1-β-d-arabinofuranosylcytosine and hydroxyurea led to an increased number of strand breaks at lower doses of captan. Strand breaks were only observed in closed systems indicating a role of a volatile product. The alkaline elution data are consistent with the induction of SCE and the visual measure of chromosomal breaks in the same cell culture system (Tezuka et al., 1980).

4.1. Thiophosgene

As early as 1972, the mutagenic activity of captan was thought to be due to a volatile breakdown product. The role of volatile captan products using bacterial mutagenesis systems was evaluated in E. coli (Bridges et al., 1972). In a series of studies, captan was placed on the lid of the Petri dish with the bacteria on agar below. Thus, only a volatile chemical reaction product of captan would be able to reach the bacteria. Mutants observed were distributed randomly on the plates, indicating that a volatile component of captan was mutagenic. Knowing that captan degrades much faster as pH rises, captan-impregnated filter paper was moistened with buffers with increasing pH values and the number of mutants increased with alkalinity. From the lack of mutants following extraction of captan from the filters and by bubbling air through saturated suspensions of captan, it was demonstrated that there was no build-up of breakdown product and the mutagenic product was short-lived. Thus, it was inferred that mutagenicity of captan was a result of a “continuous release of an unstable mutagenic breakdown product from the captan-water interface” (Bridges et al., 1972).

By using specially constructed mutants of E. coli containing mutations in recA, exrA, and uvrA, it was demonstrated again that the mutagenicity of captan was mediated by a volatile mutagen and that it was dependent on reduced DNA repair (uvrA only) capability (Bridges et al., 1973). Similar results were reported for folpet (Bridges, 1975).

In an in vitro mammalian cell study in V79 hamster cells similar to the E. coli studies already described, captan was removed from contact with the cell culture and activated by the addition of carbonate to generate volatile degradation products (Arlett et al., 1975). Both 8-azaguanine (8-AG) and ouabain (Oua) were used as selective agents. Positive results were obtained, indicating the mutagenic component was a volatile product of carbonate-reacted captan. In addition, captan cytotoxicity and mutagenicity disappeared if cells and captan were added to media containing 10% fetal calf serum, allowing the interaction of captan or captan products with serum thiols. In a meeting abstract (data not presented), captan, folpet, and thiophosgene were reported to be mutagenic in S. typhimurium (Rideg, 1982).

4.2. Ring structures and metabolites

THPI, the primary degradate of captan, was not mutagenic in bacterial test systems (Barrueco and de la Pena, 1988; Carver, 1986; EC, 2000a). THPI did not induce micronuclei or apoptotic bodies in mouse duodenal crypt cells following gavage dosing from 250 to 1500 mg/kg (Chidiac and Goldberg, 1987).

PI, the primary degradate of folpet, was not mutagenic in the Ames assay nor in yeast (EC, 2000b; Herbold, 1981; Rideg, 1982). PI was negative in the mouse lymphoma TK^+/− gene mutation assay (US EPA, 1984). In a cytogenetic assay performed using human peripheral blood lymphocytes (Pilinskaya, 1986), no chromosomal aberrations were observed. A structurally related product, phthalamide, was also negative in the mouse lymphoma assay performed in the absence of S9 (McGregor et al., 1988). In addition, EPA has judged that phthalimide is not mutagenic or clastogenic and is inactive in tissue culture assays for inhibition of cell growth, differentiation or proliferation (Schnaubelt, 1995).

Phthalic acid, the secondary degradate of PI, was evaluated as negative in the Ames assay (Rideg, 1982; Sayato et al., 1987). It showed no mutagenic effects in the Chinese hamster ovary chromosomal aberration assay nor in the in vivo mouse micronucleus assay (Lee and Lee, 2007).

5.1. Gene mutation somatic cell assays

There are very few in vivo tests for gene mutation. One test, the mouse somatic cell assay, measures gene mutation in heterozygous alleles for coat color in mouse embryos exposed in utero (Russell et al., 1981). This assay can detect genetic effects of several kinds, including intragenic mutations, minute deficiencies, chromosomal deletions, and chromosomal crossovers. The somatic cell assay is sometimes referred to as the mouse spot test or the mouse specific locus test and is distinct from the heritable gene mutation assay (Russell et al., 1979) that detects germ cell mutation. Both use the same T-strain of mice with the same heterozygous coat color allele.

Folpet was tested in the mouse somatic cell mutation assay (Moore, 1985) (Table 2). After being co-housed with T-strain male mice, C57Bl/6 pregnant female mice were fed diets containing 0, 100, 1500 or 5000 ppm folpet (equivalent to 11–647 mg/kg) during gestation days (GDs) 8.5 to 12.5. On GD 10.5, pregnant dams received an intraperitoneal injection of 50 mg/kg N-ethyl-N-nitrosourea (ENU) as a positive control. There was maternal and consequent fetal and neonatal toxicity at 5000 ppm. On days 12 and 28 of lactation, there was no significant increase in the number of pups with recessive coat spots (RCS) or differentiation spots (DS) in any group fed diets containing folpet, as contrasted to a significant increase in the number of pups with RCS delivered from the ENU-treated dams. On day 28 of lactation, there was an increase in pups with white mid-ventral spots (WMVS) in the 5000 ppm group. WMVS are due to melanocyte toxicity and vary with the age of the parental female, time of year, and diet and do not indicate mutation potential (Russell et al., 1981).

Captan was evaluated in the same assay at dietary concentrations of 100, 1000, or 5000 ppm (equivalent to 11–419 mg/kg/day) along with positive and negative controls (Nguyen and McGowan, 1980, as reported in Wilkinson et al., 2004). Pregnant dams were fed captan for 5 days beginning on day 8 and ending on day 12 of gestation. Pups were evaluated for “spots” on day 12 and at weaning. The results show that captan did not induce somatic mutations in mice at the three concentrations tested. In a similar study, it was reported in an abstract (Imanishi et al., 1987) that captan induced a frequency of recessive color spots of 2.2% after an intraperitoneal (i.p.) injection of 15 mg/kg, approximating the low dose in the Nguyen study. The background level of recessive color spots in the Nguyen study was 2.9% (Wilkinson et al., 2004), above the reported 2.2% generating the positive result (Imanishi et al., 1987). Based on the two studies, neither folpet nor captan induced in vivo somatic cell mutation (Nguyen and McGowan, 1980, as reported in Wilkinson et al., 2004; Moore, 1985).

5.2. Chromosomal somatic cell assays

Just as chromosomal aberrations can be evaluated in vitro, they can be assessed after in vivo exposure. Although chromosomal aberration can be evaluated in any tissue, the most common tissues used are the bone marrow and peripheral blood. Scoring for chromosomal aberrations is laborious and requires a trained cytogeneticist; however, the micronucleus assay is much simpler to score, and has become an effective substitute for the chromosomal aberration assay (Zeiger, 2004). Both chromosomal aberration and micronucleus assays are widely used to evaluate structural chromosomal damage. Micronuclei arise from chromosomal fragments or whole chromosomes that are not incorporated into daughter nuclei at mitosis and can be scored by differential staining and counting alone. Micronuclei are generally scored in mature erythrocytes of bone marrow or peripheral blood. As the erythrocyte matures, the nucleus is extruded leaving only micronuclei, if present. Both captan and folpet have been evaluated using both direct chromosomal aberration identification and micronuclei formation.

Folpet has been assessed for the ability to induce chromosomal aberrations in both rat and mouse bone marrow in numerous studies with negative results. Folpet showed no clastogenic activity when administered (i.p.) to Swiss albino mice at doses from 125 to 500 mg/kg; there was no increase in chromosomal aberrations in bone marrow cells at 3, 6, or 24 hours after treatment (Sirianni, 1978). This abstract reported conclusions only with no supportive data available for review. Folpet was evaluated in Sprague-Dawley rat bone marrow following gavage of doses ranging from 150 to 2000 mg/kg (Esber, 1983). Some random aberrations were seen in rats treated with folpet but were not treatment related. Folpet was shown to have no clastogenic activity in male or female rats at post-treatment times of 6, 24, or 48 hours, confirming the earlier results in mice (Sirianni, 1978).

Folpet was also evaluated for its potential to induce chromosomal damage using the bone marrow micronucleus test in CD-1 mice (Jacoby, 1985a). Male and female mice were administered 10, 50 or 250 mg/kg folpet by gavage. All groups were sacrificed at 24 hours with additional control groups, and the 250 mg/kg groups also evaluated at 48 and 72 hours. No toxicity was observed in the animals or in the bone marrow. Treatment with folpet did not result in any significant increase in the frequency of micronucleated polychromatic erythrocytes (PCEs) or normochromatic erythrocytes (NCEs) at any harvest time (1000 PCEs scored/animal).

Tezuka and co-workers showed captan did not induce chromosomal aberrations in Wistar rat bone marrow whether administered once at oral doses of 500, 1000, or 2000 mg/kg or on 5 consecutive days at 200, 400, or 800 mg/kg (Tezuka et al., 1978). When Swiss albino mice were i.p. dosed with a 50% captan formulation at 250 mg/kg, chromosomal aberrations were not found (Fry and Ficsor, 1978). In contrast to these studies, it was reported that captan induced chromosomal aberrations in spermatogonia and chromosomal aberrations and micronuclei in the bone marrow of an unspecified strain of mice (Feng and Lin, 1987). This study is of questionable quality and has been excluded from the analysis.¹

In another chromosomal aberration study, captan was administered to Sprague-Dawley rats by gavage at 0 (0.5% gum tragacanth), 200, 400, or 800 mg/kg per day for 5 consecutive days (Bootman and Whalley, 1979). There was no evidence of chromosomal damage in rat bone marrow at any of the doses tested. Likewise captan did not induce micronuclei in the bone marrow of five male or female CD-1 mice (Jacoby, 1985b). No toxicity or bone marrow cytotoxicity was observed in any treatment with the exception of the positive control group. Captan did not induce micronuclei in either sex at any harvest interval.

In more specialized studies to evaluate nuclear aberrations (micronuclei and apoptotic bodies) in duodenal crypt cells of C57Bl/6J mice exposed in the diet or in CD-1 mice exposed by gavage, it was shown that captan did not induce nuclear aberrations at the tumor target site (Chidiac and Goldberg, 1987). In a similar study, folpet did not induce nuclear aberrations in duodenal crypt cells of CD-1 mice (Gudi and Krsmanovic, 2001). This important study and others conducted in the duodenum will be discussed in detail (Section 6) in relationship to the proposed mode of tumorigenic action. Thus, neither folpet nor captan has been shown to induce chromosomal damage in vivo in studies of acceptable quality.

5.3. Ancillary somatic cell in vivo assays

It has been demonstrated that captan does not induce unscheduled DNA synthesis (UDS) in vivo (Kennelly, 1990). In this study, male Alpk:ApfSD rats were administered captan by gavage at 500, 1000, or 2000 mg/kg body weight and UDS was evaluated at 4 and 12 hours post-treatment in cultures of excised liver cells. DNA damage was also evaluated in duodenal crypt cells of CD-1 mice exposed to high doses of folpet using the Comet assay. No DNA damage was observed (Clay, 2004). This study is discussed in Section 6.

5.4. In vivo germ cell gene mutation

Although not a mammalian test system, the Drosophila sex-linked recessive lethal test is considered a measure of germ cell gene mutation in vivo. Mutations in the X-chromosome of D. melanogaster that are phenotypically expressed in males result in lethality. The test detects the occurrence of forward mutations, point mutations, and small deletions in the germ line of the insect.

Sex-linked recessive lethal (SLRL) studies have been performed with captan and folpet with no effect (Kramers and Knaap, 1973; Mollet, 1973; Vogel and Chandler, 1974). However, it was reported that captan and folpet exhibited weak activity at 2000 ppm (Valencia, 1981). In the captan study, when multiples or clusters were appropriately removed from the statistical assessment, both the chromosomal and dominant lethal endpoints were not statistically significantly different compared to controls. With the exception of the weak folpet activity (Valencia, 1981), all studies for SLRL were negative. In a different Drosophila test evaluating wing spots (somatic), captan did not result in an increase in twin spots, which are indicative of mitotic recombination; however, there was weak overall mutagenic activity (Rahden-Staron, 2002). Drosophila somatic cell assays with captan were also negative in a recombination assay (Mollet and Wurgler, 1974).

It should be noted that Drosophila does not possess the glutathione reductase system present in mammalian organisms (Kanzok et al., 2001). Because glutathione reductase systems are involved in the inactivation of folpet and thiophosgene in vivo and are in such abundance in mammalian organisms, Drosophila assays are of limited relevance to mutagenic potential of captan or folpet in higher animals and in humans.

5.5. In vivo germ cell chromosomal mutation

Several studies have been conducted to evaluate the ability of folpet and captan to induce dominant lethality in mice and rats. The term dominant lethal is used to describe a genetic change in a gamete that kills the conceptus early in development. Male animals are treated and mated with untreated females over a period of several weeks to estimate a chemical’s effect on different stages of spermatogenesis. The uterine contents are evaluated for early and late deaths and living fetuses. The induction of dominant lethality is determined by the increase in pre- and post-implantation loss of zygotes in each group compared to control groups. This assay detects major chromosomal defects resulting in fetus/pup lethality.

Neither captan nor folpet induced dominant lethal effects in ICR/SIM mice following a 7-week exposure at dietary levels of 1250, 2500, or 5000 mg/kg (Simmon et al., 1977). Negative results were also obtained when using analytical grade captan in C3H mice (Tezuka et al., 1978). Captan was negative for dominant lethality in Swiss mice; however, individual data were not presented (Epstein et al., 1972).

In contrast to these studies, somewhat mixed results were obtained in a series of two studies performed on captan (Collins, 1972b) and folpet (Collins, 1972a). Osborne-Mendel rats (captan and folpet) and CA-J mice (captan) were administered test compound to 15 male animals per dose group for 5 days by both gavage (50–200 mg/kg/day) and intraperitoneal (2.5–10 mg/kg/day) routes. Males were mated with different groups of untreated females for 10–12 weeks. Antifertility effects were minimal following captan or folpet treatment in the rat or captan treatment in the mouse by either route of administration. Likewise the numbers of mean total implants were not significantly affected in any treatment group for either chemical. In rats after gavage or i.p. dosing, neither captan nor folpet showed statistically significant increases in the percentage of litters with one or more early deaths, but increases appeared for litters with two or more early deaths during the early weeks of the study. Similar results were found in the mouse following captan treatment.

In an attempt to replicate the Collins study with folpet (Collins, 1972a), folpet (gavage doses ranging from 50 to 200 mg/kg) was evaluated for similar endpoints but with 20 Osborne-Mendel rats per group compared with the 15 animals used in the Collins studies (Bradfield, 1980). The total number of implants, corpora lutea, live implants, early deaths, and late deaths were comparable to controls. There was no increase in any folpet-treatment group of litters with two or more early deaths. The positive control, triethylenemelamine, displayed antifertility effects in addition to increases in the number of litters with two or more early deaths. Thus, Bradfield (1980) was not able to confirm the results of Collins (Collins, 1972a).

The Joint Meeting on Pesticide Residues initially noted that data on the dominant lethal effects of captan were equivocal (WHO/FAO, 1983). A later review noted that the dominant lethal studies were predominantly negative (WHO/FAO, 1990). The International Agency for Cancer Research notes reported positive dominant lethal results were not confirmed by other studies (IARC, 1983). Overall, the weight of evidence (WOE) shows folpet and captan do not induce dominant lethal effects.

6.1. Duodenal chromosomal aberration

Compounds that are clastogenic will produce nuclear aberrations when cells undergo mitosis due to chromosomal damage. This damage will show up as micronuclei and apoptotic bodies. Neither captan nor folpet induced clastogenic changes in the mouse duodenum following oral exposures (Chidiac and Goldberg, 1987; Gudi and Krsmanovic, 2001).

Nuclear aberrations in mouse duodenal crypts were measured following dietary administration of captan to mice in a series of studies (Chidiac and Goldberg, 1987). Captan was fed in the diet at 0, 8000, and 16,000 ppm for 7 days to five or six male C57B1/6J mice per group. Nuclear aberrations were scored as aberrant nuclei (apoptotic bodies and micronuclei) in mouse duodenal crypts. There was no difference in nuclear aberrations between treated and control animals. Likewise, doses of 0, 20, 200, or 2000/1000 mg/kg captan administered by gavage to 10 male CD-1 mice for 5 days did not induce nuclear aberrations. In addition, a variety of different purities and formulations of captan were evaluated and also found negative for nuclear aberrations at doses of 200 and 2,000 mg/kg. Depletion of glutathione by pretreatment with l-buthionine-S,R-sulfoximine (BSO) did not induce nuclear aberrations in captan-treated mice (Chidiac and Goldberg, 1987).

In a similar study, folpet did not induce micronuclei or apoptotic cells in the duodenal crypts in CD-1 mice (Gudi and Krsmanovic, 2001). Mice were administered vehicle control (1% gun tragacanth, 0.05% Tween 40 in water), 500, 1000, or 2000 mg/kg folpet for 5 days or the positive control, dimethylhydrazine. The numbers of apoptotic cells and micronuclei in folpet-treated groups were not statistically significant when compared to control animals, even in the presence of diffuse hyperplasia in the mucosal epithelium in the high-dose group. The incidence of nuclear aberrations in the duodenum crypts is summarized in Table 3.

6.2. Duodenal DNA damage

The comet assay, also known as single-cell gel electrophoresis, is a microgel electrophoresis technique that detects DNA damage in individual cells. The damage is represented by an increase of DNA fragments that have migrated out of the cell nucleus in the form of a characteristic streak similar to the tail of a comet. The DNA fragments are generated by DNA double-strand breaks and single-strand breaks. The length and fragment content of the tail is directly proportional to the amount of DNA damage.

Folpet-induced effects on mouse duodenum were evaluated using the comet assay (Clay, 2004). CD-1 female mice (eight per group) were dosed by gavage with the vehicle, 1000 or 2000 mg/kg folpet, or the positive control, N-methyl-N-nitrosourea (MNU) by single gavage administration. Duodenal villi were removed by scraping 2 and 6 hours post-treatment. Fifty crypt cells per slide and 150 crypt cells (where possible) per animal were evaluated for comet formation. Measurements made for each cell included head length, tail length, percent head intensity, and percent tail intensity or tail moment. The primary measure of DNA damage is tail moment (Hartmann et al., 2003; Wiklung and Agurell, 2003). Folpet did not cause statistical or biologically significant increases in head length, tail length, percent head intensity, percent tail intensity, or tail moment compared to controls. There was no DNA damage induced in the duodenum when folpet was administered at high oral doses to the mouse. Data are shown in Table 4.

A series of DNA binding studies were performed in the digestive tract following captan exposure (Pritchard and Lappin, 1991; Provan and Eyton-Jones, 1996; Provan et al., 1995). There was no evidence of captan-DNA binding in the target tissues. A detailed discussion of these studies is found in Section 7.3, DNA binding studies.

7.1. Metabolism and distribution

The metabolic fate of folpet and captan has been reviewed (Edwards et al., 1991; Gordon, 2010; HSDB, 2001; Trochimowicz et al., 2001). When ingested, folpet and captan are relatively stable in the acidic environment of the stomach; however, they are more readily hydrolyzed in the alkaline environment of the duodenum to phthalimide (PI) or tetrahydrophthalimide (THPI), respectively, and to thiophosgene.

The distribution of captan and its degradates was determined in different segments of the gastrointestinal tract of mice. CD-1 male mice (30 per exposure level) were administered [1,2-¹⁴C]cyclohexene-labeled captan in their diets at concentrations of 0, 400, or 3000 ppm for varying lengths of time (Provan and Eyton-Jones, 1996). Radioactivity was measured in the contents and the epithelial tissue of the different segments of the gastrointestinal tract; segments studied were the stomach, duodenum, 10-cm segments of the rest of the small intestine, cecum, and the cecum to anus segment. A low, steady-state concentration of radiolabel was observed along the entire small intestine in both the 400 ppm and 3000 ppm groups. The radiolabel was associated with the duodenal contents rather than the epithelial tissues and did not accumulate with time. Parent captan was detectable only in the stomach of 3/6 mice receiving the 3000 ppm diets; in mice receiving 400 ppm, only captan degradates were found. Consistent with the rapid degradation of captan observed in the stomach, only THPI and its metabolites were found in duodenum, blood, and urine.

7.2. Interactions with thiols

As was briefly discussed in Sections 2 through 6, glutathione and other thiol-containing molecules play an essential role in the inactivation of folpet and captan (Davidek and Siefert, 1975; Gordon et al., 2001). The role of thiols was noted when folpet and captan were tested for genotoxicity in Salmonella and E. coli repair-deficient strains (Moriya et al., 1978). It was also demonstrated, using the S. typhimurium tester strains, that captan was degraded in rat or human blood during a 45-minute incubation (Ficsor et al., 1977). Rat plasma inactivated captan but not as readily as whole blood. This observation correlated with higher amounts of glutathione in whole blood compared to plasma (Michelet et al., 1995). Captan treatment in V79 cells resulted in a 23% decrease in non-protein sulfhydryl groups, mainly reduced glutathione (GSH) (Rahden-Staron et al., 1994). It can be concluded from these studies that in the presence of thiol-containing systems, captan is degraded and subsequent mutagenic activity in vitro is diminished or eliminated.

If folpet or captan remains intact long enough to enter the systemic circulation, each will be quickly degraded by thiols in the blood. Studies have shown that thiophosgene also reacts rapidly with other functional groups such as amines, amides, imides, and alcohols (Sharma, 1978, 1986; Tilles, 1966) and EPA has concluded that thiophosgene is so labile that residues after oral ingestion of captan are not quantifiable (US EPA, 1999b).

Concentrations of GSH in the cytoplasm of most animal tissues range from 0.1 to 10 mM (Aukerman et al., 1984; Pastore et al., 2001) and thus are in sufficient quantity to deactivate folpet, captan, and thiophosgene. Furthermore, the cytoplasm of most animal cells contains high levels of GSH S-transferases, which regenerate GSH, resulting in more available GSH for the reactions with reactive chemicals (Levy et al., 1993).

From a toxicological standpoint, the rapid nonenzymatic reactions of folpet, captan, and thiophosgene with tissue and blood thiols result in the effective elimination of these compounds.

GSH levels have been evaluated following captan and folpet treatment in rats. Single oral bolus doses of captan result in an early decrease in GSH in the duodenum (measured in minutes) followed by a rebound increase in GSH levels (Katz et al., 1982). In mice, continued administration of captan in the feed resulted in a higher sustained GSH level (Miaullis et al., 1980).

More elaborate studies of the role of GSH have been conducted with folpet (Chasseaud, 1991). Because folpet shares a common mechanism of toxicity with captan with regard to duodenal tumor induction in mice (along with in vitro mutagenicity), these data help to further understand the likely effects of captan on GSH levels. Folpet was administered as a bolus dose at 7.6, 72, and 668 mg/kg to CD-1 male mice, and GSH levels were measured in the stomach, duodenum, jejunum, ileum, and liver ranging from 0.5 to 24 hours post-treatment (Chasseaud, 1991). Depletion of GSH was observed 0.5 hours post-treatment in the mid- and high-dosage groups in the duodenum, jejunum, and ileum (Table 5). The depletion was the greatest in the duodenum and jejunum. It was still depleted in the duodenum 1 hour post-exposure. After 2 hours, the levels of GSH had rebounded to within control levels in all tissues and at all doses with the exception of the liver. Six hours after exposure, the GSH levels statistically significantly increased in the duodenum, beginning at the low dose and increased in a dose-responsive manner. In the mid- and high-dose groups, GSH levels in the jejunum and ileum were significantly increased. By 24 hours, at the mid-dose, GSH levels in all tissues were declining but still above control values. In the high-dose groups, the level of GSH in the gastrointestinal tract was higher at 24 hours than at 6 hours. No effects, either increases or decreases in GSH, were seen in the stomach at any dose or time period. These data strongly suggest that folpet-GSH interaction occurs in the small intestine following oral administration. It also implies that as the folpet, captan, or thiophosgene move through the intestinal tract, there is time for these products to be deactivated by GSH and other thiols, unless capacity is overwhelmed by very large oral doses.

The liver continued to show decreased levels of GSH at the high dose at 24 hours post-treatment, whereas all the other tissues were showing rebounding effects at 24 hours. This does not likely represent binding of the GSH in the liver as much as the liver being the major source of circulating GSH (Meister, 1988).

7.3. DNA binding studies

Studies with captan have been conducted with the ¹⁴C and ³⁵S labels on the trichloromethylthio side chain. Both label locations are not ideal: subsequent to degradation, the ¹⁴C enters the one-carbon pool as ¹⁴CO₂ and the ³⁵S undergoes sulfur exchange. This has the potential to result in the appearance of overall binding when it could be the incorporation of these free available radiolabels into macromolecules.

Captan was reported to bind to nucleotides to produce a 7-(trichloromethylsulphenyl) guanine adduct when added to both purine and pyrimidine nucleotides, which were dissolved in saline (Anderson and Rosenkranz, 1974). When analyzed by paper chromatography, only the migration of deoxyguanosine was affected. The reaction product, identified by spectral properties, was presumed to be the 7-(trichloromethylsulphenyl)-guanine adduct. To confirm their in vitro results, ³⁵S-captan was administered to a single mouse. The DNA reaction product was reported to be the same presumptive 7-guanine adduct (administered dose, specific activity, and tissue source of DNA were not provided) (Anderson and Rosenkranz, 1974).

Captan was reported to form DNA adducts in vitro when treated in cell-free systems (Snyder, 1992). When 3 mg of ¹⁴C-captan (methyl labeled) was reacted with 4 or 4.6 mg herring sperm DNA for 1 hour at 37°C in 7.5 ml of buffer, DNA adducts were detected at an unusually high level of binding. Excess radiolabeled captan was not fully removed prior to the analysis of DNA binding, confounding the interpretation of results. Standards were not available for confirmation of DNA adducts.

Studies in Osborne-Mendel rats and CD-1 mice fed 300 and 1600 mg/kg ¹⁴C-captan (trichloromethylthio labeled) investigated the ability of captan to interact with DNA (Selsky, 1981). Insufficient radioactivity was present for analysis. An additional dose of 156 mg/kg containing a higher specific activity of radiolabeled captan was administered to mice. Although there was an association of radiolabel with DNA of mice, dialysis experiments showed the majority of the detected radioactivity was noncovalently bound.

A series of DNA binding studies were performed in the digestive tract following captan exposure (Pritchard and Lappin, 1991; Provan and Eyton-Jones, 1996; Provan et al., 1995). Captan was evaluated for the ability to bind to DNA in the stomach, duodenum, jejunum, liver, and bone marrow of CD-1 mice exposed to either ³⁵S- or ¹⁴C- (methyl position) labeled captan (Pritchard and Lappin, 1991). Presumed DNA binding was observed in the bone marrow, liver, stomach, jejunum, and duodenum. No attempt was made to digest the DNA to confirm the presence of captan-DNA adducts. Subsequent purification of the DNA by cesium chloride gradient separation showed the radioactivity was not covalently bound to the DNA but was associated with a material co-purified with the DNA. These results were similar to those previously reported (Selsky, 1981).

To better understand the nature of this “bound” co-eluting material, the elution patterns of DNA from CD-1 mice fed either ³⁵S-captan and ³⁵S-N-acetylcysteine or two thiazolidine derivatives were compared (Provan et al., 1995). To elucidate the nature of the association of radiolabel with DNA, cesium chloride elution patterns from the liver and duodenum were analyzed. Although the bound material from captan or thiol containing products co-eluted in different positions on the cesium chloride gradient, none was found to co-elute with DNA. It was concluded the DNA binding observed in the GI tract after exposure to captan was due to the incorporation of free sulfur into protein or sulfur-containing macromolecules that was contaminating the DNA preparations. Thus, there was no evidence of captan-DNA binding. In a follow-up study on the disposition of (¹⁴C-cyclohexene-labeled) captan through the digestive tract, radiolabel was found to be associated with the contents of the GI tract and not the tissue (Provan and Eyton-Jones, 1996).

DNA-adduct formation has been noted in vitro (Anderson and Rosenkranz, 1974), giving a presumptive 7-(trichloromethylsulphenyl)-guanine adduct, but DNA-adduct formation in vivo following ¹⁴C- or ³⁵S-captan administration has not been convincingly demonstrated (Pritchard and Lappin, 1991; Provan and Eyton-Jones, 1996; Provan et al., 1995).

7.4. DNA-specific proteins

Under specific in vivo administration conditions (i.p. injection), folpet and captan can bind to histones and interfere with DNA unwinding and renaturation (Couch and Siegel, 1977; Couch et al., 1977; Grilli et al., 1995).

Folpet is capable of inducing DNA damage in rats when dosed by i.p. at 1/3 to 1/6 the LD₅₀ (Grilli et al., 1995). Folpet induced single-strand breaks, demonstrated by DNA unwinding as measured by fluorescence comparing rates of renaturation of DNA after physical denaturation. Twenty years earlier, it was demonstrated that the resultant binding of folpet and captan to the lysine-rich histones in vivo (rat liver, i.p. dosed) and in vitro (calf thymus DNA) changed the configuration of the histones, which altered the ability of the nuclear histones to stabilize DNA structure (Couch and Siegel, 1977; Couch et al., 1977). Interestingly, the binding to histones was greater at pH 9 and could be washed out by the addition of dithiothreitol (Couch and Siegel, 1977). Binding to histones also altered the extraction characteristics of the nuclear protein fractions. The DNA strand break results (Grilli et al., 1995), when interpreted in light of reported binding of folpet to histones that directly affects the ability of the DNA to renature are consistent with the results of Couch et al. (1977). Secondly, animals were dosed by i.p. injection, which increases bioavailability of folpet and limits the neutralizing actions of cellular thiols. Because i.p. injection bypasses the normal routes of exposure, it does not simulate anticipated human exposure.

DNA polymerases, similar to histones, are integrally linked to DNA integrity. In vitro biochemical investigations have demonstrated that captan inhibits DNA polymerase by competing for the same site as the DNA while not interfering with the fidelity of the DNA polymerase I copy of the DNA template (Dillwith and Lewis, 1982). Using captan as an inhibitor of viral reverse transcriptase, the differential effects of captan on DNA polymerase and RNase H activity could be determined (Freeman-Wittig et al., 1986). RNase H activity was 10-fold more sensitive to captan than either DNA-dependent or RNA-dependent DNA polymerase. Based on the observation that dithiothreitol prevented captan inhibition, it was concluded that the trichloromethylthio moiety of captan was involved in the inhibitory action (Freeman-Wittig et al., 1986). Captan was subsequently found to be active at but not limited to the nucleoside triphosphate binding site and DNA binding site of the RNA polymerase (Luo and Lewis, 1992).

A number of in vitro studies have been conducted with the goal of elucidating key elements associated with folpet and captan’s mutagenicity. The reactivity of folpet, captan, and thiophosgene is such that within an in vitro environment reactions will occur with available thiols. The in vitro environment is distinct from the intact animal. Although these mechanistic studies speak to possible modes of mutagenic action, they are not performed in systems comparable to the in vivo setting where captan and folpet mutagenicity is absent.