The liver and testes were the primary target organs of DBA-induced toxicity in rats and mice. In the liver, exposure to DBA produced increased incidences or severity of hepatocellular cytoplasmic vacuolization. Testicular lesions induced in rats in the 2-week and 3-month studies were similar to those described by Linder et al. (1994b
and were characterized by a delay in spermiation with retention of step 19 spermatids in the seminiferous epithelium adjacent to the lumen of stage IX and stage X tubules of the spermatogenic cycle. These changes were accompanied by the presence of atypical residual bodies. The formation of atypical residual bodies in rats exposed to DBA was suggested to be a result of impaired degradative function in Sertoli cells (Linder et al., 1997
). The lowest concentration of DBA showing testicular effects in the 3-month study was 500 mg/L, equivalent to a mean daily dose of 40 mg/kg. Decreased testicular weight and testicular atrophy were observed in the 2000 mg/L group (166 mg/kg). Linder et al. (1997)
observed altered spermiation at 10 mg/kg or higher doses in Sprague-Dawley rats exposed to DBA by gavage for 31 or 79 days and seminiferous tubule atrophy at 250 mg/kg/day. The difference in dose-response between the present study and that of Linder et al. (1997)
might be due differences in rat strains used in the two studies or different methods of exposure (drinking water versus gavage). Sperm morphology evaluations in the present study showed reductions in epididymal sperm counts, sperm motility, and spermatid heads at 2000 mg/L (166 mg/kg) but not at 1000 mg/L (90 mg/kg). Hypospermia was also diagnosed in the epididymis of rats in the present study. Linder et al. (1994b)
observed significant reductions in sperm motility and altered sperm morphology after 14 daily gavage doses of 270 mg/kg.
Testicular lesions induced in mice exposed to 1000 or 2000 mg/L were similar to those observed in rats. These lesions, which had not been reported previously in mice, were characterized as spermatid retention and large atypical residual bodies in seminiferous tubules. The atypical residual bodies in mice were generally larger than those observed in rats; however, this effect was detected at a lower average daily dose in rats (40 mg/kg) than in mice (115 mg/kg). As noted above, the formation of these bodies was probably due to impaired degradative functions carried out by Sertoli cells. Evaluations of sperm from exposed mice did not demonstrate any effects on morphology, motility, or epididymal counts. Neither testicular atrophy nor decreases in testicular weight were observed in mice exposed to 2000 mg/L, the concentration of DBA that caused these effects in rats.
In the 2-year studies, neoplastic effects were induced at multiple sites in rats and mice exposed to DBA. The incidence of malignant mesotheliomas in the 1000 mg/L group of male rats (20%) was significantly increased and exceeded the NTP’s historical control rate for mesotheliomas in 2-year drinking water studies (mean historical incidence in male rats: 6.0% ± 4.2%). This malignant lesion was observed at multiple sites throughout the abdominal cavity. Chemically induced mesotheliomas have been observed predominantly in male rats compared to female rats or mice of either sex. Of over 500 carcinogenicity studies reported by the NTP, 16 agents produced positive evidence of chemically induced neoplasms in the mesothelium; of those agents, 15 were active in male rats, two in female rats and mice, and one in male mice. Both mutagenic and non-mutagenic chemicals induced these neoplasms. Phenoxybenzamine was the only chemical that induced neoplasms in the abdominal cavity mesothelium of both sexes of rats and mice (NCI, 1978
). Potassium bromate also induced mesotheliomas of the peritoneum in male F344 rats (Kurokawa et al., 1983
). Based on the above studies, there is no apparent relationship between chemical structure and this neoplastic response.
In a subsequent 2-year study of potassium bromate administered to male F344 rats in drinking water at concentrations ranging from 0.02 to 0.4 g/L, mesothelioma was detected on the tunica vaginalis testis in one animal euthanized after 52 weeks of treatment (0.2 g/L); whereas, after 78 weeks of treatment, mesotheliomas were present on multiple abdominal organs (Wolf et al., 1998
). Based on these findings, Wolf et al. (1998)
suggested that the tunica vaginalis might be the site of origin of bromate-induced mesotheliomas. However, the time-dependent incidence of preneoplastic lesions and mesotheliomas in male rats exposed to potassium bromate was consistent with either the tunica vaginalis or the spleen as the site of origin of these neoplasms (Crosby et al., 2000b
). In the present study of DBA, the earliest death of a male rat with mesothelioma occurred after 73 weeks of exposure. Neoplastic lesions were observed on multiple organs throughout the abdominal cavity in early death animals in the DBA study. Thus, it is not possible at this time to draw a definitive conclusion on the site of origin of the abdominal mesotheliomas.
MCL is a commonly occurring hematopoietic neoplasm in F344 rats. It is a rapidly progressive, lethal neoplastic disease that first develops in the spleen, with infiltrates of neoplastic cells occurring in the liver, lung, lymph nodes, bone marrow, and other organs (Elwell et al., 1996
). A significantly increased incidence and positive trend for MCL was observed in female rats. The incidence in the 1000 mg/L group (44%) far exceeded the historical control rate for 2-year drinking water studies in which female rats were given NTP-2000 diet (mean incidence of 23.5% ± 4.4%). In male rats, the incidence of MCL was increased significantly in the 50 mg/L group (62% versus 34% in controls), and a marginal increase was observed in the 500 mg/L group (48%). Both of these incidences exceeded the historical control rate for 2-year drinking water studies in male rats fed NTP-2000 diet (mean incidence of 31.6 ± 3.3%). There was no difference in the incidence of MCL between controls and high-dose males (1,000 mg/L) (26%). The MCL response in female rats was considered to be related to DBA exposure based on the positive trend, the significantly increased incidence that far exceeded historical rates, and the equivocal evidence of DBA-related MCLs in male rats.
A nonmonotonic dose response does not always occur with chemically induced MCLs in male rats. In most instances in which such a response was observed, survival was reduced in the higher dose groups (e.g., NTP 1988
). However, in the DBA study, survival of male rats was not significantly reduced in any exposure group. Because there is no satisfactory explanation for the significantly increased incidence of MCL in only the low-dose group, the MCL response in male rats was considered to represent equivocal evidence of carcinogenic activity. This conclusion is consistent with that of other 2-year studies in which MCL was increased in the low-dose or mid-dose groups (e.g., NTP 1997
Exposure of mice to DBA produced significant dose-related increases in hepatocellular adenomas or carcinomas in males and females and hepatoblastomas in male mice. Of particular note is that the increase in hepatocellular neoplasms in male mice was significant even at the lowest exposure concentration tested (50 mg/L), which is equivalent to an average daily dose of approximately 5 mg/kg. Administration of DCA in drinking water for two years also induced hepatocellular neoplasms in male B6C3F1 mice (DeAngelo et al., 1991
); however, DCA did not induce hepatocellular neoplasms at doses as low as that of DBA in the current study. The potent liver tumor response observed in this study provides clear evidence of carcinogenic activity for DBA in mice. The increases in hepatocellular neoplasms were not associated with hepatocellular necrosis or regenerative hyperplasia.
DBA also induced a significant increase in alveolar/bronchiolar neoplasms in male mice and a nonsignificant dose-related increase in lung neoplasms in female mice. Also observed in male mice was a dose-related increase in alveolar epithelial hyperplasia, a lesion that is considered to be part of the continuum of proliferative changes in lung neoplasia. Thus, the lung neoplastic response in male mice is likely related to DBA exposures, while the weaker response observed in female mice may have been related to DBA exposure.
The 2-years studies presented here demonstrate that DBA is a multiple organ carcinogen in laboratory animals; the primary sites of tumor induction identified are the abdominal cavity mesothelium of male rats, hematopoietic system in female rats, and the liver and lung of mice. The mechanism(s) of neoplasm induction by DBA or the related compound, DCA, are not known. For DCA, Carter et al. (2003)
suggested that the induction of liver tumors in mice is due to selective growth of a phenotypic cell-type that does not respond to mitoinhibitory homeostatic control mechanisms. Neither hepatocellular necrosis nor regenerative hyperplasia was associated with the development of preneoplastic lesions or liver tumors in DCA-treated or DBA-treated mice. An early increase in hepatocyte proliferation is not likely involved in the mode of action for DBA, because no increases in the hepatocyte DNA labeling index were observed in mice exposed for 26 days (results not shown), and a slight increase that occurred in male F344 rats was not accompanied by an increase in liver tumor response. DNA hypomethylation and increased expression of c-myc and IGF-II genes were suggested as possible early events in the hepatocarcinogenicity of dihaloacetic acids in mice (Pereira et al., 2001
; Tao et al., 2004
). DNA damage due to oxidative stress in the livers of mice exposed to halogenated acetic acids, including DBA (Austin et al., 1996
), may also contribute to the hepatocarcinogenicity of these chemicals. A study of gene expression in immortalized rat peritoneal mesothelial cells incubated with potassium bromate detected increases in oxidative stress responsive genes, as well as changes in genes that regulate DNA repair and cell cycling (Crosby et al., 2000a
). The carcinogenicity of DBA may involve a genotoxic mechanism since this chemical induces DNA damage in E. coli (Giller et al., 1997
), mutations in Salmonella typhimurium
strain TA98 and TA100 with and without metabolic activation (Kargalioglu et al., 2002
), and DNA strand breaks in Chinese hamster ovary cells (Plewa et al., 2002
). In addition, glyoxylate, a metabolite of dihaloacetate biotransformation, is mutagenic in Salmonella TA100, TA102, and TA104 (Sayato et al., 1987
). Thus, it is possible that oxidative stress, DNA damage, and selective growth of preneoplastic cells are involved in the carcinogenesis of DBA.