Genetic toxicology of phthalate esters: mutagenic and other genotoxic effects.

The effects of DEHP on sperm morphology and on peripheral blood micronuclei were studied for 12 weeks following five subacute IP injections of DEHP at 1/6, 1/12, and 1/60 of the LD50 per day. Sperm morphology was examined in both adult mice and rats, while peripheral blood micronuclei were scored in mice up to 4 weeks after treatment. In mice, DEHP at 1/6 LD50 significantly depressed body weight gain for up to 12 weeks after treatment, and reduced epididymal sperm number by 4 weeks. Numbers of morphologically abnormal sperm did not differ from controls in the 12 weeks following treatment. In addition, DEHP did not increase the numbers of peripheral blood micronuclei. Studies in the rat indicated that exposure to doses of 1/6 and 1/12 of the LD50 per day of DEHP resulted in a reduced gain in body weight compared to controls. Testis weight, sperm number, and numbers of morphologically abnormal sperm were unaffected by DEHP following treatment. In separate experiments, DEHP did not induce sister chromatid exchange (SCE) or DNA damage in Chinese hamster ovary (CHO) cells. Although DEHP is known to cause testicular atrophy in rats and to a lesser extent in mice, it did not cause an increase in abnormal sperm in either species. Together with the CHO and micronucleus data, these findings suggest that DEHP has a low probability of causing genetic damage capable of being transmitted through the male germ line.


Introduction
A valid decision on the genotoxicity of phthalates, as with any chemical substance, requires data from a carefully chosen battery of tests. Such a battery would ideally contain tests for the two basic types of genetic effects (i.e., gene mutations and chromosomal mutations) with redundancy in terms of cell type or organisms where practicable. In terms of making any decisions concerning potential hazard to human beings, it is imperative to recognize that the results of in vitro tests can only demonstrate the potential to cause the critical effects in vivo. Thus, the actual confirmation of any in vitro effects requires the conduct of the appropriate in vivo tests. Table 1 is a representative, but not necessarily exhaustive, summary of studies which have been reported in the scientific literature on the genotoxicity of phthalates. While it is difficult to generalize about a broad chemical class, certain observations can be made. The in vitro and in vivo data are equivocal and give no clearcut indication of genotoxicity. Where redundancy in endpoints with the same chemical does exist, there is often a lack of agreement. Discrepancies in the in vivo tests do not seem to be due to differences in the route of admin-*Mutagenesis Section, Environmental Health Directorate, Department of National Health and Welfare, Environmental Health Centre, Ottawa, Canada, KlA OL2. tPresent address: Toxicological Evaluation Division, Foods Directorate, Department of National Health and Welfare, Ottawa, Canada. istration, since both chromosome aberrations and dominant lethality are observed after topical, subcutaneous, or intraperitoneal injection, as well as following oral administration.
There are extensive data on a number of other indicators of toxicity, many of which are discussed elsewhere in this issue. Examples of some of these effects which relate to the purpose of this paper are shown in Table 2. Apart from the indication of liver tumors and teratogenicity, which may or may not be related to the genotoxic effects observed, there is clear evidence of cytotoxicity both in vitro and in vivo; the most important perhaps being testicular cell damage (Table 3). Such effects appear to be related to the age and species as well as to the route of administration.
Because the existing data suggested a genotoxic effect and the clear evidence of testicular damage, we reasoned that one of the most appropriate tests for further study of the potential genotoxic effects of phthalic acid esters would be the sperm abnormality assay (39). A number of carcinogens and mutagens are known to induce abnormal sperm morphology, and some abnormalities are inherited by the offspring of treated males (39,40). However, in addition, increased temperature and food deprivation are also known to induce abnormally shaped sperm (41,42). We have conducted a study on both mice and rats using DEHP as a representative phthalate. In order to assist in the determination of the cause of any al. (14) aAbbreviations for Tables: DEHP,   effects of sperm morphology in the mouse study, circulating red blood cells were assayed for the presence of micronuclei (43), the products of a subset of chromosomal damage. In separate in vitro experiments with CHO cells, the effect of DEHP on the formation of SCE and the induction of alkaline sucrose gradient-detectable DNA lesions were also determined.

Methods and Materials
Mice of hybrid strain B6C3F1 and Sprague-Dawley strain rats were obtained from Charles River Breeders (Lasalle, Quebec). Mice were 6 to 8 weeks old and weighed 18 to 25 g when treated, while rats of the same age weighed 300 to 375 g when injected. Subacute IP injections of DEHP (Cat. #D20,115-4; Aldrich, Milwaukee, WI) in olive oil (#3016; Eastman Kodak, Rochester, NY; or Fisher Scientific, Fair Lawn, NJ) were administered to each animal, along with untouched and solvent negative controls, and MMS (methyl methanesulfonate; #7936, Eastman Kodak, Rochester, NY), in normal saline as a positive control. The groups consisted of three animals (rats) or five animals (mice) per control group per week and five animals per treatment group per week. Animals were injected on five consecutive days with maximum volumes of 0.2 mL/mouse and 2.5mL/rat. Doses of DEHP administered were as follows: mice, 6.0g/kg/day (1/6LD50),3.0g/kg/day(¼/12LD50),0.6g/kg/day('/60LD50);rats; 5.2 g/kg/day (1/6 LD50), 2.6 g/kg/day (/12 LD50), 0.52 g/kg/ day (1/60 LD50). MMS was administered at 75 mg/kg/day for the positive control mice. Since several mice were lost during the course of the study, rats were exposed to 75 mg/kg/day on the first injection day, but to 37.5 mg/kg/ day on the remaining four treatment days. The "week 0" group of animals was sacrificed following the injections on the 5th day of treatment and every 2 weeks thereafter for up to 12 weeks. Mice were sacrificed by cervical dislocation and rats by CO2 asphyxiation.
Body weights and testes weights were recorded at sacrifice, and epididymal sperm smears stained in 1% eosin Y were prepared according to Wyrobek and Bruce (39), except that rat epididymes were minced in 10 mL of PBS per cauda. Sperm number was obtained using a hemacytometer. 500 sperm or more were scored for abnormalities per animal. Data are presented as the means for each treatment group, and analyzed using the G-test  (38) or Students's t-test (44). In the mouse experiment, 1000 monochromatic erythrocytes/animal were scored for the presence of micronuclei from the 0, 2, and 4 week groups according to the method described by Schlegel and MacGregor (43). SCE and alkaline sucrose gradient-detectable DNA damage was detected in CHO cells as described previously (45). Cells were treated for 1 hour in minimal MEM minus calf serum with DEHP dissolved in DMSO.

DNA Damage and SCE in CHO Cells
No increase in the level of SCE (Table 4) or in DNA damage detectable by alkaline sucrose gradient analysis (Table 5) was observed after treatment of CHO cells.

Body Weights and Gross Changes
Mice exposed to the highest dose of DEHP (6.0 g/kg/ day) showed significant reductions in body weight gain (t-test; p < 0.05) compared with other treatment groups and with control groups at the second sacrifice and continuing for the duration of the study (p < 0.01) (Fig. la). A decline in body weight gain was also seen in rats for 12 weeks after exposure to 5.25 g/kg/day (p < 0.01) or 2.6 g/kg/day (p < 0.05) (Fig. lb). Both species showed extensive whitish deposits in the diaphragm, abdominal wall and in the scrotal fat which persisted throughout the study. In the later weeks, the scrotal sac was filled with a milky colored fluid in some of the high and middle dose animals from both species.
Peripheral Blood Micronuclei in Mice Table 6 presents the results obtained when mice exposed to DEHP were sacrificed 0, 2, or 4 weeks after the final injection. The positive control, MMS, displayed Testis Weight Mice exposed to 6.0 g/kg DEHP per day showed a decline in testis weight compared to the solvent or un- MMS, on the other hand, caused a significant decrease in testis weight 2 weeks (p < 0.001) and 4 weeks (p < 0.01) after treatment, which then returned to within normal range (Fig. la). No such decline in testis weight (p > 0.05) was seen for any of the treatment groups in the rat as indicated in Figure la. Sperm Number Mice exposed to DEHP at 6.0 g/kg/day showed a significant reduction in sperm number per epididymis from 4 weeks following exposure to the end of the study (p < 0.01) (Fig. la). MMS, on the other hand, showed a reversible decline in sperm number, that attained control values by the final weeks (Fig. la). By 10 weeks after exposure, sperm numbers in the rats were not significantly reduced (p > 0.05) at any of the doses of DEHP (Fig. lb). A decline in sperm number was not significant (p > 0.05) following exposure to MMS in the rats.  bThe numbers in parentheses indicate the number of animals analyzed.

Sperm Morphology
The results of the sperm abnormality studies are shown in Figure 2. In general, DEHP did not exert striking increases in the levels of abnormally shaped sperm in either species, over the 12-week period examined.

Discussion
In this investigation, two in vitro assays were performed with CHO cells to assist in the determination of genotoxic potential; alkaline sucrose gradient analysis, which detects a spectrum of different types of DNA dam- age, and the SCE assay, which must be regarded as an indicator of DNA damage. Accordingly, in the interpretation of these data, it must be realized that neither of their effects represent mutations per se. Furthermore, as these are in vitro tests, their results must be tempered with the wisdom of experimental data from whole animal models. The lack of an effect in the peripheral blood micronucleus assay in mice complements the negative findings of the in vitro assays. The sperm abnormality assay performed both in mice and rats represents a departure from the protocol established by Wyrobek and Bruce (39,46), where a single sacrifice time of 5 weeks following the first treatment was employed. In the present study, animals were sacrificed every 2 weeks up to 12 weeks following the last dose. The advantage of this protocol can be seen by the observation that the positive control (MMS) produced the maximum response at 2 weeks in these studies rather than at approximately 4 weeks. These data demonstrate that optimal responses can vary from one agent to another, and point out the necessity of using multiple sacrifice times.
No sperm abnormalities were induced by DEHP in either mice or rats, even though toxic effects were seen in the testis (i.e., reduced testis weight and sperm count in mice) and body weight gains were reduced on both species. To the extent that abnormal sperm morphology is an indicator of genotoxicity (47), it would appear that DEHP would have little probability of causing inherited genetic damage in the offspring of treated animals. The other negative data in this study support this conclusion. Furthermore, although toxic effects other than mutagenicity are known to induce sperm abnormalities (41,42), such appears not to be the case with DEHP The genotoxicity that is observed with phthalates by others (3,(5)(6)(7)12,13,15) occurs at very high concentrations. Accordingly, the influences of limited solubility, impurities, or an indirect genotoxic effect resulting from primary toxicity should be considered seriously as a causative factor in these cases. For instance, in Seed's study (6) of phthalates using bacterial strains of Salmonella typhimurium, no absolute increases in the mutant frequencies in S. typhimurium treated with a number of phthalates were observed; increased mutagenicity was obtained only when data were corrected for viability.
On the other hand, no genotoxicity was observed in the present experiments, even at very high, toxic levels of DEHP. This suggests that toxic effects are not the only determinants influencing the genotoxic effects of phthalates in other studies. The contradictory results reported for some assay systems (1,(3)(4)(5)(6)(13)(14)(15) suggest that variability in general experimental conditions, culture media, cell types or species could account for some of the reported ambiguities.