Trichloroethylene,Trichloroacetic Acid,and Dichloroacetic Acid: Do They Affect Eye Development in the Sprague-Dawley Rat?

Abstract

Maternal exposure to high doses of trichloroethylene (TCE) and its oxidative metabolites, trichloroacetic acid (TCA) and dichloroacetic acid (DCA), has been implicated in eye malformations in fetal rats, primarily micro-/anophthalmia. Subsequent to a cardiac teratology study of these compounds (Fisher et al. 2001, Int. J. Toxicol. 20:257–267), their potential to induce ocular malformations was examined in a subset of the same experimental animals. Pregnant, Sprague-Dawley Crl:CDR BR rats were orally treated on gestation days (GDs) 6 to 15 with bolus doses of either TCE (500 mg/kg/day), TCA (300 mg/kg/day), DCA (300 mg/kg/day), or all-trans retinoic acid (RA; 15 mg/kg/day). The heads of GD 21 fetuses were not only examined grossly for external malformations, but were sectioned using a modified Wilson’s technique and subjected to computerized morphometry that allowed for the quantification of lens area, globe area, medial canthus distance, and interocular distance. Gross ocular malformations were essentially absent in all treatment groups except for the RA group in which 26% of fetuses exhibited micro-/anophthalmia. Using the litter as the experimental unit of analysis, lens area, globe area, and interocular distance were statistically significantly reduced in the DCA treatment group. Statistically significant reductions in lens and globe areas also occurred in the RA treatment group, all four ocular measures were reduced in the TCA treatment group but none significantly so, and TCE was without effect. Because DCA, TCA, and RA treatments were associated with significant reductions in fetal body weight (bw), data were also statistically analyzed after bw adjustment. Doing so dramatically altered the results of treatment group comparisons, but the severity of bw reduction and the degree of change in ocular measures did not always correlate. This suggests that bw reduction may not be an adequate explanation for all the changes observed in ocular measures. Thus, it is unclear whether DCA specifically disrupted ocular development even under these provocative exposure conditions. Clearly, however, if TCE is capable of disrupting ocular development in the Sprague-Dawley rat, a higher dose than that employed in the present study is required.

Trichloroethylene (TCE), trichloroacetic acid (TCA), and dichloroacetic acid (DCA) are commonly found in finished drinking water or drinking water sources such as groundwater and surface water. The presence of TCE is the result of its historical use as a degreaser, whereas TCA and DCA are by-products of the chlorination of municipal water (Coleman et al. 1984). TCA and DCA are also products of TCE’s oxidative metabolism. Although the toxicity of TCE has been scrutinized for years, it has recently been reassessed by the U.S. Environmental Protection Agency (U.S. EPA 2001) and is currently under review by an expert panel convened by the National Academy of Sciences. Among the noncancer toxicities at issue are ocular malformations in the pups of rats treated during gestation with TCE or its metabolites, TCA and DCA. In particular, the study of TCE by Narotsky et al. (1995) that reported a dose-dependent increase in the percentage of Fischer-344 rat pups with eye defects is thought to provide the most robust data set for quantitative risk assessment (Barton and Das 1996; Barton and Clewell 2000). It is one of many oral studies cited by U.S. EPA as supporting the development of a reference dose for TCE. On average, the study reported eye defects in 1.0%, 5.3%, 9.2%, 11.7% and 30.0% of pups per litter at 0, 475, 633, 844, and 1125^* mg/kg/day TCE, respectively (^*denotes statistically significant). Micro-/anophthalmia was identified by observing a reduction or absence of ocular bulge with subsequent confirmation by eye dissection to examine the size of the orbit. In the present study, however, it was of interest to determine whether treatment that did not result in grossly observable micro-/anophthalmia would nonetheless result in measurable differences in the size of internal ocular structures typically modified in micro-/anophthalmic conditions. Thus, four end points were selected for measurement, including lens and globe areas, medial canthus distance (i.e., distance between medial canthi as measured on the intact head), and interocular distance (i.e., distance between medial globes as measured after sectioning).

MATERIALS AND METHODS

All studies involving live animals were conducted under a program of animal care accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, International, and in accordance with the Guide for the Care and Use of Laboratory Animals, National Research Council (1996).

Pregnant, Sprague-Dawley Crl:CDR (SD) BR rats approximately 3 months of age were dosed on gestational days (GDs) 6 to 15 by oral bolus intubation with either TCE (500 mg/kg/day), TCA (300 mg/kg/day), or DCA (300 mg/kg/day). All-trans retinoic acid (RA), a known ocular teratogen (Sulik et al. 1995), was used as a positive control (15 mg/kg/day). TCE and RA were administered in a soybean oil vehicle, whereas TCA and DCA were administered in ion exchange/reverse osmosis-treated water. Additional methodological details are available in Fisher et al. (2001), a cardiac teratology study of 1677 rat fetuses from 115 dams. For the present study, a subset of the fetuses examined by Fisher and colleagues was randomly selected and processed for ocular examination (i.e., 1185 fetuses [71%] from 108 dams). Reducing the number of fetuses undergoing ocular examination by approximately 30% compared to the cardiac study was deemed necessary given the laborious nature of the assessment technique employed.

Heads from the decapitated GD 21 fetuses were fixed in Bouin’s solution and stored in 70% ethyl alcohol and were examined grossly for external abnormalities. The mandible was then resectioned along a line extending posteriorly from the commissures of the lips of the mouth to a point approximately 1 mm below the ventral demarcation of the ear. This facilitated subsequent manipulations of the samples. Using an Olympus SZ 40 dissection microscope and a Leica Quantimet 570C Image Analysis System, an image of the head was captured and the distance between the medial canthi of the eyes measured. The heads were then hand-sectioned using a modified Wilson’s technique. An initial, vertical cut was made perpendicular to the long axis of the head along a plane visually determined to pass through the center of each ocular globe. A blade guide was used to assist in standardizing the sectioning. Subsequently, two cuts were made parallel and approximately 2 mm anterior and posterior to the initial cut, thus freeing the anterior and posterior facets of the original cut for further examination. The freed sections were placed in a drop of water on a glass microscope slide and digital images were obtained using a Sony (DXC-760MD) CCD camera for future assessment. If the cut through either of the globes was markedly off-center, the sample was excluded from morphometric assessment. The term “markedly off-center” refers to a cut that captured only a segment of globe that was clearly too small for consideration as a representative sample. Alternatively, the cut resulted in a globe that was clearly elliptical due to oblique orientation of the fetal head relative to the sectioning blade. In addition, an a priori decision was made to exclude all micro-/anophthalmic and exencephalic fetuses from morphometric assessment and statistical analysis, respectively.

The following measurements on head sections were determined: interocular distance (i.e., distance between medial globes as measured after sectioning), total area of the cut surface (i.e., cross-sectional area of the head with the mandible removed), areas of left and right lenses, and areas of left and right globes. For statistical comparisons, only the largest measurement for each feature was used from the two digitized sections of each fetal head. Also, left and right lens areas and left and right globe areas were averaged, as mean lens area of the smaller eye was 98.1% that of the larger eye and mean globe area of the smaller eye was 96.7% that of the larger eye. Calibration of the morphometric system was done prior to each measurement session using a standard microscopic metric rule (American Optics). The litter was used as the experimental unit of analysis. The Kruskal-Wallis one-way layout test (chi-square approximation with 5 degrees of freedom) was used to determine the significance of differences among the six treatment groups. The Wilcoxon rank sum test (two-tailed) was used to determine the significance of paired comparisons. A p value of ≤.05 was considered to be statistically significant.

RESULTS

Despite the a priori decision to exclude exencephalic fetuses from statistical analysis, they were nonetheless subjected to computerized morphometry and showed significant reductions in lens (18%) and globe areas (14%) compared to fetuses with no such malformation. As seen in Table 1, fetuses with exencephaly or micro-/anophthalmia were almost exclusively isolated to the RA treatment group. Of the 1185 fetuses selected for ocular examination, 574 (48%) had clean cuts through the middle of the eyes deemed necessary for further evaluation. Excluding the RA treatment group, the percentages of fetuses deemed suitable for morphometric evaluation ranged from 41% to 56% (Table 2). Mean fetal body weight (bw) was statistically significantly reduced in the TCA, DCA, and RA treatment groups. Mean maternal bw was also reduced in these same treatment groups, but not significantly so (Table 3).

Mean fetal lens and globe areas were statistically significantly reduced in the DCA and RA treatment groups (Table 4). These results are displayed graphically in Figure 1, with and without adjustment for bw and cross-sectional area of the fetal head. Adjusting for bw alters dramatically the outcome of treatment group comparisons, as bw reductions were more severe than reductions in lens and globe areas when examined as a percentage of control. When mean lens and globe areas were adjusted to the cross-sectional area of the fetal head, none of the treatment groups was significantly different from its corresponding control. Note that the DCA and RA treatment groups are similar in fetal lens and globe areas, despite a dramatic difference in fetal bw.

As with mean lens and globe areas, mean medial canthus and interocular distances were reduced among the TCA and DCA treatment groups, with statistical significance achieved in the case of DCA and interocular distance only (Table 5). These results are displayed graphically in Figure 2, with and without bw adjustment. Again, adjusting for bw alters dramatically the outcome of treatment group comparisons as bw reductions were more severe than reductions in medial canthus and interocular distances when examined as a percentage of control. Note that RA treatment did not reduce interocular distance, despite dramatically reducing fetal bw.

DISCUSSION

Despite a prolonged and intensive research effort to better characterize its toxicity, very few studies of TCE (or TCA and DCA) are informative on the issue of ocular development. As previously mentioned, Narotsky et al. (1995) reported dose-dependent increases in the percentage of Fischer-344 rat pups with eye defects at all TCE doses tested (475, 633, 844, and 1125^* mg/kg/day; ^*denotes statistically significant), despite no significant decreases in pup weight. Maternal weight gain during gestation was significantly decreased at all doses, however. These researchers diagnosed micro-/anophthalmia by observing the reduction or absence of eye bulge with subsequent confirmation by eye dissection to examine the size of the orbit. No information on how orbital size was assessed was provided. Narotsky and Kavlock (1995) made a similar report of micro-/anophthalmia at 1125 and 1500 mg/kg/day TCE in Fischer-344 rats, but reported significant reductions in pup weight on postnatal day 1. As for TCE’s metabolites, Smith et al. (1989, 1992) administered TCA (330, 800, 1200^*, and 1800^* mg/kg/day; ^*denotes statistically significant) or DCA (14, 140, 400, 900^* , 1400^* , 1900^*, and 2400 mg/kg/day; ^*denotes statistically significant) to Long-Evans rats on GDs 6 to 15 and noted statistically significant increases in soft tissue and skeletal orbital malformations, but only at doses that significantly reduced fetal bw. These results indicate that TCE and its oxidative metabolites can disrupt ocular development in rats at extreme doses. Yet, effects have been statistically significant only under conditions that impact maternal or fetal bw and only at doses at least two- to four fold higher than those used in the present study.

To date, relatively few teratology studies have relied upon a means of assessing the eye other than gross observation, including those studies of TCE and its metabolites discussed above. Yet, like most terata, eye anomalies can occur on a continuum from severe and visible on gross inspection to subtle and detectable only with mechanical or computerized measurement. As noted by Sulik et al. (1995) in their teratology evaluation of low doses of RA in mice, gross observation does not allow for subtle ocular malformations to be taken into account. Ozeki and Shirai (1997) also point out in their investigation of RA in the mouse that distinguishing anophthalmia from microphthalmia by gross examination using a stereomicroscope is often very difficult. Clearly, the use of morphometry in ocular studies can be of value. For instance, Stull and Walker (2000) used an image analysis system to demonstrate that perturbations in RA availability alter the size and shape of the retina in mice. Morphometry has also been effectively used in fetal rats to assess optic nerve hypoplasia secondary to fetal alcohol syndrome (Pinazo-Duran et al. 1997) and to detect mild degrees of microphthalmia by measuring the perimeter of eyes after maternal exposure to the fungicide benomyl and/or protein deprivation (Hoogenboom et al. 1991). Interestingly, Hoogenboom and colleagues mention that the results of their latest study differ from those of previous experimental efforts with benomyl and attribute this to a previous reliance on macroscopic rather than microscopic examinations to detect abnormalities. Similar to the bw adjustment made in the present study, these researchers examined ocular perimeter to bw ratios in order to determine the relationship of body size reduction to ocular perimeter reduction. The perimeters of eyes from exposed pups were significantly smaller than those of control pups, but when these size changes in the eyes were related to changes in overall bw, no significant decrease in eye size was found. Finally, lens and orbital measurements at various gestational ages have also been made in humans in an effort to collect normative data for fetal growth and development, data that are potentially helpful in detecting fetal ocular anomalies (Goldstein et al. 1998; Dilmen et al. 2002).

Unquestionably, it is less than optimal that only a subset of the fetuses examined for cardiac malformations by Fisher et al. (2001) underwent ocular assessment in the present study. However, of the 1185 fetuses selected for ocular examination, about one-half were judged to have been hand-sectioned through the eyes in such a way that their inclusion could have potentially resulted in misleading statistical comparisons. This is particularly true given the very small quantitative differences that exist between treatment groups, even among some of those that statistically significantly differ. Despite the difference in the number of fetuses examined in the cardiac and ocular studies, the number of dams represented was about the same (TCE: 18 versus 20; TCA: 17 versus 19; DCA: 20 versus 20; RA: 8 versus 12; soybean oil: 21 versus 21; water: 19 versus 19). Thus, entire litters were not typically excluded from analysis. In addition, compared to the number of fetuses examined for cardiac malformations, the percentage of fetuses assessed for ocular malformations was quite comparable across treatment groups (TCE: 28%; TCA: 32%; DCA: 27%; RA: 19%; soybean oil: 30%; water: 48%). It is also noteworthy that excluding improperly cut histological sections from further analysis is not unprecedented. For instance, Hoogenboom et al. (1991) made ocular perimeter measurements on serial sections of fetal rats and discarded those eyes in which sections did not transect the optic nerve or in which sections were clearly elliptical due to oblique orientation. Therefore, although considerably smaller in number, the subset of fetuses on which ocular measures were made is believed to be representative of those originally assessed for cardiac malformations.

Unlike previous studies with TCE and its metabolites, the present study sought to detect mild degrees of microphthalmia not detectable by gross evaluation. It is supportive of an association between DCA and significant decreases in fetal lens and globe areas and interocular distance. Given the small changes observed in these ocular measures, however, the specificity of the association remains unclear as it may simply be a function of decreased fetal bw or size. Clearly, TCE did not affect any ocular measure. This suggests that the dose of TCE was too low to produce an effect itself or to result in a sufficient amount of metabolically derived DCA or TCA to do so. Although it seems unlikely that humans will encounter doses of TCE, DCA, or TCA as high as those used in the present study, the risk for subtle ocular changes such as those reported herein remains unknown, as does their clinical significance, if any.