Results of testing fifteen glycol ethers in a short-term in vivo reproductive toxicity assay.

Fifteen glycol ethers were investigated for their potential to cause adverse reproductive toxic effects using an in vivo mouse screening bioassay. Pregnant mice were orally dosed once per day on days 7 through 14 of gestation at concentrations causing 0 to 41% maternal mortality. Reproductive endpoints included pup survival in utero (percent of live litters/pregnant survivors), pup perinatal and postnatal survival (number of live pups per litter, number of dead pups per litter, and pup survival to 2.5 days of age), and pup body weight statistics (weight at birth and weight at 2.5 days of age). The study was conducted in two phases: a dose range-finding phase using nonpregnant female mice, and a definitive reproductive phase using time-mated mice. The range-finding phase sought to identify, for each chemical, the maternal LD10 as the target dose. However, based upon reproductive phase results, such an exact dose was impractical to achieve. Thus, a range from the LD5 to the LD20 was considered a sufficient challenge dose that would not affect results due to high mortality, i.e., greater than the LD20. Glycol ethers were assigned to groups having different priorities for further testing based upon whether a sufficient challenge dose was administered and the degree of effects recorded for each chemical.(ABSTRACT TRUNCATED AT 250 WORDS)


Introduction
Conventional reproductive testing is expensive, involves complex scheduling, and requires the commit-  (1). For this test, pregnant mice are treated during organogenesis with high doses of the test chemicals. Females are then allowed to deliver their litters and the number of live-born pups, their birth weight, and growth and survival to 2 to 3 days of age are monitored. While the test is not appropriate for labeling a compound as safe or unsafe, it may serve to generate data useful in establishing priorities for conventional testing. It may also be useful for rapidly surveying structure-activity relationships. As part of the NTP evaluation of potential reproductive toxins, four contracts were awarded by NIOSH under which a total of 30 chemicals were tested. Fifteen of the chemicals were glycols or glycol ethers and the results of those tests are summarized here. Of these 15, three contractors tested 4 chemicals each and one contractor tested the remaining 3 (see Table 1). None were tested in more than one laboratory.

Methods
All contractors used CD-1 mice purchased from Charles River Breeding Laboratories, Inc. (Wilmington, MA) throughout these studies. Chemicals were evaluated in two phases: a preliminary dose-finding study in nonpregnant mice followed by the reproductive phase using time-mated females. In both phases chemicals were evaluated in blocks of two to four chemicals with a shared concurrent vehicle control group. Some blocks included only glycols, others included other chemicals. Only data on glycols are reported here. Chemicals were provided to contractors by NIOSH and were tested in blind, with the chemicals identified by an arbitrary code number. Table 1 summarizes the glycols tested and their abbreviations, structural formulas, chemical purity and the laboratories that performed the investigations. diEGdiBE was administered in corn oil due to its insolubility in water. Distilled water served as the vehicle for all other glycols tested. Dosage was by oral gavage in all instances.
The dose range-finding study was conducted at 5 dose levels using ten mice, 6 to 8 weeks old, per treated or control group (except for diEGBE and diEGdiEE where the mice were 60 to 80 days old*). Upon receipt, mice were weighed and marked for individual identification, then formally randomized to treatment groups. Mice were group-housed, five per cage, throughout the range-finding study. Standard laboratory rodent chow and untreated tap water were available ad libitum. Bedding of a type known not to induce microsomal enzymes was changed as needed or at least once per week. Oral doses were administered once daily for 8 consecutive days using a constant dose volume of 10 mL/kg body weight. Body weights were recorded on days 1 and 8 of the dosing period and on days 4 and 8 of an 8-day post-dosing observation period. Group mean or individual body weights taken on the first day of dosing were used to calculate treatment volumes over the entire 8-day dosing period. Survivors were sacrificed immediately following the last weighing on the 8th post-dosing day. All mice that died before that time were necropsied for evidence of dosing error as a cause of death. Based on the results of these dose-finding studies, the estimated LD10 dose was selected for the reproductive phase.
Reproductive studies were conducted in time-mated CD-1 mice, 6 to 8 weeks of age, orally dosed on days 7 to body weights were taken on day 7 of gestation immediately before dosing, on day 18 of gestation, and on day 3 postpartum. The weight of the animal on gestation day 7 was used to calculate the dosage for the entire period. All mice were housed individually throughout the reproductive study. Food, water and bedding were provided as in the range-finding studies except bedding was not changed after gestation day 18.
Females were observed with minimal disturbances twice daily beginning on day 18 of gestation. As soon as possible after litters were delivered (within 12 hr), the number of live and stillborn pups was recorded. Maternal body weight was recorded and all live pups were weighed together. Pups and their dams were then returned to nest boxes and were left undisturbed until 48 hr after the initial weighing, at which time the number of live pups, their total weight, and maternal body weight were again recorded.
Details of the statistical analyses varied from one contractor to another, but generally the procedures were similar. Body weight data were analyzed by analysis of variance. The proportion of surviving pregnant mice that gave birth to viable litters (one or more live-born pups) was evaluated by the Fisher-Irwin exact test. Numbers of live and stillborn pups and percent survival to 2.5 days of age were analyzed by analysis of variance and the Student's t-test.

Results
Tables 2 and 3 present the findings of the postnatal screen of the 15 glycol ethers. Because these chemicals were investigated by four independent laboratories at different times and in some cases concurrently with other chemicals, seven control groups were used (some controls served more than one chemical). Control values were generally consistent across these seven groups. There was only one maternal death in all control groups and that was a result of a gavage error. Reproductive success in the controls ranged from 91 to 100% of all pregnant survivors. The average litter size ranged from 9 to 11 pups per litter. The percent pup postnatal survival (to 2.5 days postpartum) ranged from 98 to 100%. The average pup weight gain over days 1 to 3 postpartum ranged from 0.4 g to 1.1 g, and the average pup birth weight ranged from 1.6 g to 1.7 g.
Nine of fifteen glycols tested affected the viable litter index. Pregnant mice treated with EGME, EGEE, EGdiME, diEGdiME, and triEGdiME produced no viable litters. Mice treated with EG, EGBE, EGdiEE  and diEGME showed a significant reduction (p < 0.05) in viable litters produced (41, 77, 11, and 16 percent viable litters produced, respectively). The remaining glycols and glycol ethers produced no effect on that reproductive index at the concentrations administered. Postnatal observations varied widely in those groups with viable litters. EG and EGdiEE reduced the number of live pups per litter, increased the number of dead pups per litter, reduced pup survival, reduced pup birth weight and reduced pup weight gain over days 1 to 3 postpartum. EGdiEE data were not analyzed statistically because of the small sample size. diEGME significantly reduced (p < 0.05) the number of live pUps per litter and pup survival over days 1 to 3 postpartum. diEGdiEE significantly increased (p < 0.05) the number of dead pups per litter and significantly reduced (p < 0.05) pup birth weight. diEGdiBE increased (p < 0.05) the number of dead pups per litter. diEG reduced (p < 0.05) pup weight gain over day 1 to 3 postpartum and triEG and diEGEE reduced (p < 0.05) mean pup birth weight.

Discussion
Because it is our intent to use this bioassay to screen chemicals for their potential to cause reproductive toxicity in pregnant females, it is necessary to employ clearly toxic doses. If a chemical is evaluated as a low priority, one wants to be relatively confident that it was tested at a sufficiently severe challenge level. Clear maternal toxicity does not mean that reproductive toxicity follows. diEG and triEG produced 4% maternal mortality; diEGdiBE produced 8% mortality; and diEGEE produced 14% mortality. None of these showed strong evidence of reproductive toxicity. In fact, diEGEE treatment did not adversely affect any of the reproductive indices.
As noted, the estimated maternal LD10 was chosen as the challenge dose to be used for the reproductive studies. In practice, mortality will vary somewhat, and a response in the LD520 range was considered acceptable in the reproductive phase. If reproductive effects are noted in the presence of more than 20% maternal mortality, the test probably should be repeated at a lower dose. Conversely, if there is no reproductive toxicity and less than 5% maternal mortality, the test probably should be repeated at a higher dose. This LD=20 range, however, is only a tentative suggestion and further experience with this test may suggest other criteria for judging the appropriateness of the challenge dose.
The six end points examined for determining the priority of chemicals for further testing can be condensed into three levels of consideration. Most important is pup survival in utero, i.e., percent viable litters delivered by pregnant survivors. Of second importance is pup perinatal and postnatal survival, i.e., the number of live and dead pups per litter at birth, and pup survival over days 1 to 3 post partum. Final consideration is given to pup body weight end points, i.e., pup weight gain over days 1 to 3 and pup weight at birth. Chemicals were assigned to groups having different priorities for further testing, based upon whether a sufficient dose was administered (the maternal LD520) and the degree of effects recorded for each chemical. triEGdiME therefore has a very high priority for further testing because the administered dose of 4% maternal mortality was less than the LD!20, and the results were profound, i.e., no viable litters delivered. Other chemicals deserving high priority include EG, EGME, EGEE, EGdiEE, and diEGME in that all were tested within the range of 5 to 20% maternal mortality and all showed a drastic reduction of viable litters. A middle to high level priority group would include EGdiME and diEGdiME because both produced no viable litters but they received a dose greater than the LD520 which could have influenced the results due to maternal toxicity effects. A middle level priority group would include EGBE, diEG, diEGdiEE and triEG. EGBE significantly reduced viable litters with no other effects when dosed at the upper limit of the range of acceptable mortality. diEG, diEGdiEE and triEG mice received less than the LD520 dose but still produced some lesser effects: reduced pup weight gain for diEG; reduced pup weight gain and decreased number of live pups per litter for triEG; and an increased number of dead pups per litter plus a reduced pup birth weight for diEGdiEE. A low priority group would include diEGEE and diEGdiBE because the LD=20 dose was achieved and only lesser effects were seen, i.e., a reduced pup birth weight for diEGEE and an increased number of dead pups per litter for diEGdiBE. diEGBE mice did not receive a maternally toxic dose and no effects were found, thus diEGBE should be repeated until an LD.20 dose is achieved.
It is important to note that, for the purposes of this screen, comparisons among chemicals are based upon maternal mortality, i.e., doses less than, greater than, or within the LD,20 range. Comparisons as to the potential exposure hazard are not made. For example, although EG is designated to a higher priority group than EGBE, it may be that an individual is more likely to receive an 1180 mg/kg dose of EGBE (LD20) than he is to receive an 11,090 mg/kg dose of EG (LDIo). Thus EGBE may be more of a potential hazard than EG.
Previous investigations have shown several of the glycol ethers assessed in the current work to be teratogenic in traditional teratology test systems. EGME (designated in our high priority group) induced skeletal anomalies in the offspring of mice receiving, by oral gavage, as low as 31 mg/kg/day over days 7 through 14 of gestation (2). Gross (external) anomalies and the incidence of embryonic death were greatly increased in the offspring of mice receiving 250 mg/kg/day. In an inhalation study, 200 ppm EGME for 7 hr/day on gestation days 7 through 15 caused complete embryonic death in rats (3). Reduced fetal weights, skeletal and cardiovascular defects, as well as increased embryonic death occurred at both 50 and 100 ppm. EGEE, also in the high priority group, induced skeletal anomalies, increased embryonic death, and decreased pup weight in offspring of rats treated by oral gavage on days 1 through 21 of gestation (4). Doses ranged from 12 to 327 mg/kg/day. Inhalation of 160 ppm EGEE for 7 hr/day on days 1 through 18 of gestation increased embryonic death and skeletal anomalies in rabbits (5). At 615 ppm complete embryonic death occurred. Offspring of rats receiving a 1.0 mL or 2.0 mL dermal application of EGEE on days 7 through 16 of gestation showed greatly increased embryonic deaths, and cardiovascular and skeletal anomalies (2). Offspring of rats receiving 100 ppm of EGEE for 7 hr/day on days 7 through 13 or 14 through 20 of gestation showed altered behavioral patterns and altered brain neurochemical concentrations (7). EGdiME, placed in our middle to high level priority group, administered to pregnant mice by oral gavage on days 7 through 10 of gestation caused increased embryonic death at concentrations of 250, 350, and 490 mg/kg/day (8). Skeletal and external anomalies occurred in the 490 mg/kg group. EGBE, in our middle priority group, caused no apparent adverse effect on the offspring of rats exposed to 200 ppm EGBE for 7 hr/day on gestation days 7 through 15 (3). diEGEE, in our low priority group, caused no adverse effect on the offspring of rats exposed to 100 ppm for 7 hr/day on days 7 through 15 of gestation (3).
Some general statements can be made regarding structure-activity relationships. All mice receiving glycol ethers having terminal methyl groups, i.e., EGME, EGdiME, diEGME, diEGdiME and triEGdiME produced few viable litters (0, 0, 16, 0, and 0%, respectively). Only EGEE and EGdiEE having terminal ethyl groups produced similar results (0 and 11% viable litters, respectively). The remaining ethyl ethers (diEGEE and diEGdiEE), the butyl ethers (EGBE, diEGBE, diEGdiBE), and the glycol ethers with terminal hydroxy groups (EG, diEG and triEG) did not produce such profound fetotoxicity. Maternal toxicity of the glycols was sharply increased by the addition of an alkyl group. EGBE was more toxic than EGME which was more toxic than EGEE. All three showed greater toxicity than EG. The diEG mono-and diEGdi-alkyl ethers were more toxic than diEG, and triEGdiME was more toxic than triEG. The methyl ethers were generally more toxic than the ethyl or butyl ethers with the exception of EGBE.
We would caution again that neither NIOSH nor the NTP regards these results as definitive. Instead, they are suggestive, when considered along with other information on each chemical, of the urgency with which these chemicals should be considered for more detailed conventional testing. From the current available information on the glycol ethers, a correlation can be seen between the known positive teratogens (EGME, EGEE, and EGdiME) and their results in this in vivo screen where these compounds were designated either in a high priority or middle to high priority group. Also, EGBE and diEGEE, which showed no reproductive toxicity in conventional tests, were given a lower priority ranking (middle or low group) following testing in this screen.