FUNDAMENTAL
AND
APPLIED
TOXICOLOGY
19, 186-196
(1992)
Assessment of a Short-Term Reproductive and Developmental Toxicity Screen MARTHA
ROBERT E. CHAPIIU,’ ANN C. LOCKHART,* AND MICHAEL P. JOKINEN? With the technical assistance of Janet D. Allen and Eric A. Haskins
W. HARRIS,
Developmental and Reproductive Toxicology Group, *Computer Sciences Corporation, and tChemica1 Carcinogenesis Branch, NIEHS, Research Triangle Park, North Carolina 27709 Received
October
7, 199 1; accepted
Assessment of a Short-Term Reproductiveand Developmental Toxicity Screen. HARRIS, M. W., CHAPIN, R. E., LOCKHART, A. C., AND JOKINEN, M. P. (1992). Fundum. Appl. Toxicof. 19, 186-196. Short-term testsfor reproductive and developmentaltoxicity are neededto provide preliminary data on the toxicity of chemicals about which little or no data exist. An ideal designwould test all aspectsof reproduction and identify the target process in a short time period. One potential designhasbeenevaluated using four chemicals of varying reproductive/developmental toxicity. Swissmice were mated for 3 days prior to chemical exposureto producetime-matedfemalesfor gestationalexposure and to ascertain fertility of the untreated males.The group of time-mated femaleswas treated during Gestation Days 8-14 and allowed to litter for observations through Postnatal Day (PND) 4. Endpoints observedincluded pup number and body weightson PND 0, 1, and 4 and number of uterine implantation siteson PND 4. A secondgroup of femaleswas doseddaily for 19 days. After 7 days, these females (n = lo/group) were cohabited with male mice who had been treated for 5 days prior to this secondmating. Daily chemical dosingcontinued during the 5-day cohabitation. This secondgroup of femaleswaskilled
after 19 days of treatment and the number of live and dead fetuses and implantation sites was recorded. After 17 days of do;ing, malemice were killed and the reproductive systemevaluatedby organ weights,total epididymal spermcounts and motility, and testicularhistology.All four chemicalstested,boric acid,ethylene glycol, ethyleneglycol monomethyl ether, and theophylline, were found to be toxic to development or reproduction when tested previously by conventional developmentaltoxicity or continuous breeding protocols. This short-term (21 day) designcorrectly identified three of thesefour chemicalsasreproductive and developmental toxicants and distinguished the potent toxicants from the lesseffective compounds.This designcan be used to prioritize chemicalsfor further study, or to delineatethe relative toxicities of structurally related chemicals,and to identify the ’ To whom correspondence should be addressed NIEHS, P.O. Box 12233, RTP, NC 27709.
0 1992 by the Society of Toxicology.
All rights of reproduction
Drop
El-02.
186
0272-0590/92 $5.00 Copyright
at Mail
in any form reserved.
March
6, 1992
proper dose range for subsequent toxicity studies. o 1992Society of Toxicology.
While rodent oogenesis requires approximately 5 weeks from recruited oocyte to released ovum (Pederson, 1970), spermatogenesis from stem cell spermatogonia, and epididymal transit of sperm, requires ca. 9 weeks (Amann, 1982). Thus, tests to evaluate the potential reproductive toxicity of a compound are generally lengthy, lasting from lo- 13 weeks (Lamb, 1988). Numerous strategies exist for evaluating the reproductive toxicity of chemicals, but as the chemical evaluation process (Heindel, 1988) or the study design (Gray et al., 1988) become more complex, the total time and cost of the process increase. This, in turn, reduces the number of chemicals that can be assessed for reproductive toxicity. The efficiency of the screening process can be improved by shortening the test, and/or getting more information per exposure. Effects on spermatogenesis can be identified by histologic evaluation without waiting for decreased sperm output to result in fewer offspring, thus shortening the exposure time. Simultaneous screening for both reproductive and developmental (R/D) endpoints would increase the information produced by the test without necessarily doubling the resources required. The design detailed below is an answer to a challenge to construct a test that would, in a short period of time (several weeks), be able to separate the more “potent” R/ D toxicants from chemicals that are less toxic. This design attempts to identify those chemicals in which R/D effects are predominant in the profile of toxicities. It exposes as many vulnerable processes as possible to putative toxicant effects within a time frame of 21 days. As such, the processes of spermatogenesis, oogenesis, mating, sperm motility, fertilization, implantation, and organogenesis and early postnatal development are all subject to chemical exposure and subsequent evaluation.
SHORT-TERM Study
REPRODUCTIVE
DESIGN
187
Days
Group A Females
Males
Group B Females
FIG. 1. A schematic of the design used for these studies. For each dose level, two groups of females are used: Group A is treated with test agent continuously throughout the experiment, while group B females are treated only during GD 8-14. The one group of males per dose level started treatment after 3 days of mating with group B females. The endpoints are provided in the text. Co, cohabit: gd, gestation day; _ , days of chemical treatment; Bi, birth; Net, necropsy.
Four chemicals were chosen to evaluate the design. All four had been tested previously by conventional developmental toxicity and Reproductive Assessment by Continuous Breeding (RACB) protocols. Two chemicals [boric acid (BA) and ethylene glycol monomethyl ether (EGME)] produced sterility in the RACB studies, and EGME has been the subject of numerous publications documenting the profound adverse effects on the male, the female, and the conceptus (Lamb et al., 1984; Foster et al., 1984; Hanley et al., 1984; Doe, 1984; Nagano et al., 198 1). The other two chemicals [theophylline (THEO) and ethylene glycol (EG)] were chosen because there were published data on their reproductive toxicity, and compared to EGME and BA, both produced relatively mild effects in the RACB studies (Morrissey et al., 1988; Lamb et al., 1985; Price et al., 1985). In addition, EG, BA, and EGME produced significant adverse effects on the developing embryo, while THEO appears slightly less toxic. Thus, the chemicals were chosen to span a spectrum of toxicity. The questions to be addressed by the present studies were: (1) would the design correctly identify the R/D toxicity of the more potent chemicals, and if so, (2) what type of effects would be found? In addition, (3) would THEO and EG produce adverse effects, and if so, how much change would there be? Finally (4) would the test distinguish developmental toxicity from reproductive toxicity? METHODS
AND
MATERIALS
Animals. Swiss Crl:CD-1 mice, 12- 14 weeks of age, were purchased from Charles River Lab (Portage, MI). Male and female mice were allowed to acclimate for at least 2 weeks prior to start of dosing. Each animal was uniquely identified with a numerical tatoo at the base of the tail and was individually housed. Food (NIH 3 I Rodent Chow, Purina Mills) and distilled deionized water were available ad libitum. The animal rooms were maintained under controlled conditions of temperature (23 + 1“C), humidity (50 f 10%) and lighting (12: 12, 1ight:dark).
Design(Fig. 1). One group of male (n = 10) and two groups of female mice (designated A and B) were used at each dose level. Untreated males and untreated females (B) were cohabited for 3 days to provide time-mated females (n = 10 pregnant females/dose) for treatment during organogenesis [Gestational Day (GD) g-141. The dams were allowed to deliver and litters were evaluated on Postnatal Days (PND) 0, 1, and 4. Treatment ofthe males began on Study Day 3 (SD 3) and continued until SD 20. Group A females (n = IO/dose) were treated from SD 0 until SD 20. On SD 8 these females were cohabited for 5 consecutive days with the males on test. This group of females (A) was killed on SD 21 by an overdose of carbon dioxide. The male mice were killed on SD 20. Body weights were recorded for all animals every fourth day during exposure and at necropsy. In addition, percentage pregnant and number of live implants were measured in the group A females. At termination of group B females and their litters on PND 4, data on the following endpoints were collected: number of females delivering/number plugged, number of implantation sites, litter size, and body weights on PND 0. 1, and 4. Since pregnancy can alter the microscopic structure of systemic organs, and could thus confound histologic interpretations, histopathology was not performed on the females. However, a more comprehensive examination was made of the male mice: organ weights of liver. right kidney, right testis, and right epididymis were measured. Histopathology of the liver, left kidney, and left testis and epididymis was performed following fixation in 10% neutral-buffered formalin of liver and kidney sections and Bouin’s fixative for the testis and epididymis. Tissues were embedded in paraffin and sectioned at 6 pm, and liver and kidney sections were stained with hematoxylin/eosin. while sections of testis and epididymis were stained with PAS/H. Total right cauda epididymal sperm counts were measured as described (Chapin et al., 1985). Epididymal sperm motility was evaluated by a modification ofthe method described by Dunnick et a/. (1986): rather than immediate evaluation, sperm were videotaped, and the tape was subsequently evaluated for moving/not-moving sperm. A total of at least 200 sperm were evaluated for each animal. The endpoints assessedin this 21-day design are compared with those measured in developmental toxicity and continuous breeding (RACB) studies and presented in Table 1. This table shows that the present design assesses many of the endpoints found in each of the other designs. The primary differences are in the durations of exposure: RACB studies expose F0 animals to the test agent for 15 weeks during mating and generate up to five litters (rev. in Lamb, 1988). Developmental toxicity studies traditionally expose pregnant females on GD 6-15. These exposure times were shortened to
188
HARRIS ET AL.
TABLE 1 Comparison of the Parameters of a Developmental Toxicity Study (DEV), Reproductive Assessment by Continuous Breeding Study (RACB), and the 21-Day Screen DEV
RACB
SCREEN
J
J
J
J
J
J
J
J
J
J
J
J
J
r,
J
J
-
J
J
-
General toxicity Body weight gain Terminal body weight Kidney weight Liver weight Mortality Water consumption Feed consumption Reproductive toxicity Fertility index Mean litters per pair Live pups/litter Live pup weight (F,) Days to litter Days between litters Testis weight Epididymis weight Cauda epididymis weight Prostate weight Seminal vesicle weight Testicular histology Cauda sperm density Sperm motility % Abnormal sperm Estrous cycle length Ovary weight Developmental toxicity No. implantation sites/dam Postimplantation loss Number of resorptions Number live pups Live pup weight Neonatal viability to Day 4 Neonatal death External malformations Internal malformations External variations Internal variations Weight at 21 days Fertility index
-
J
-
J J
J
J
J
J
-
J
J
-
J
-
J
J -
-
J
J
-
J
J
-
J
-
J
-
-
J
-
-
J
J
-
J
J
-
J
J
-
J
-
J
-
-
J
-
J
J
-
J
-
J
J
-
J
J
-
J
J
-
J
J
J
J
J
J
J
J
J -
J J
J
-
J -
J J
J
-
J
J
decrease the length and cost of the study. consistent with the intent of a “first-pass” study. Chemicals. Each chemical was tested at three dose levels, chosen based on literature-derived LD50 values, and not on previous in-house data. Approximately one-third of the reported LD50 was selected as the high concentration and a log difference for the low level; the middle dose was approximately one-third of the high level. All chemicals were administered daily by gavage at a volume of 0. I ml/IO g body weight. Dosing solutions were prepared by Radian Chemical Containment Laboratories, (Morrisville. NC). Boric acid (BA. CAS No. 10043-35-3) was administered in corn oil at 0. 120. 400, and 1200 mg/kg/day. EC (CAS No. 107-21-I) in water was
tested at 0, 250, 700, and 2500 mg/kg/day. EGME (CAS No. 109-86-4) was administered in water at 0, 70, 250, and 700 mg/kg/day. The fourth chemical, THEO (CAS No. 58-55-9) was tested at 0, 20, 60, and 200 mg/kg/day, using corn oil as the vehicle. All formulations were analyzed by gas chromatography after dosing and were found to be 93-100% of target concentrations. Histologic evaluation. Because of the shortened time frame, this study design relied heavily on histologic evaluation as a complement to the functional evaluations. All tissueswere assessedby observers unaware of treatment group. Liver and kidney sections were analyzed for deviations from “normal” by standard pathology criteria used in the National Toxicology Program. Deviations were graded as minimal, mild, and moderate. Testicular histology was evaluated by the criteria of Oakberg (1956). Eight control sections of testis from two of the studies were evaluated first, to provide a baseline from which to judge effects. The control slides were then placed back with their respective studies. and the testis sections were scored using the following classification system: 0 for no observable effect: I for minor changes in 20 to 50% of the tubules, consisting of increases in residual body size, and/or up to 25% cell exfoliation or death (necrosis), and/or slight decrease in cell number, or small visible changes in synchronization of cells within a stage (less than a 3 stage missmatch); 2 for bigger changes in timing (> a 3 stage discrepancy). and/or disruption/exfoliation in up to 50% of the tubules: 3 for exfoliation/disruption in up to 50% of tubules, or the presence of some end-stage tubules. and/or up to 50% overall germ cell loss; and 4 for >50% end-stage tubules, and/or massive germ cell loss. Stafistical analyses. Nonparametric statistical analysis was used for all endpoints. The pregnancy rates of group A females were evaluated using the Cochran-Armitage test for linear trend (Armitage, 197 I), followed by Fisher’s exact test (Reiss, 198 I) when significance was observed in the trend statistic. The remaining endpoints for group A females, as well as those for the group B females and the males, were analyzed using Kruskal-Wallis analysis of variance procedures to determine if overall differences existed between dose groups. Dose-response trends were assessedusing Jonckheere’s test against ordered alternatives (Hollander and Wolfe, 1973). When significance was noted in either the trend test or the Kruskal-Wallis ANOVA. pairwise comparisons were made using the Mann-Whitney U test (Hollander and Wolfe. 1973).
RESULTS Males
Control males in all four studies gained an average of 0.5% body weight during the course of each study. Weight gain was not inhibited by EGME or by BA; animals in the top dose group of the EG study lost 0.5% of their weight [no significant difference (NSD) from controls], and males in the top dose in the THEO study lost 1.2% of their weight (NSD). No males showed adverse clinical signs. There were no treatment-related deaths among males, although one died early in the study from gavage error. Administration of BA and EGME led to a reduction of testis weight (p < 0.05) after 19 days of dosing (Table 2). Neither chemical reduced epididymal weight, although epididymal sperm density and motility were reduced in the group receiving EGME at 700 mg/kg (p < 0.05, Table 2). Significant histopathology was noted in the testes of animals treated with BA and EGME (Fig. 2). The testes of mice receiving 250 mg EGME/kg/day (the middle dose group)
SHORT-TERM
REPRODUCTIVE
189
DESIGN
TABLE 2 Male Organ Weights, Sperm Parameters, and Histology Scores following Chemical Exposure in a 2 1-Day Reproductive/Developmental Screen Chemical”
Testis* weight (mg)
Epididymis* weight (mg)
No. sperm/g cauda. (X 106)
% Motile sperm
Histology score
EGME 0 70 250 700 Trend
130 + 5.6 130 * 7.3 85 f 5.6* 42 + 2,s p < 0.05
60.0 67.8 53.0 51.2
f f f f NS
5.1 7.0 6.5 7.4
917 f 167 1069 f 282 1347 f 480 380 f 97* p < 0.05
14.1 k 1.0 81.1 f 3.4 13.3 f 6.6 12.4 + 3.9* p < 0.05
0.5 0.1 2.1 5.0
zk f f f
0 250 700 2500 Trend
117 124 121 115
k 5.3 + 5.6 k 7.6 + 3.3 NS
44.6 45.8 45.2 43.8
k k + f NS
1.5 1.7 1.4 1.6
1179* III 1119* 57 1160+ 97 962 k 121 NS
90.6 87.5 89.0 89.5
0.4 0.2 0 I .o
i 0.2 + 0. I *o t 0.3
0 120 400 1200 Trend
1 I I i 3.8 I10 + 6.0 91 f 4.7” 87 k 5.5* p < 0.01
42.4 40.9 42.2 37.0
+ * * + NS
4.2 2.4 4.2 3.0
1761 1396 2368 972
f k f k NS
492 391 797 228
0.3 0.2 0.4 2.8
f f + *
0.2 0. I 0.2 0.2
118 119 112 123
46.0 45.9 44.9 54.1
+ 1.4 AC 3.2 * 3.3 f 3.2 NS
1080 1085 887 888
f f f + NS
101 116 118 134
0.2 0.5 0.2 0.7
f 2 f f
0.1 0.2 0. I 0.2
0.2 0.1 0.2 0.0
EC i k + f NS
2.2 2.7 3.2 2.3
BA ND ND ND ND
THEO 0 20 60 200 Trend
+I 5.5 F 5.0 * 4.9 f 5.3 NS
Note. Trend, test performed using Jonckheere’s test: NS. not significantly ’ Chemical dose in mg/kg/day. * Mean f SEM. )I = 9 for 1200 mg/kg BA; IO for all other groups. * Different from controls, p i 0.05.
showed extensive spermatocyte and spermatid cell loss, although complete spermatogenesis (through elongated spermatids) was present in many tubules (Fig. 2D). The highest dose of EGME resulted in near total germ cell loss (Fig. 2C); only scattered tubules contained any postspermatogonial germ cells in lO/ 10 testes. These effects are reflected in the histology scores (Table 2). The effects of BA were not as severe, but still marked: the high dose (1.2 g/kg/day) of BA caused exfoliation/disruption in >50% of tubules, with up to 50% germ cell loss in 7/9 mice receiving 1.2 g/kg/day for 19 days (Fig. 2B, Table 2). This was accompanied by asynchronous development: round spermatids were occasionally seen in late-stage tubules that should not contain these cells. The middle dose group of BA was generally indistinguishable from controls. THEO produced milder changes in the epithelium that consisted primarily of asynchronous germ cell development and focal loss of one or more generations of germ cells within individual tubules (Fig. 2E). Round spermatids were noted in stages X, XI, and XII; some of these
different
from
control;
78.5 81.3 81.8 80.4
ND,
f +k f NS
3.2 4.6 4.1 4.5
not determined,
round cells were undergoing karyolysis. Focal germ cell loss involved all germ cell generations, although usually only one generation was missing per tubular cross section. No treatment-related lesions were noted in testes from EG-treated mice. Histologic evaluation of liver and kidney sections could identify no treatment-related effects for any of the chemicals. Group A Females
No females in the EGME and EG groups showed adverse clinical signs. One THEO female and three BA females showed rough hair coat prior to death during chemical exposure; one of those females was euthanized moribund (Table 3). An effect (p < 0.05) on impregnation was found for EGME at the high dose level (Table 3). EGME also produced a slight and insignificant decrease in implant number. Females treated with the top dose of EG had significantly fewer live, and more dead implants, with 2/6 litters totally resorbed. None of the chemicals decreased the total number of im-
190
HARRIS ET AL.
FIG. 2. (A) Section of testis from a control mouse. All steps of germ cells were present. with minimal necrosis in stage XII. Scale bar, 30 pm. (B) Testis section from a mouse treated with I .2 g BA/kg/day. Note the round spermatids present with dividing spermatocytes (arrowheads) and the lack of elongating spermatids. Scale bar, 30 Frn. (C) Testis section from a mouse treated with 700 mg EGME/kg/day. Note the absence of postspermatogonial germ cells in most of the tubules. Scale bar, 30 +m. (D) Testis section from a mouse treated with 250 mg EGME/kg/day. Note the absence of round and elongating spermatids (tubule A) and a decreased number of germ cells in the remaining tubules. Scale bar, 30 pm. (E) Testis section from a mouse treated with 200 mg THEO/kg/day. Note the aberrant presence and position of round spermatids in conjunction with dividing spermatocytes (small arrowheads) or misplaced within the epithelium (large arrowhead. left). Scale bar, 30 grn.
plantation sites in these females, who had been mated during chemical exposure. Group B Females All of these females (impregnated prior to the start of dosing) had equivalent numbers of uterine implantations regardless of chemical treatment (Table 4). Treatment with EGME (250 and 700 mg/kg/day) or BA (1200 mg/kg/day) significantly reduced the number of females delivering and decreased the number of live pups. Although none of these chemicals caused neonatal mortality between PND 1 and PND 4, pups from mothers treated with EC during gestation had lower body weights on these postnatal days. DISCUSSION The data presented above demonstrate that the design employed in this study can identify chemicals that have significant R/D toxicity. Administration of BA or EGME resulted in significant effects on male reproduction and on the group B females (number of litters, number of young). Ethylene glycol administration also resulted in changes in postnatal litter weights. The ability of this design to differentiate
developmental toxicity from reproductive toxicity (effects on gametogenesis) is shown in the findings that fertilization and implantation were not affected (there was no change in the number of implantation sites), but the decrease in the number of pups in some groups (EC) shows that development wczsadversely affected. The chemicals used for these pilot experiments were chosen to represent two levels of effect, based on results of preceeding developmental toxicity and RACB studies. A summary of the effects noted in previous studies is presented in Table 5, where these effects are compared to those seen in the present study. As regards the fertility indices, both BA and EGME had produced complete infertility in previous RACB studies. With each chemical, males and females were affected (Fail et al., 1989; Lamb et al., 1984, respectively). In the RACB study, theophylline consistently decreased pup number in all dose groups, starting with the first litter (Morrissey et al., 1988) from a 22% decrease in the low dose group to a 42% decrease in the high dose group. The high dose THEO group in the RACB study also had lowered pup weight, increased F0 male liver weights, and a 20% reduction in sperm count. The high dose of EC, in the RACB study, produced an 18%
SHORT-TERM
REPRODUCTIVE
FIG.
DESIGN
191
2-Confinued
decrease in overall pups per litter and lower pup weight in all dose groups (Lamb et al., 1985) with no change in F0 male or female organ weights. Thus, in longer reproductive toxicity tests that utilized more endpoints and collected more data, BA and EGME were clearly more toxic to the reproductive systems than were THEO or EG. While there are occasional divergences in the data from the two study designs (see Table 5) it is noteworthy that the relative toxicity dif-
ferential previously seen in the RACB studies was also reflected by the data from the current design. Increased resorptions were seen after treatment with EG or BA in both the present study and in previous reports (Price et al., 1985; Heindel et al., 1991, respectively). Lindstrom et al. (1990) report increased resorptions after THEO exposure to Swiss mice at 372 and 396 mg/kg/day, doses considerably higher than those used in the present
192
HARRIS
FIG.
ET
AL.
2-Continued
study. Previous studies with EGME found increased fetal deaths and resorptions at higher doses (Nagano et al., 1981) while lower doses produced skeletal and visceral malfor .mations and variations. (Nagano et al., 198 1; Hanley et /al., 1984). It is perhaps not surprising that the developn nental toxicity studies identified structural malformatiot 1s and variations that were not identified by the screen . The design presented here is more closely related
to the Chernoff-Kavlock test (Chernoff and K avlock, 1982) than to a Segment II study: more reliance is; placed upon the relatively gross indices of pup weight, n umber, viability, and survival than on structural effects. T ‘he lack of pup/fetal dissection in the present studies mea ns that visceral or skeletal malformations or variations co uld not have been detected. Thus, a toxicant whose effect is only visible at dissection would not be identified as aL devel-
SHORT-TERM
REPRODUCTIVE
TABLE 3 Impregnation and Uterine Implant Data from Continuously Exposed Females following Chemical Exposure in a 2 1-Day Reproduction/Developmental Screen
DESIGN
193
change in the number of pups in the THEO 2 1-day study, while there was a clear effect on that endpoint in the RACB study, starting with the first litter. In addition, both chemicals produced effects in developmental toxicity studies that were not seen using the 21-day design (Table 5). No. live No. dead Total The question of sensitivity is also addressed by setting No. pregnant implants per implants per implants Chemical” (No. treated) femaleb femaleb per femaleb doses to assure some overt toxicity. The highest doses employed in the present studies were set at one-third the reEGME ported LD50. Compounds that produce no developmental 0 IO (10) 10.3 * 1.0 0.5 + 0.2 10.8 f 0.9 or reproductive toxicity at these doses would not be ex70 10.8 f 0.4 0.8 f 0.2 11.5 f 0.4 8 (10) pected to pose a risk of severe reproductive toxicity when 250 10 (10) 7.3 AI 1.8 3.8 f 1.4 11.1 kO.7 tested under other designs. This does not mean that there 700 6 (lo)* 6.7 f 1.9 0.2 + 0.2 6.8 f 2.0 Trend p < 0.05 NS NS NS would not br effects: rather, it indicates that this design is intended to identify those chemicals that are likely to be EC highly toxic to the conceptus and/or the reproductive sys0 10.0 f 0.9 0.4 * 0.2 10.4 & 1.0 9(10) 250 10.7 f 0.4 0.8 t 0.1 11.4 + 0.3 tem. Whereas positive effects in this design indicate that 9 (10) 700 11.1 ztzo.5 0.3 + 0.3 11.4 + 0.4 7(10) another, longer. test would also find R/D toxicity, the lack 2500 7.4 + 1.2* 1.4 t 0.4* 8.9 f 1.1 9 (10) of effects do not necessarily assure that a subsequent, Trend NS p < 0.05 p < 0.05 NS longer, test would also be negative. This is especially true BA for toxicants that bioaccumulate; such compounds would 0 10 (10) 12.3 f 0.6 0.3 2 0.1 12.6 + 0.5 be incorrectly identified as nontoxic. This is a drawback 120 12.1 + 0.5 0.4 ?I 0.2 12.6 f 0.4 9 (10) of any short-term evaluation. 400 12.7 + 0.6 0.3 +_ 0.2 13.0 f 0.5 9 (9) Sensitivity in detecting effects upon spermatogenesis 1200 9.8 f 3.1 3.0 + 2.3' 12.8 f 1.8 6 (7)[31 NS NS NS Trend NS might be enhanced by the use of glycol methacrylateembedded sections, which allows for improved visualizaTHEO tion of the seminiferous epithelium (e.g., Hess, 1990). By 0 10.0 f 0.5 0.4 i 0.2 10.4 + 0.4 9 (10) 0.3 f 0.2 10.3 + 1.1 20 10.0 f 1.1 9(10) using a better means of testicular evaluation, one need not 60 10.4 * 0.4 1.0 f 0.4 11.4 + 0.3 9 (10) treat the animal for the full cycle of spermatogenesis and 1.2 + 0.7 11.2 f 1.2 200 10.0 f 1.1 6 (9)Ill epididymal transit time, and then evaluate pups produced Trend NS NS NS NS from those sperm, to detect an effect upon spermatogenesis. Rather, high-resolution light microscopy can allow Note. Trend, test performed using Jonckheere’s test; NS, not significantly different from controls; [n], number of mice dying during chemical exposure. an evaluation of all the cells in the testis after a short exa Chemical dose in mg/kd/day. posure, giving similar information in a much reduced time b Mean f SEM. frame. It should be noted that genetic toxicities will re’ Two of six litters totally resorbed. main undetected by histologic means alone, but might well * Different from controls, p < 0.05. be identified by other short-term i~z 11ivo or in vitro tests. This design has a limited ability to address behavioral opmental toxicant. While this study identified only EG and BA as developmental toxicants, much higher doses of effects. Compounds affecting libido and/or mating behavior would cause a decrease in the number of pregnant THEO have been required to demonstrate fetal toxicity. group A females. While additional tests would be required And while EGME was not found to be developmentally toxic, it’s antifertility effects were clear; EGME was cor- to identify the affected process, an effect would have been found, which could be considered one key requirement of rectly identified by the present design as toxic. any pilot study. Other shortcomings of a short-term design One of the primary questions of a new testing approach is, will it identify chemicals that are developmental or re- include the inability to identify second-generation changes productive toxicants? This design appears to do that. An- in fertility or behavior and the fact that the dosing period other question is, where is the limit of sensitivity, or how is too short to allow for a full expression of testicular toxicity to be manifest as large testicular weight loss or large toxic does a chemical need to be in order to be detected? increases in abnormal sperm morphology. Indeed, because One might logically expect a decrease in sensitivity with the time frame of the study is so short relative to spera short period of dosing and only one mating period. This is borne out by the results with EG and THEO; the effects matogenesis, sperm morphology was considered to be a evaluation of of both chemicals were more profound in the RACB stud- poorly responsive endpoint. Morphologic ies than in the present design. For example, there was no sperm was omitted from these studies, but can be added
194
HARRIS
ET
AL.
TABLE 4 Neonatal and Uterine Implant Data for Females Exposed during Gestation following Chemical Exposure during a 21-Day Reproduction/Developmental Screen Total
No. live neonates Chemical”
No. females littering (No. Rx)
PND
Ob
PND
1
PND
4
PND
litter
1
wt. PND
No. implantation sites per female
4
EGME 0 70 250 700 Trend
9 (10) 11 (11) 12 (12) O(ll)
0 250 700 2500 Trend
12 12 12 14
0.5 0.4 0.7 0.6
NS
1 I .6 -+ 11.5 * 11.5 * 12.4 + NS
19.2 f 1.1 19.8 + 0.8 19.6 + 0.7 15.3 f 0.9* p < 0.05
31.7 I 1.3 34.1 k 1.1 33.4 f 1.2 26.6 f l.5* p < 0.05
12.0 k 1 I .9 + 12.5 zk 11.9 f NS
0.5 0.5 0.4 0.5
10.9 f 0.6 I I .2 f 0.7 10.4 f 0.4 o.o* p < 0.01
20.3 k 0.8 20.2 f 0.8 19.6 k 0.8 -
33.5 rt 1.3 33.5 i 1.2 33.4 k 1.4
NS
NS
I I.0 f 0.6 11.2+0.7 10.8 f 0.4 1 I .3 + 0.5 NS
11.4 * 0.4 9.3 + 0.8 11.0*0.7 10.8 f 0.5 NS
20.4 17.8 19.9 18.7
10.9 t 0.6’ 10.7 * 0.4 0.4 f 0.2* o.o* p -=z 0.01
10.9 + 0.6 10.7 f 0.4 o.o* o.o* p < 0.01
10.6 k 0.7 10.6 f 0.3 o.o* o.o* p < 0.01
18.1 + 1.0 16.8 k 0.9 -
28.6 k 2.1 24.9 + 1.9 -
NS
11.3 11.3 11.0 10.3
+_ 0.6 + 0.6 * 1.0 F 0.7 NS
11.2kO.6 1 I .3 * 0.6 10.9 + I.0 10.2 t 0.7 NS
11.2 11.3 10.9 10.1
? 0.6 Y!Z0.6 f 1.0 t 0.7 NS
10.9 + 0.6 11.2kO.7 10.7 * 0.4 o.o* p < 0.01
10.9 + 0.6 1 1.2 * 0.7 10.4 * 0.4 o.o* p < 0.01
11.4 9.4 11.0 10.9
11.4 + 0.4 9.3 + 0.8 11.ot0.7 10.8 + 0.5 NS
EG (12) (12) (12) (14)
BA 0 120 400 1200 Trend
10 (10) 11 (II) 12 (12) O(11)
THEO 0 20 60 200 Trend
14 13 15 14
(14) (13) (15) (14)
NOIE. Trend, test performed using Jonckheere’s ’ Chemical dose in mg/kg/day. b Postnatal day. c Mean f SEM. * Different from controls, p < 0.05.
* 0.4 + 0.8 * 0.7 + 0.5 NS
test; NS, not significantly
to studies of other chemicals as desired. Additionally, the design uses only 10 animals per group. It has been shown that, especially for reproductive endpoints, increasing the number of animals increases the quality of the data (Schwetz et al., 1980). This is also obvious from a statistical viewpoint. Thus, this design is not as sensitive as much longer tests, but it also costs considerably less and takes less time. The optimal design for a given situation depends on the question being asked, the resources available, and the confidence required in the answer. Additional endpoints can be added to address specific interests or applications. For example, the NTP is using this design to evaluate AIDS therapeutics and has added immunotoxicology endpoints to identify possible immune system effects. Additional reproductive or systemic endpoints could be added, such as computer-assisted sperm motion analysis or other organ-specific tests. We have
different
from
controls;
k + + k NS
0.8 0.5 1.3 0.9
No. Rx. number
34.8 31.4 32.9 32.5
f ? f + NS
1.2 0.8 1.5 1.2
of females
12.2 11.1 12.0 Il.3
treated.
t It + + NS
0.5 0.3 0.6 0.5
in parentheses.
housed mice in metabolism cages for the last 24 hr of dosing to assess possible kidney changes. Specific nonlethal developmental abnormalities could be identified by terminal dissection of pups on PND 4. We view the approach detailed above as a basic model to which options can be added as appropriate. Finally, the usefulness of these data will depend largely upon the context in which they are generated. We view this test as being most useful for generating preliminary data that will ultimately become part of a larger data set. For example, a compound of unknown toxicity that produces some developmental toxicity in this design would then be subject to a more complete developmental toxicity assessment; the focus would be on the developmental toxicity rather than female or male reproductive toxicity. The design could also be used to rank the toxicity of members of a structurally related chemical class. The data from this screen would thus
SHORT-TERM
REPRODUCTIVE
TABLE 5 Comparison of Chemical Effects in Reproductive Assessment by Continuous Breeding (RACB), 21-Day Screen, and Developmental Toxicity Studies EGME
EG
BA
Reproductive endpoints Body weight Kidney weight Liver weight Mortality Fertility index Live pups/litter Pup weight Days to litter Testis weight Cauda weight Epididymis weight Testis hisology No. sperm/g cauda Sperm motility Live births Neonate death Viability External malformations
Aadb NC/NC t/t NC/NC $14 J/J NC/NC NC/NC $14 NC/NC $/NC t/t $14 CA 414 t/t &/NC NC/NC
NC/h NC/NC NC/NC NC/NC NC/NC NC/NC 414 NC/NC 4/J NC/NC NC/NC t/t NC/NC $/NC NC/NC NC/NC NC/NC f/NC
J/t4 NC/NC +/NC t/t J-14 J/JJ/NC NC/NC i/c J-/NC h/NC t/t J/NC i/ND r/J NC/NC NC/NC NC/NC
e/JNC/NC C/NC NC/NC NW $/NC NC/NC NC/NC NC/NC NC/NC NC/NC ND/t h/NC NC/NC h/NC t/NC NC/NC NC/NC
Developmental endpoints Body weight Liver weight Kidney weight Mortality Implantation sites/dam Number of resorptions Malformation& Variations
NCc/Jb ND/t ND/NC NC/NC NC/NC NC/NC NC/NC t/NC
414 $/NC NC/NC NC/NC NC/NC t/t f/NC t/ND
w NC/NC t/NC NC/t NC/NC
4/c NC/NC ND/NC NC/NC NC/NC
t/t
t/NC h/ND
and and Kim this
Dr. Bernard Schwetz and the other members of the Developmental Reproductive Toxicology Group (Drs. Rick Morrissey, Jerry Heindel, Treinen) for stimulating discussions during the formative stages of project.
REFERENCES
THEO
,N::!: t/ND
Note. NC, no change compared to controls; ND, not determined; tNC, equivocal increase; t, increased, compared to controls; 4, decreased, compared to controls; EGME, ethylene glycol monomethyl ether: EG, ethylene glycol: BA. boric acid; THEO, theophylline. a RACB data. ’ 2 I -Day screen data. ’ Developmental toxicity. d Only external malformations were assessedin the screen.
permit the deployment of resources toward the highestpriority questions. While no short-term test can replace one using a longer exposure or a complete developmental assessment, it appears that this design shows promise for being able to detect those chemicals that have significant reproductive and/or developmental toxicity. Continued use of this design will improve our understanding of the merits and faults of this test and allow for its use in the most efficient manner.
195
DESIGN
Amann, R. P. ( 1982). Use of animal models for detecting specific alterations in reproduction. Fundam. Appl. Toxicol. 2, 13-26. Armitage, P. (197 I). Cochran-Armitage trend in proportions. In Statistical Methods in Medical Research, pp. 362-365. Blackwell, Great Britain. Chapin. R. E., Dutton, S. L., Ross, M. D., and Lamb IV, J. C. (1985). Effects of ethylene glycol monomethyl ether on mating performance and epididymal sperm parameters in F344 rats. Fundam. Appl. To,xicol. 5, 182189.
ChemotT N., and Kavlock, R. J. (1982). An in vivo teratology screen utilizing pregnant mice. J. To.xicol. Environ. Health 10, 541-550. Doe, J. E. (1984). Ethylene glycol monoethyl ether and ethylene glycol monoethyl ether acetate teratology studies. Environ. Health Perspect. 57, 3341. Dunnick, J. K.. Harris, M. W.. Chapin, R. E., Hall. L. B., and Lamb IV, J. C. ( 1986). Reproductive toxicology of methyldopa in male F344/N rats. Toxicology
41, 305-318.
Fail, P. A.. George. J. D., Grizzle, T. B., and Heindel, J. J. (1989). Reproductive toxicity of Boric Acid (BORA) in mice as evaluated by the Continuous Breeding protocol. Toxicologist 9, 266. Reiss, J. L. (198 1). Statistical Methods,for Rates and Proportions, 2nd ed., pp. 25-29. Wiley, New York. Foster. P. M. D.. Creasy, D. M., Foster, J. R., and Gray, T. J. B. (1984). Testicular toxicity produced by ethylene glycol monomethyl and monoethyl ethers in the rat. Environ. Health Perspect. 57, 207217.
Gray, L. E.. Jr., Ostby, J., Sigmon. R., Ferrell, J.. Rehnberg, G., Linder. R.. Cooper, R.. Goldman, J., and Laskey, J. (1988). The development of a protocol to assessreproductive effects of toxicants in the rat. Reprod. Toxicol.
2, 281-287.
Hanley, T. R., Jr., Yano, B. L., Nitschke, K. D.. and John J. A. (1984). Comparison of the teratogenic potential of inhaled ethylene glycol monomethyl ether in rats, mice and rabbits. Toxicol. Appl. Pharmacol. 75,409422.
Heindel, J. J. (1988). The National Toxicology Program chemical nomination selection and testing process. Reprod. To-xicol. 2, 273-279. Heindel, J. J., Price. C. J., Field, E. A., Marr, M. C., Myers, C. B., Morrissey, R. E.. and Schwetz, B. A. (199 I ). Developmental toxicity of boric acid in mice and rats. Fundam. .4ppl. Toxicol. 18, 266-277. Hess, R. A. (1990). Quantitative and qualitative characteristics of the stages and transitions in the cycle of the rat seminiferous epithelium: Light microscopic observations of perfusion fixed and plastic-embedded testes. Biol. Reprod.
43, 525-542.
Hollander, M.. and Wolfe, D. A. (1973). Nonparametric Statistical Methods, pp. 27-33 and 120-123. Wiley, New York. Lamb IV, J. C., Gulati, D. K., Russell, V. S., Hommel, L.. and Sabharwal, P. S. (1984). Reproductive toxicity of ethylene glycol monoethyl ether tested by continuous breeding of CD-1 mice. Environ. Health Perspect. 57,85-90.
ACKNOWLEDGMENTS We are grateful to Mr. Brad Collins (NTP) and Dr. Bob Moseman (Radian) for assistance with dosing solution formulation and analysis
Lamb IV, J. C., Maronpot, R. R., Gulati, D. K., Russell, V. S., HommelBarnes, L., and Sabharwal. P. S. (1985). Reproductive and developmental toxicity of ethylene glycol in the mouse. Toxicol. Appl. Pharmacol. 81, 100-I
12.
196
HARRIS ET AL
Lamb IV, J. C. (1988). Fundamentals of male reproductive toxicity testing. In Physio1ug.vand To.uicologv qfMale Reproduction (J. C. Lamb IV and P. M. D. Foster, Eds), p. 137-154. Academic Press, San Diego. Lindstrom. P.. Monissey, R. E., George, J. D., Price, C. J., Mat-r, M. C., Kimmel. C. A., and Schwetz, B. A. (1990). The developmental toxicity of orally administered theophylline in rats and mice. Fundam. iippnpl.To.*-id
14, 167-178.
Morrissey, R. E.. Collins, J. J., Lamb IV, J. C.. Manus, A. G., and Gulati, D. K. ( 1988). Reproductive effectsof theophylline in mice and rats. Fundam. .4pp/. Tosicol. 10, 525-536. Nagano. K., Nakayama, E., Oobayashi, H.. Yamada, T., Adachi. H., Nishizawa. T.. Ozawa, H., Nakaichi. M., Okuda, H., Minami, K.. and Ya-
mazaki, K. ( 198 1). Embryotoxic effects of ethylene glycol monomethyl ether in mice. To.rico/ogy 20, 335-343. Oakberg, E. F. (1956). A description of spermiogenesis in the mouse and its use in analysis of the cycle of the seminiferous epithelium and germ cell renewal. Am. J. Anat. 99, 391-413. Pederson, T. (1970). Follicle kinetics in the ovary of the cyclic mouse. Acta Endocrinol. 64, 304-323. Price, C. J.. Kimmel, C. A., Tyl, R. W., and Marr, M. C. (1985). The developmental toxicity of ethylene glycol in rats and mice. To.xico/. Appl. Pharmacol. 81, I I 3- 127. Schwetz, B. A., Rao, K. S.. and Park, C. N. (1980). Insensitivity of tests for reproductive problems. J. Erwiron. Pathol. Toxicoi. 3, 8 l-98.
AND
APPLIED
TOXICOLOGY
19, 186-196
(1992)
Assessment of a Short-Term Reproductive and Developmental Toxicity Screen MARTHA
ROBERT E. CHAPIIU,’ ANN C. LOCKHART,* AND MICHAEL P. JOKINEN? With the technical assistance of Janet D. Allen and Eric A. Haskins
W. HARRIS,
Developmental and Reproductive Toxicology Group, *Computer Sciences Corporation, and tChemica1 Carcinogenesis Branch, NIEHS, Research Triangle Park, North Carolina 27709 Received
October
7, 199 1; accepted
Assessment of a Short-Term Reproductiveand Developmental Toxicity Screen. HARRIS, M. W., CHAPIN, R. E., LOCKHART, A. C., AND JOKINEN, M. P. (1992). Fundum. Appl. Toxicof. 19, 186-196. Short-term testsfor reproductive and developmentaltoxicity are neededto provide preliminary data on the toxicity of chemicals about which little or no data exist. An ideal designwould test all aspectsof reproduction and identify the target process in a short time period. One potential designhasbeenevaluated using four chemicals of varying reproductive/developmental toxicity. Swissmice were mated for 3 days prior to chemical exposureto producetime-matedfemalesfor gestationalexposure and to ascertain fertility of the untreated males.The group of time-mated femaleswas treated during Gestation Days 8-14 and allowed to litter for observations through Postnatal Day (PND) 4. Endpoints observedincluded pup number and body weightson PND 0, 1, and 4 and number of uterine implantation siteson PND 4. A secondgroup of femaleswas doseddaily for 19 days. After 7 days, these females (n = lo/group) were cohabited with male mice who had been treated for 5 days prior to this secondmating. Daily chemical dosingcontinued during the 5-day cohabitation. This secondgroup of femaleswaskilled
after 19 days of treatment and the number of live and dead fetuses and implantation sites was recorded. After 17 days of do;ing, malemice were killed and the reproductive systemevaluatedby organ weights,total epididymal spermcounts and motility, and testicularhistology.All four chemicalstested,boric acid,ethylene glycol, ethyleneglycol monomethyl ether, and theophylline, were found to be toxic to development or reproduction when tested previously by conventional developmentaltoxicity or continuous breeding protocols. This short-term (21 day) designcorrectly identified three of thesefour chemicalsasreproductive and developmental toxicants and distinguished the potent toxicants from the lesseffective compounds.This designcan be used to prioritize chemicalsfor further study, or to delineatethe relative toxicities of structurally related chemicals,and to identify the ’ To whom correspondence should be addressed NIEHS, P.O. Box 12233, RTP, NC 27709.
0 1992 by the Society of Toxicology.
All rights of reproduction
Drop
El-02.
186
0272-0590/92 $5.00 Copyright
at Mail
in any form reserved.
March
6, 1992
proper dose range for subsequent toxicity studies. o 1992Society of Toxicology.
While rodent oogenesis requires approximately 5 weeks from recruited oocyte to released ovum (Pederson, 1970), spermatogenesis from stem cell spermatogonia, and epididymal transit of sperm, requires ca. 9 weeks (Amann, 1982). Thus, tests to evaluate the potential reproductive toxicity of a compound are generally lengthy, lasting from lo- 13 weeks (Lamb, 1988). Numerous strategies exist for evaluating the reproductive toxicity of chemicals, but as the chemical evaluation process (Heindel, 1988) or the study design (Gray et al., 1988) become more complex, the total time and cost of the process increase. This, in turn, reduces the number of chemicals that can be assessed for reproductive toxicity. The efficiency of the screening process can be improved by shortening the test, and/or getting more information per exposure. Effects on spermatogenesis can be identified by histologic evaluation without waiting for decreased sperm output to result in fewer offspring, thus shortening the exposure time. Simultaneous screening for both reproductive and developmental (R/D) endpoints would increase the information produced by the test without necessarily doubling the resources required. The design detailed below is an answer to a challenge to construct a test that would, in a short period of time (several weeks), be able to separate the more “potent” R/ D toxicants from chemicals that are less toxic. This design attempts to identify those chemicals in which R/D effects are predominant in the profile of toxicities. It exposes as many vulnerable processes as possible to putative toxicant effects within a time frame of 21 days. As such, the processes of spermatogenesis, oogenesis, mating, sperm motility, fertilization, implantation, and organogenesis and early postnatal development are all subject to chemical exposure and subsequent evaluation.
SHORT-TERM Study
REPRODUCTIVE
DESIGN
187
Days
Group A Females
Males
Group B Females
FIG. 1. A schematic of the design used for these studies. For each dose level, two groups of females are used: Group A is treated with test agent continuously throughout the experiment, while group B females are treated only during GD 8-14. The one group of males per dose level started treatment after 3 days of mating with group B females. The endpoints are provided in the text. Co, cohabit: gd, gestation day; _ , days of chemical treatment; Bi, birth; Net, necropsy.
Four chemicals were chosen to evaluate the design. All four had been tested previously by conventional developmental toxicity and Reproductive Assessment by Continuous Breeding (RACB) protocols. Two chemicals [boric acid (BA) and ethylene glycol monomethyl ether (EGME)] produced sterility in the RACB studies, and EGME has been the subject of numerous publications documenting the profound adverse effects on the male, the female, and the conceptus (Lamb et al., 1984; Foster et al., 1984; Hanley et al., 1984; Doe, 1984; Nagano et al., 198 1). The other two chemicals [theophylline (THEO) and ethylene glycol (EG)] were chosen because there were published data on their reproductive toxicity, and compared to EGME and BA, both produced relatively mild effects in the RACB studies (Morrissey et al., 1988; Lamb et al., 1985; Price et al., 1985). In addition, EG, BA, and EGME produced significant adverse effects on the developing embryo, while THEO appears slightly less toxic. Thus, the chemicals were chosen to span a spectrum of toxicity. The questions to be addressed by the present studies were: (1) would the design correctly identify the R/D toxicity of the more potent chemicals, and if so, (2) what type of effects would be found? In addition, (3) would THEO and EG produce adverse effects, and if so, how much change would there be? Finally (4) would the test distinguish developmental toxicity from reproductive toxicity? METHODS
AND
MATERIALS
Animals. Swiss Crl:CD-1 mice, 12- 14 weeks of age, were purchased from Charles River Lab (Portage, MI). Male and female mice were allowed to acclimate for at least 2 weeks prior to start of dosing. Each animal was uniquely identified with a numerical tatoo at the base of the tail and was individually housed. Food (NIH 3 I Rodent Chow, Purina Mills) and distilled deionized water were available ad libitum. The animal rooms were maintained under controlled conditions of temperature (23 + 1“C), humidity (50 f 10%) and lighting (12: 12, 1ight:dark).
Design(Fig. 1). One group of male (n = 10) and two groups of female mice (designated A and B) were used at each dose level. Untreated males and untreated females (B) were cohabited for 3 days to provide time-mated females (n = 10 pregnant females/dose) for treatment during organogenesis [Gestational Day (GD) g-141. The dams were allowed to deliver and litters were evaluated on Postnatal Days (PND) 0, 1, and 4. Treatment ofthe males began on Study Day 3 (SD 3) and continued until SD 20. Group A females (n = IO/dose) were treated from SD 0 until SD 20. On SD 8 these females were cohabited for 5 consecutive days with the males on test. This group of females (A) was killed on SD 21 by an overdose of carbon dioxide. The male mice were killed on SD 20. Body weights were recorded for all animals every fourth day during exposure and at necropsy. In addition, percentage pregnant and number of live implants were measured in the group A females. At termination of group B females and their litters on PND 4, data on the following endpoints were collected: number of females delivering/number plugged, number of implantation sites, litter size, and body weights on PND 0. 1, and 4. Since pregnancy can alter the microscopic structure of systemic organs, and could thus confound histologic interpretations, histopathology was not performed on the females. However, a more comprehensive examination was made of the male mice: organ weights of liver. right kidney, right testis, and right epididymis were measured. Histopathology of the liver, left kidney, and left testis and epididymis was performed following fixation in 10% neutral-buffered formalin of liver and kidney sections and Bouin’s fixative for the testis and epididymis. Tissues were embedded in paraffin and sectioned at 6 pm, and liver and kidney sections were stained with hematoxylin/eosin. while sections of testis and epididymis were stained with PAS/H. Total right cauda epididymal sperm counts were measured as described (Chapin et al., 1985). Epididymal sperm motility was evaluated by a modification ofthe method described by Dunnick et a/. (1986): rather than immediate evaluation, sperm were videotaped, and the tape was subsequently evaluated for moving/not-moving sperm. A total of at least 200 sperm were evaluated for each animal. The endpoints assessedin this 21-day design are compared with those measured in developmental toxicity and continuous breeding (RACB) studies and presented in Table 1. This table shows that the present design assesses many of the endpoints found in each of the other designs. The primary differences are in the durations of exposure: RACB studies expose F0 animals to the test agent for 15 weeks during mating and generate up to five litters (rev. in Lamb, 1988). Developmental toxicity studies traditionally expose pregnant females on GD 6-15. These exposure times were shortened to
188
HARRIS ET AL.
TABLE 1 Comparison of the Parameters of a Developmental Toxicity Study (DEV), Reproductive Assessment by Continuous Breeding Study (RACB), and the 21-Day Screen DEV
RACB
SCREEN
J
J
J
J
J
J
J
J
J
J
J
J
J
r,
J
J
-
J
J
-
General toxicity Body weight gain Terminal body weight Kidney weight Liver weight Mortality Water consumption Feed consumption Reproductive toxicity Fertility index Mean litters per pair Live pups/litter Live pup weight (F,) Days to litter Days between litters Testis weight Epididymis weight Cauda epididymis weight Prostate weight Seminal vesicle weight Testicular histology Cauda sperm density Sperm motility % Abnormal sperm Estrous cycle length Ovary weight Developmental toxicity No. implantation sites/dam Postimplantation loss Number of resorptions Number live pups Live pup weight Neonatal viability to Day 4 Neonatal death External malformations Internal malformations External variations Internal variations Weight at 21 days Fertility index
-
J
-
J J
J
J
J
J
-
J
J
-
J
-
J
J -
-
J
J
-
J
J
-
J
-
J
-
-
J
-
-
J
J
-
J
J
-
J
J
-
J
-
J
-
-
J
-
J
J
-
J
-
J
J
-
J
J
-
J
J
-
J
J
J
J
J
J
J
J
J -
J J
J
-
J -
J J
J
-
J
J
decrease the length and cost of the study. consistent with the intent of a “first-pass” study. Chemicals. Each chemical was tested at three dose levels, chosen based on literature-derived LD50 values, and not on previous in-house data. Approximately one-third of the reported LD50 was selected as the high concentration and a log difference for the low level; the middle dose was approximately one-third of the high level. All chemicals were administered daily by gavage at a volume of 0. I ml/IO g body weight. Dosing solutions were prepared by Radian Chemical Containment Laboratories, (Morrisville. NC). Boric acid (BA. CAS No. 10043-35-3) was administered in corn oil at 0. 120. 400, and 1200 mg/kg/day. EC (CAS No. 107-21-I) in water was
tested at 0, 250, 700, and 2500 mg/kg/day. EGME (CAS No. 109-86-4) was administered in water at 0, 70, 250, and 700 mg/kg/day. The fourth chemical, THEO (CAS No. 58-55-9) was tested at 0, 20, 60, and 200 mg/kg/day, using corn oil as the vehicle. All formulations were analyzed by gas chromatography after dosing and were found to be 93-100% of target concentrations. Histologic evaluation. Because of the shortened time frame, this study design relied heavily on histologic evaluation as a complement to the functional evaluations. All tissueswere assessedby observers unaware of treatment group. Liver and kidney sections were analyzed for deviations from “normal” by standard pathology criteria used in the National Toxicology Program. Deviations were graded as minimal, mild, and moderate. Testicular histology was evaluated by the criteria of Oakberg (1956). Eight control sections of testis from two of the studies were evaluated first, to provide a baseline from which to judge effects. The control slides were then placed back with their respective studies. and the testis sections were scored using the following classification system: 0 for no observable effect: I for minor changes in 20 to 50% of the tubules, consisting of increases in residual body size, and/or up to 25% cell exfoliation or death (necrosis), and/or slight decrease in cell number, or small visible changes in synchronization of cells within a stage (less than a 3 stage missmatch); 2 for bigger changes in timing (> a 3 stage discrepancy). and/or disruption/exfoliation in up to 50% of the tubules: 3 for exfoliation/disruption in up to 50% of tubules, or the presence of some end-stage tubules. and/or up to 50% overall germ cell loss; and 4 for >50% end-stage tubules, and/or massive germ cell loss. Stafistical analyses. Nonparametric statistical analysis was used for all endpoints. The pregnancy rates of group A females were evaluated using the Cochran-Armitage test for linear trend (Armitage, 197 I), followed by Fisher’s exact test (Reiss, 198 I) when significance was observed in the trend statistic. The remaining endpoints for group A females, as well as those for the group B females and the males, were analyzed using Kruskal-Wallis analysis of variance procedures to determine if overall differences existed between dose groups. Dose-response trends were assessedusing Jonckheere’s test against ordered alternatives (Hollander and Wolfe, 1973). When significance was noted in either the trend test or the Kruskal-Wallis ANOVA. pairwise comparisons were made using the Mann-Whitney U test (Hollander and Wolfe. 1973).
RESULTS Males
Control males in all four studies gained an average of 0.5% body weight during the course of each study. Weight gain was not inhibited by EGME or by BA; animals in the top dose group of the EG study lost 0.5% of their weight [no significant difference (NSD) from controls], and males in the top dose in the THEO study lost 1.2% of their weight (NSD). No males showed adverse clinical signs. There were no treatment-related deaths among males, although one died early in the study from gavage error. Administration of BA and EGME led to a reduction of testis weight (p < 0.05) after 19 days of dosing (Table 2). Neither chemical reduced epididymal weight, although epididymal sperm density and motility were reduced in the group receiving EGME at 700 mg/kg (p < 0.05, Table 2). Significant histopathology was noted in the testes of animals treated with BA and EGME (Fig. 2). The testes of mice receiving 250 mg EGME/kg/day (the middle dose group)
SHORT-TERM
REPRODUCTIVE
189
DESIGN
TABLE 2 Male Organ Weights, Sperm Parameters, and Histology Scores following Chemical Exposure in a 2 1-Day Reproductive/Developmental Screen Chemical”
Testis* weight (mg)
Epididymis* weight (mg)
No. sperm/g cauda. (X 106)
% Motile sperm
Histology score
EGME 0 70 250 700 Trend
130 + 5.6 130 * 7.3 85 f 5.6* 42 + 2,s p < 0.05
60.0 67.8 53.0 51.2
f f f f NS
5.1 7.0 6.5 7.4
917 f 167 1069 f 282 1347 f 480 380 f 97* p < 0.05
14.1 k 1.0 81.1 f 3.4 13.3 f 6.6 12.4 + 3.9* p < 0.05
0.5 0.1 2.1 5.0
zk f f f
0 250 700 2500 Trend
117 124 121 115
k 5.3 + 5.6 k 7.6 + 3.3 NS
44.6 45.8 45.2 43.8
k k + f NS
1.5 1.7 1.4 1.6
1179* III 1119* 57 1160+ 97 962 k 121 NS
90.6 87.5 89.0 89.5
0.4 0.2 0 I .o
i 0.2 + 0. I *o t 0.3
0 120 400 1200 Trend
1 I I i 3.8 I10 + 6.0 91 f 4.7” 87 k 5.5* p < 0.01
42.4 40.9 42.2 37.0
+ * * + NS
4.2 2.4 4.2 3.0
1761 1396 2368 972
f k f k NS
492 391 797 228
0.3 0.2 0.4 2.8
f f + *
0.2 0. I 0.2 0.2
118 119 112 123
46.0 45.9 44.9 54.1
+ 1.4 AC 3.2 * 3.3 f 3.2 NS
1080 1085 887 888
f f f + NS
101 116 118 134
0.2 0.5 0.2 0.7
f 2 f f
0.1 0.2 0. I 0.2
0.2 0.1 0.2 0.0
EC i k + f NS
2.2 2.7 3.2 2.3
BA ND ND ND ND
THEO 0 20 60 200 Trend
+I 5.5 F 5.0 * 4.9 f 5.3 NS
Note. Trend, test performed using Jonckheere’s test: NS. not significantly ’ Chemical dose in mg/kg/day. * Mean f SEM. )I = 9 for 1200 mg/kg BA; IO for all other groups. * Different from controls, p i 0.05.
showed extensive spermatocyte and spermatid cell loss, although complete spermatogenesis (through elongated spermatids) was present in many tubules (Fig. 2D). The highest dose of EGME resulted in near total germ cell loss (Fig. 2C); only scattered tubules contained any postspermatogonial germ cells in lO/ 10 testes. These effects are reflected in the histology scores (Table 2). The effects of BA were not as severe, but still marked: the high dose (1.2 g/kg/day) of BA caused exfoliation/disruption in >50% of tubules, with up to 50% germ cell loss in 7/9 mice receiving 1.2 g/kg/day for 19 days (Fig. 2B, Table 2). This was accompanied by asynchronous development: round spermatids were occasionally seen in late-stage tubules that should not contain these cells. The middle dose group of BA was generally indistinguishable from controls. THEO produced milder changes in the epithelium that consisted primarily of asynchronous germ cell development and focal loss of one or more generations of germ cells within individual tubules (Fig. 2E). Round spermatids were noted in stages X, XI, and XII; some of these
different
from
control;
78.5 81.3 81.8 80.4
ND,
f +k f NS
3.2 4.6 4.1 4.5
not determined,
round cells were undergoing karyolysis. Focal germ cell loss involved all germ cell generations, although usually only one generation was missing per tubular cross section. No treatment-related lesions were noted in testes from EG-treated mice. Histologic evaluation of liver and kidney sections could identify no treatment-related effects for any of the chemicals. Group A Females
No females in the EGME and EG groups showed adverse clinical signs. One THEO female and three BA females showed rough hair coat prior to death during chemical exposure; one of those females was euthanized moribund (Table 3). An effect (p < 0.05) on impregnation was found for EGME at the high dose level (Table 3). EGME also produced a slight and insignificant decrease in implant number. Females treated with the top dose of EG had significantly fewer live, and more dead implants, with 2/6 litters totally resorbed. None of the chemicals decreased the total number of im-
190
HARRIS ET AL.
FIG. 2. (A) Section of testis from a control mouse. All steps of germ cells were present. with minimal necrosis in stage XII. Scale bar, 30 pm. (B) Testis section from a mouse treated with I .2 g BA/kg/day. Note the round spermatids present with dividing spermatocytes (arrowheads) and the lack of elongating spermatids. Scale bar, 30 Frn. (C) Testis section from a mouse treated with 700 mg EGME/kg/day. Note the absence of postspermatogonial germ cells in most of the tubules. Scale bar, 30 +m. (D) Testis section from a mouse treated with 250 mg EGME/kg/day. Note the absence of round and elongating spermatids (tubule A) and a decreased number of germ cells in the remaining tubules. Scale bar, 30 pm. (E) Testis section from a mouse treated with 200 mg THEO/kg/day. Note the aberrant presence and position of round spermatids in conjunction with dividing spermatocytes (small arrowheads) or misplaced within the epithelium (large arrowhead. left). Scale bar, 30 grn.
plantation sites in these females, who had been mated during chemical exposure. Group B Females All of these females (impregnated prior to the start of dosing) had equivalent numbers of uterine implantations regardless of chemical treatment (Table 4). Treatment with EGME (250 and 700 mg/kg/day) or BA (1200 mg/kg/day) significantly reduced the number of females delivering and decreased the number of live pups. Although none of these chemicals caused neonatal mortality between PND 1 and PND 4, pups from mothers treated with EC during gestation had lower body weights on these postnatal days. DISCUSSION The data presented above demonstrate that the design employed in this study can identify chemicals that have significant R/D toxicity. Administration of BA or EGME resulted in significant effects on male reproduction and on the group B females (number of litters, number of young). Ethylene glycol administration also resulted in changes in postnatal litter weights. The ability of this design to differentiate
developmental toxicity from reproductive toxicity (effects on gametogenesis) is shown in the findings that fertilization and implantation were not affected (there was no change in the number of implantation sites), but the decrease in the number of pups in some groups (EC) shows that development wczsadversely affected. The chemicals used for these pilot experiments were chosen to represent two levels of effect, based on results of preceeding developmental toxicity and RACB studies. A summary of the effects noted in previous studies is presented in Table 5, where these effects are compared to those seen in the present study. As regards the fertility indices, both BA and EGME had produced complete infertility in previous RACB studies. With each chemical, males and females were affected (Fail et al., 1989; Lamb et al., 1984, respectively). In the RACB study, theophylline consistently decreased pup number in all dose groups, starting with the first litter (Morrissey et al., 1988) from a 22% decrease in the low dose group to a 42% decrease in the high dose group. The high dose THEO group in the RACB study also had lowered pup weight, increased F0 male liver weights, and a 20% reduction in sperm count. The high dose of EC, in the RACB study, produced an 18%
SHORT-TERM
REPRODUCTIVE
FIG.
DESIGN
191
2-Confinued
decrease in overall pups per litter and lower pup weight in all dose groups (Lamb et al., 1985) with no change in F0 male or female organ weights. Thus, in longer reproductive toxicity tests that utilized more endpoints and collected more data, BA and EGME were clearly more toxic to the reproductive systems than were THEO or EG. While there are occasional divergences in the data from the two study designs (see Table 5) it is noteworthy that the relative toxicity dif-
ferential previously seen in the RACB studies was also reflected by the data from the current design. Increased resorptions were seen after treatment with EG or BA in both the present study and in previous reports (Price et al., 1985; Heindel et al., 1991, respectively). Lindstrom et al. (1990) report increased resorptions after THEO exposure to Swiss mice at 372 and 396 mg/kg/day, doses considerably higher than those used in the present
192
HARRIS
FIG.
ET
AL.
2-Continued
study. Previous studies with EGME found increased fetal deaths and resorptions at higher doses (Nagano et al., 1981) while lower doses produced skeletal and visceral malfor .mations and variations. (Nagano et al., 198 1; Hanley et /al., 1984). It is perhaps not surprising that the developn nental toxicity studies identified structural malformatiot 1s and variations that were not identified by the screen . The design presented here is more closely related
to the Chernoff-Kavlock test (Chernoff and K avlock, 1982) than to a Segment II study: more reliance is; placed upon the relatively gross indices of pup weight, n umber, viability, and survival than on structural effects. T ‘he lack of pup/fetal dissection in the present studies mea ns that visceral or skeletal malformations or variations co uld not have been detected. Thus, a toxicant whose effect is only visible at dissection would not be identified as aL devel-
SHORT-TERM
REPRODUCTIVE
TABLE 3 Impregnation and Uterine Implant Data from Continuously Exposed Females following Chemical Exposure in a 2 1-Day Reproduction/Developmental Screen
DESIGN
193
change in the number of pups in the THEO 2 1-day study, while there was a clear effect on that endpoint in the RACB study, starting with the first litter. In addition, both chemicals produced effects in developmental toxicity studies that were not seen using the 21-day design (Table 5). No. live No. dead Total The question of sensitivity is also addressed by setting No. pregnant implants per implants per implants Chemical” (No. treated) femaleb femaleb per femaleb doses to assure some overt toxicity. The highest doses employed in the present studies were set at one-third the reEGME ported LD50. Compounds that produce no developmental 0 IO (10) 10.3 * 1.0 0.5 + 0.2 10.8 f 0.9 or reproductive toxicity at these doses would not be ex70 10.8 f 0.4 0.8 f 0.2 11.5 f 0.4 8 (10) pected to pose a risk of severe reproductive toxicity when 250 10 (10) 7.3 AI 1.8 3.8 f 1.4 11.1 kO.7 tested under other designs. This does not mean that there 700 6 (lo)* 6.7 f 1.9 0.2 + 0.2 6.8 f 2.0 Trend p < 0.05 NS NS NS would not br effects: rather, it indicates that this design is intended to identify those chemicals that are likely to be EC highly toxic to the conceptus and/or the reproductive sys0 10.0 f 0.9 0.4 * 0.2 10.4 & 1.0 9(10) 250 10.7 f 0.4 0.8 t 0.1 11.4 + 0.3 tem. Whereas positive effects in this design indicate that 9 (10) 700 11.1 ztzo.5 0.3 + 0.3 11.4 + 0.4 7(10) another, longer. test would also find R/D toxicity, the lack 2500 7.4 + 1.2* 1.4 t 0.4* 8.9 f 1.1 9 (10) of effects do not necessarily assure that a subsequent, Trend NS p < 0.05 p < 0.05 NS longer, test would also be negative. This is especially true BA for toxicants that bioaccumulate; such compounds would 0 10 (10) 12.3 f 0.6 0.3 2 0.1 12.6 + 0.5 be incorrectly identified as nontoxic. This is a drawback 120 12.1 + 0.5 0.4 ?I 0.2 12.6 f 0.4 9 (10) of any short-term evaluation. 400 12.7 + 0.6 0.3 +_ 0.2 13.0 f 0.5 9 (9) Sensitivity in detecting effects upon spermatogenesis 1200 9.8 f 3.1 3.0 + 2.3' 12.8 f 1.8 6 (7)[31 NS NS NS Trend NS might be enhanced by the use of glycol methacrylateembedded sections, which allows for improved visualizaTHEO tion of the seminiferous epithelium (e.g., Hess, 1990). By 0 10.0 f 0.5 0.4 i 0.2 10.4 + 0.4 9 (10) 0.3 f 0.2 10.3 + 1.1 20 10.0 f 1.1 9(10) using a better means of testicular evaluation, one need not 60 10.4 * 0.4 1.0 f 0.4 11.4 + 0.3 9 (10) treat the animal for the full cycle of spermatogenesis and 1.2 + 0.7 11.2 f 1.2 200 10.0 f 1.1 6 (9)Ill epididymal transit time, and then evaluate pups produced Trend NS NS NS NS from those sperm, to detect an effect upon spermatogenesis. Rather, high-resolution light microscopy can allow Note. Trend, test performed using Jonckheere’s test; NS, not significantly different from controls; [n], number of mice dying during chemical exposure. an evaluation of all the cells in the testis after a short exa Chemical dose in mg/kd/day. posure, giving similar information in a much reduced time b Mean f SEM. frame. It should be noted that genetic toxicities will re’ Two of six litters totally resorbed. main undetected by histologic means alone, but might well * Different from controls, p < 0.05. be identified by other short-term i~z 11ivo or in vitro tests. This design has a limited ability to address behavioral opmental toxicant. While this study identified only EG and BA as developmental toxicants, much higher doses of effects. Compounds affecting libido and/or mating behavior would cause a decrease in the number of pregnant THEO have been required to demonstrate fetal toxicity. group A females. While additional tests would be required And while EGME was not found to be developmentally toxic, it’s antifertility effects were clear; EGME was cor- to identify the affected process, an effect would have been found, which could be considered one key requirement of rectly identified by the present design as toxic. any pilot study. Other shortcomings of a short-term design One of the primary questions of a new testing approach is, will it identify chemicals that are developmental or re- include the inability to identify second-generation changes productive toxicants? This design appears to do that. An- in fertility or behavior and the fact that the dosing period other question is, where is the limit of sensitivity, or how is too short to allow for a full expression of testicular toxicity to be manifest as large testicular weight loss or large toxic does a chemical need to be in order to be detected? increases in abnormal sperm morphology. Indeed, because One might logically expect a decrease in sensitivity with the time frame of the study is so short relative to spera short period of dosing and only one mating period. This is borne out by the results with EG and THEO; the effects matogenesis, sperm morphology was considered to be a evaluation of of both chemicals were more profound in the RACB stud- poorly responsive endpoint. Morphologic ies than in the present design. For example, there was no sperm was omitted from these studies, but can be added
194
HARRIS
ET
AL.
TABLE 4 Neonatal and Uterine Implant Data for Females Exposed during Gestation following Chemical Exposure during a 21-Day Reproduction/Developmental Screen Total
No. live neonates Chemical”
No. females littering (No. Rx)
PND
Ob
PND
1
PND
4
PND
litter
1
wt. PND
No. implantation sites per female
4
EGME 0 70 250 700 Trend
9 (10) 11 (11) 12 (12) O(ll)
0 250 700 2500 Trend
12 12 12 14
0.5 0.4 0.7 0.6
NS
1 I .6 -+ 11.5 * 11.5 * 12.4 + NS
19.2 f 1.1 19.8 + 0.8 19.6 + 0.7 15.3 f 0.9* p < 0.05
31.7 I 1.3 34.1 k 1.1 33.4 f 1.2 26.6 f l.5* p < 0.05
12.0 k 1 I .9 + 12.5 zk 11.9 f NS
0.5 0.5 0.4 0.5
10.9 f 0.6 I I .2 f 0.7 10.4 f 0.4 o.o* p < 0.01
20.3 k 0.8 20.2 f 0.8 19.6 k 0.8 -
33.5 rt 1.3 33.5 i 1.2 33.4 k 1.4
NS
NS
I I.0 f 0.6 11.2+0.7 10.8 f 0.4 1 I .3 + 0.5 NS
11.4 * 0.4 9.3 + 0.8 11.0*0.7 10.8 f 0.5 NS
20.4 17.8 19.9 18.7
10.9 t 0.6’ 10.7 * 0.4 0.4 f 0.2* o.o* p -=z 0.01
10.9 + 0.6 10.7 f 0.4 o.o* o.o* p < 0.01
10.6 k 0.7 10.6 f 0.3 o.o* o.o* p < 0.01
18.1 + 1.0 16.8 k 0.9 -
28.6 k 2.1 24.9 + 1.9 -
NS
11.3 11.3 11.0 10.3
+_ 0.6 + 0.6 * 1.0 F 0.7 NS
11.2kO.6 1 I .3 * 0.6 10.9 + I.0 10.2 t 0.7 NS
11.2 11.3 10.9 10.1
? 0.6 Y!Z0.6 f 1.0 t 0.7 NS
10.9 + 0.6 11.2kO.7 10.7 * 0.4 o.o* p < 0.01
10.9 + 0.6 1 1.2 * 0.7 10.4 * 0.4 o.o* p < 0.01
11.4 9.4 11.0 10.9
11.4 + 0.4 9.3 + 0.8 11.ot0.7 10.8 + 0.5 NS
EG (12) (12) (12) (14)
BA 0 120 400 1200 Trend
10 (10) 11 (II) 12 (12) O(11)
THEO 0 20 60 200 Trend
14 13 15 14
(14) (13) (15) (14)
NOIE. Trend, test performed using Jonckheere’s ’ Chemical dose in mg/kg/day. b Postnatal day. c Mean f SEM. * Different from controls, p < 0.05.
* 0.4 + 0.8 * 0.7 + 0.5 NS
test; NS, not significantly
to studies of other chemicals as desired. Additionally, the design uses only 10 animals per group. It has been shown that, especially for reproductive endpoints, increasing the number of animals increases the quality of the data (Schwetz et al., 1980). This is also obvious from a statistical viewpoint. Thus, this design is not as sensitive as much longer tests, but it also costs considerably less and takes less time. The optimal design for a given situation depends on the question being asked, the resources available, and the confidence required in the answer. Additional endpoints can be added to address specific interests or applications. For example, the NTP is using this design to evaluate AIDS therapeutics and has added immunotoxicology endpoints to identify possible immune system effects. Additional reproductive or systemic endpoints could be added, such as computer-assisted sperm motion analysis or other organ-specific tests. We have
different
from
controls;
k + + k NS
0.8 0.5 1.3 0.9
No. Rx. number
34.8 31.4 32.9 32.5
f ? f + NS
1.2 0.8 1.5 1.2
of females
12.2 11.1 12.0 Il.3
treated.
t It + + NS
0.5 0.3 0.6 0.5
in parentheses.
housed mice in metabolism cages for the last 24 hr of dosing to assess possible kidney changes. Specific nonlethal developmental abnormalities could be identified by terminal dissection of pups on PND 4. We view the approach detailed above as a basic model to which options can be added as appropriate. Finally, the usefulness of these data will depend largely upon the context in which they are generated. We view this test as being most useful for generating preliminary data that will ultimately become part of a larger data set. For example, a compound of unknown toxicity that produces some developmental toxicity in this design would then be subject to a more complete developmental toxicity assessment; the focus would be on the developmental toxicity rather than female or male reproductive toxicity. The design could also be used to rank the toxicity of members of a structurally related chemical class. The data from this screen would thus
SHORT-TERM
REPRODUCTIVE
TABLE 5 Comparison of Chemical Effects in Reproductive Assessment by Continuous Breeding (RACB), 21-Day Screen, and Developmental Toxicity Studies EGME
EG
BA
Reproductive endpoints Body weight Kidney weight Liver weight Mortality Fertility index Live pups/litter Pup weight Days to litter Testis weight Cauda weight Epididymis weight Testis hisology No. sperm/g cauda Sperm motility Live births Neonate death Viability External malformations
Aadb NC/NC t/t NC/NC $14 J/J NC/NC NC/NC $14 NC/NC $/NC t/t $14 CA 414 t/t &/NC NC/NC
NC/h NC/NC NC/NC NC/NC NC/NC NC/NC 414 NC/NC 4/J NC/NC NC/NC t/t NC/NC $/NC NC/NC NC/NC NC/NC f/NC
J/t4 NC/NC +/NC t/t J-14 J/JJ/NC NC/NC i/c J-/NC h/NC t/t J/NC i/ND r/J NC/NC NC/NC NC/NC
e/JNC/NC C/NC NC/NC NW $/NC NC/NC NC/NC NC/NC NC/NC NC/NC ND/t h/NC NC/NC h/NC t/NC NC/NC NC/NC
Developmental endpoints Body weight Liver weight Kidney weight Mortality Implantation sites/dam Number of resorptions Malformation& Variations
NCc/Jb ND/t ND/NC NC/NC NC/NC NC/NC NC/NC t/NC
414 $/NC NC/NC NC/NC NC/NC t/t f/NC t/ND
w NC/NC t/NC NC/t NC/NC
4/c NC/NC ND/NC NC/NC NC/NC
t/t
t/NC h/ND
and and Kim this
Dr. Bernard Schwetz and the other members of the Developmental Reproductive Toxicology Group (Drs. Rick Morrissey, Jerry Heindel, Treinen) for stimulating discussions during the formative stages of project.
REFERENCES
THEO
,N::!: t/ND
Note. NC, no change compared to controls; ND, not determined; tNC, equivocal increase; t, increased, compared to controls; 4, decreased, compared to controls; EGME, ethylene glycol monomethyl ether: EG, ethylene glycol: BA. boric acid; THEO, theophylline. a RACB data. ’ 2 I -Day screen data. ’ Developmental toxicity. d Only external malformations were assessedin the screen.
permit the deployment of resources toward the highestpriority questions. While no short-term test can replace one using a longer exposure or a complete developmental assessment, it appears that this design shows promise for being able to detect those chemicals that have significant reproductive and/or developmental toxicity. Continued use of this design will improve our understanding of the merits and faults of this test and allow for its use in the most efficient manner.
195
DESIGN
Amann, R. P. ( 1982). Use of animal models for detecting specific alterations in reproduction. Fundam. Appl. Toxicol. 2, 13-26. Armitage, P. (197 I). Cochran-Armitage trend in proportions. In Statistical Methods in Medical Research, pp. 362-365. Blackwell, Great Britain. Chapin. R. E., Dutton, S. L., Ross, M. D., and Lamb IV, J. C. (1985). Effects of ethylene glycol monomethyl ether on mating performance and epididymal sperm parameters in F344 rats. Fundam. Appl. To,xicol. 5, 182189.
ChemotT N., and Kavlock, R. J. (1982). An in vivo teratology screen utilizing pregnant mice. J. To.xicol. Environ. Health 10, 541-550. Doe, J. E. (1984). Ethylene glycol monoethyl ether and ethylene glycol monoethyl ether acetate teratology studies. Environ. Health Perspect. 57, 3341. Dunnick, J. K.. Harris, M. W.. Chapin, R. E., Hall. L. B., and Lamb IV, J. C. ( 1986). Reproductive toxicology of methyldopa in male F344/N rats. Toxicology
41, 305-318.
Fail, P. A.. George. J. D., Grizzle, T. B., and Heindel, J. J. (1989). Reproductive toxicity of Boric Acid (BORA) in mice as evaluated by the Continuous Breeding protocol. Toxicologist 9, 266. Reiss, J. L. (198 1). Statistical Methods,for Rates and Proportions, 2nd ed., pp. 25-29. Wiley, New York. Foster. P. M. D.. Creasy, D. M., Foster, J. R., and Gray, T. J. B. (1984). Testicular toxicity produced by ethylene glycol monomethyl and monoethyl ethers in the rat. Environ. Health Perspect. 57, 207217.
Gray, L. E.. Jr., Ostby, J., Sigmon. R., Ferrell, J.. Rehnberg, G., Linder. R.. Cooper, R.. Goldman, J., and Laskey, J. (1988). The development of a protocol to assessreproductive effects of toxicants in the rat. Reprod. Toxicol.
2, 281-287.
Hanley, T. R., Jr., Yano, B. L., Nitschke, K. D.. and John J. A. (1984). Comparison of the teratogenic potential of inhaled ethylene glycol monomethyl ether in rats, mice and rabbits. Toxicol. Appl. Pharmacol. 75,409422.
Heindel, J. J. (1988). The National Toxicology Program chemical nomination selection and testing process. Reprod. To-xicol. 2, 273-279. Heindel, J. J., Price. C. J., Field, E. A., Marr, M. C., Myers, C. B., Morrissey, R. E.. and Schwetz, B. A. (199 I ). Developmental toxicity of boric acid in mice and rats. Fundam. .4ppl. Toxicol. 18, 266-277. Hess, R. A. (1990). Quantitative and qualitative characteristics of the stages and transitions in the cycle of the rat seminiferous epithelium: Light microscopic observations of perfusion fixed and plastic-embedded testes. Biol. Reprod.
43, 525-542.
Hollander, M.. and Wolfe, D. A. (1973). Nonparametric Statistical Methods, pp. 27-33 and 120-123. Wiley, New York. Lamb IV, J. C., Gulati, D. K., Russell, V. S., Hommel, L.. and Sabharwal, P. S. (1984). Reproductive toxicity of ethylene glycol monoethyl ether tested by continuous breeding of CD-1 mice. Environ. Health Perspect. 57,85-90.
ACKNOWLEDGMENTS We are grateful to Mr. Brad Collins (NTP) and Dr. Bob Moseman (Radian) for assistance with dosing solution formulation and analysis
Lamb IV, J. C., Maronpot, R. R., Gulati, D. K., Russell, V. S., HommelBarnes, L., and Sabharwal. P. S. (1985). Reproductive and developmental toxicity of ethylene glycol in the mouse. Toxicol. Appl. Pharmacol. 81, 100-I
12.
196
HARRIS ET AL
Lamb IV, J. C. (1988). Fundamentals of male reproductive toxicity testing. In Physio1ug.vand To.uicologv qfMale Reproduction (J. C. Lamb IV and P. M. D. Foster, Eds), p. 137-154. Academic Press, San Diego. Lindstrom. P.. Monissey, R. E., George, J. D., Price, C. J., Mat-r, M. C., Kimmel. C. A., and Schwetz, B. A. (1990). The developmental toxicity of orally administered theophylline in rats and mice. Fundam. iippnpl.To.*-id
14, 167-178.
Morrissey, R. E.. Collins, J. J., Lamb IV, J. C.. Manus, A. G., and Gulati, D. K. ( 1988). Reproductive effectsof theophylline in mice and rats. Fundam. .4pp/. Tosicol. 10, 525-536. Nagano. K., Nakayama, E., Oobayashi, H.. Yamada, T., Adachi. H., Nishizawa. T.. Ozawa, H., Nakaichi. M., Okuda, H., Minami, K.. and Ya-
mazaki, K. ( 198 1). Embryotoxic effects of ethylene glycol monomethyl ether in mice. To.rico/ogy 20, 335-343. Oakberg, E. F. (1956). A description of spermiogenesis in the mouse and its use in analysis of the cycle of the seminiferous epithelium and germ cell renewal. Am. J. Anat. 99, 391-413. Pederson, T. (1970). Follicle kinetics in the ovary of the cyclic mouse. Acta Endocrinol. 64, 304-323. Price, C. J.. Kimmel, C. A., Tyl, R. W., and Marr, M. C. (1985). The developmental toxicity of ethylene glycol in rats and mice. To.xico/. Appl. Pharmacol. 81, I I 3- 127. Schwetz, B. A., Rao, K. S.. and Park, C. N. (1980). Insensitivity of tests for reproductive problems. J. Erwiron. Pathol. Toxicoi. 3, 8 l-98.
