Reproductive Toxicology, Vol. 11, No. I, 29-36, 1997 Copyright 0 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0X90-6238197 $17.00 + .OO ELSEVIER
DISRUPTION OF SPERMATOGENESIS IN RABBITS CONSUMING ETHYLENE GLYCOL MONOMETHYL ETHER
WILLIAME.BERNDTSON* and ROBERT H.FOOTE~ of Animal and Nutritional Sciences, University of New Hampshire, Durham, New Hampshire TDepartment of Animal Science, Cornell University, Ithaca, New York
Abstract-Effects of ethylene glycol monomethyl ether (EGME) on spermatogenesis were examined using groups of six to seven Dutch rabbits that received 0, 12.5, 25.0, 37.5, or 50.0 mg of EGME/kg body weight, respectively, in their drinking water Sd/week. After 12 weeks, animals were euthanized and their testes were removed, weighed, and processed to permit germ cell numbers to be quantified. Spermatogenesis was depressed by EGME in a dose-dependent manner; numbers of round spermatids per Sertoli cell (a measure of spermatogenie efficiency) averaged 8.86, 8.87, 6.20, 2.38, and 7.42 for the 0, 12.5, 25.0, 37.5, and 50.0 mg/kg dosages, respectively. The latter value of 7.42 represents an overestimation of sperm production because it is based on only two unexpected outlier rabbits. Nearly complete destruction of spermatogenesis occurred in the other five animals in this highest dosage group, precluding evaluation by the histometric method. Numbers of homogenization-resistant elongated spermatids per testis, measurable on all animals, averaged 231,256, 195, 52, and 67 x 106, respectively. The correlation between the predicted sperm production, based on the elongated spermatids at necropsy, and the number of sperm ejaculated by the males during week 12 was 0.92. Thus, EGME impaired rabbit spermatogenesis in a dose-dependent manner. Generally, rabbit spermatogenesis was at least 10 times more sensitive to EGME than previously reported for rats and mice. 0 1997 Elsevier Science Inc. Ke)] Wur-ds; EGME; rabbits; spermatogenesis;
is of importance
human health and fertility
is a problem
The Environmental Protection Agency (7) has increased efforts to develop new models and new procedures to enhance the ability to evaluate the effect of environmental agents on male reproductive function. One aspect addressed was to develop multiple species comparisons and particularly to utilize a species where repeated semen collections were possible. The rabbit is well suited for this purpose and, in addition, provides a nonrodent model to compare with the extensive studies performed in rats and mice (8). Other aspects include the development or identification of the most sensitive and reliable criteria by which one may evaluate agents for potential reproductive toxicity. As a potent derivative of the ethylene glycols, EGME was considered a potentially useful model agent by which to assess the relative sensitivity of specific reproductive end points to this class of chemicals. EGME had been shown to be a potent reproductive toxicant in several species, but few studies had been conducted with the male rabbit. Based on data from a subchronic inhalation study, Miller et al. (9,lO) concluded that rabbits were more sensitive to EGME than rats. However, the effects of EGME had not been examined when administered orally in this species, quantita-
for many couples of reproductive
age, with about one-half of this infertility reportedly attributable to the male partner. Although a number of factors or combinations of factors may be involved, many cases of infertility have been attributed to a low sperm count, and many authorities believe that sperm production is only marginally adequate in a high proportion of adult men (l-5). Spermatogenesis is a complex process that appears readily susceptible to disruption by a variety of physical and chemical agents. Indeed, severe impairments of spermatogenesis have often been observed after exposure to known or suspected toxicants in the absence of detectable effects on other organs or body systems. If human sperm production is already limited, additional decreases resulting from environmental exposure to antispermatogenic agents could cause sperm production to become deficient for the exposed population. Thus, the investigation of chemicals for potential toxicity
Address correspondence to William E. Bemdtson, Department of Animal & Nutritional Sciences, University of New Hampshire, Durham, NH 03824. Received 30 April 1996; Revision received 1 I September 1996; Accepted 9 October 1996. 29
tive histometric assessments of germ cell numbers were lacking, and estimates of sperm production derived from testicular evaluations had not been compared to actual sperm output. Such information was of interest in enabling comparisons of the relative sensitivity of the rabbit to that of other species, or of rabbits administered EGME orally or by inhalation. Such information was also needed to identify the specific points in spermatogenesis and types of germ cells that are susceptible to disruption by this chemical. Therefore, the primary objectives of this study were 1) to characterize and quantify, via several different end points, the changes in spermatogenesis in rabbits consuming EGME in their drinking water, 2) to compare these to the corresponding changes in seminal ch~acteristics and fertility of these same rabbits [reported in a companion manuscript (I I)], and 3) to compare the responses of rabbits to those reported for other species.
Animak Thirty-one sexually mature Dutch rabbits of approximately 2 kg body weight were selected from among a larger group raised in an AAALAC approved facility maintained at 19 to 21°C under a 1212 1ight:dark cycle. All males had been trained to ejaculate into an artificial vagina, and had been judged to be normal based on libido, seminal quality, and testicular palpation. All animals were ranked, based on this pretest infonnntion and body weight, and were assigned to one of two replicates of 16 and 15 animals, respectively. Individuals from each successive group of five animals in rank within each replicate were assigned at random to each of the five treatment groups, to receive 0, 12.5, 25.0, 37.5, and 50.0 mg EGME per kg of body weight, as recently described (I 1). A sixth animal was assigned at random to the high est dosage group in one replicate. The two replicates were conducted sequentially.
EGME administration Each rabbit was weighed on Monday mornings, and the weight was used to calculate the dosage of EGME required for that week. EGME was administered in the drinking water on 5 d per week (Monday through Friday) by the system developed by Foote et al. (12), avoiding the negative effect of gavaging on sexual behavior. For this, a 0.1% (v/v) EGME solution was prepared by adding 107 pL of EGME solution (Sigma Chemical Co., St. Louis, MO, catalog number E5378, containing 930 mg of EGME/mL) to 100 mL of water. This solution, or dilutions of it, was administered to provide the appropriate dosage of EGME in a volume of approximately 100 mL
11,Number I, 1997
of drinking water, as described previously (11). EGME was administered 5 d per week for 12 weeks. Necrupsy and tissue processing procedures After 12 weeks of treatment, animals were weighed and euthanized with an overdose of Surital (thiamylal sodium, Park-Davis Co.) administered via the ear vein. Testes were removed and weighed. One testis from each rabbit, chosen on an alternating basis, was placed in Zenker-form01 solution. After about 30 min, the poles were removed to enhance penetration of the fixative, and the tissue was returned to the fixative for a total of 24 h. Then, the tissue was rinsed in running tap water for 24 h, and stored in 70% ethanol for at least 24 additional h, prior to subsequent processing for histologic evaluation. The remaining testis was placed in a vial and stored frozen until thawed for use in determining the numbers of homogenization-resistant spermatids (13).
Portions of fixed tissue were dehydrated, embedded with parafin, sectioned at 4 pm, stained with periodic acid Schiff’s reagent, and counterstained with hematoxylin. Tissues from each rabbit were scanned at random until 20 round tubular cross sections were identified at stage I of the cycle of the seminiferous epithelium, according to the classification system of Swierstra and Foote (14). The numbers of Sertoli cell nuclei with nucleoli and germ cell nuclei present within each tubular cross-section were determined by direct counts (15). Because these “crude counts” contained both whole nuclei (or Sertoli nucleoli) and fragments produced by sectioning, values were converted to “true counts” (whole nuclei or nucleoli equivalents) by application of Abercrombie’s formula (15,16).The mean nuclear or nucleolar dimensions required for application of Abercrombie’s equation were obtained for each animal by measuring five cells of each type with an ocular micrometer. These measurements were used to convert crude counts to true counts for each individual animal. Because the numbers of cells per tubular cross-section are influenced by seminiferous tubular dimensions, which can be altered due to experimental treatment or by tissue processing, germ cell numbers were Sertoli cell corrected (15) by expression on a per Sertoli cell basis. Additionally, the ratios of specific germ cells to their progenitor cell types (germ cell:germ cell ratios) were calculated from true counts per Sertoli cell to enable assessments of spermatogenic efficiency. Tt should be noted that the germinal epithelium of stage I seminiferous tubules is comprised of spermatogonia and two generations of both primary spermatocytes and spermatids (14). The younger and older generations of stage I primary spermatocytes are at the preleptotene and pachytene phases of develop-
EGME and Spermatogenesis
0 W. E. BERNDTSONand R. H. FOOTE
ment, while the two generations of spermatids consist of young, round spermatids and more developmentally advanced elongated spermatids. Elongated spermatids were not counted because their shape and arrangement precludes accurate counting and because Abercrombie’s correction is only applicable for spherical structures.
Statistical analyses Data for each variable were subjected to a simple one-way analysis of variance to test for the main effect of dosage (0.0, 12.5, 25.0, 37.5, and 50.0 mg EGMEPg). Where a significant treatment effect was found, differences among individual group means were tested by the Student-Newman-Keuls method. Additionally, relationships among various selected end points of interest were examined by linear regression analyses.
Quantification of homogenization-resistant spermatids The numbers of homogenization-resistant spermatids per testis were determined by the method of Amann and Lambiase (13). For this, one previously frozen testis from each rabbit was thawed, decapsulated, and the weight of the capsule and parenchyma was determined. The entire parenchyma was transferred to a Waring blender microhomogenization vessel containing 10 mL of 0.9% NaCl:0.05% (v/v) Triton X- 100 homogenization fluid, and homogenized for 1 min. The total volume of each homogenate equaled 10 mL plus the volume of the parenchyma. The testicular parenchyma was assumed to have a specific gravity of 1.O (I 7) to convert weight to volume. An aliquot of each sample was diluted further (usually sixfold), as necessary, to provide a suitable concentration of spermatids for counting via hemacytometry. Based on recommended procedures derived from critical analyses with the bull and stallion (18), homogenization-resistant spermatids were enumerated in duplicate by each of four independent evaluators. Evaluators performed additional counts where duplicate values differed by more than 10 to 12%. These data were converted to estimated daily sperm production (13) to allow comparison to the number of sperm collected per male during week 12, previously reported (11).
RESULTS Germ cell numbers The effects of EGME on spermatogenesis are summarized in Table 1. Severe disruption of spermatogenesis in five of seven rabbits that had received the 50.0 mg/kg dosage precluded the identification of stages and the ability to obtain cell count data by the histometric method. Additionally, some disrupted tubules were found in one, one, two, and five of the six rabbits in the 0.0, 12.5, 25.0, and 37.5 mg/kg groups, respectively. While occasional disruption might be expected even among normal, untreated individuals, the incidence and severity of disruption was clearly dose related. Where such disruption precluded identification of the stage of the cycle of the seminiferous epithelium, the magnitude of EGME effects will be underestimated to the extent that cell count data in Table 1 do not include data for the most severely affected individuals or reflect conditions within a select population of tubules within individuals with some tubular disruption. Data for the highest dosage are clearly biased by these factors, as they were neces-
EGME (mg/kg) Variable Bucks with assayable testes Assayable testes with disrupted tubules Germ cells/Sertoli cell Spermatogonia Young 1” spermatocytes Old 1” spermatocytes Round spermatids Germ cell:germ cell ratios Round spermatids/old 1” spermatocytes Round spermatids/young 1” spermatocytes Round spermatids/spermatogonia‘ Old 1” spermatocytes/young lo spermatocytes Old lo spermatocytes/spermatogonia Young 1” spermatocytes/spermatogonia Bucks with assayable testes Elongated spermatids/testis (x10’)
0.17 kO.02 2.54 + 0.13 2.43”*0.14 8.86” f 0.58
0.16+0.01 2.52 + 0.13 2.47” + 0.21 8.87” & 0.59
3.67” 3.51” 54.0* 0.96 15.0 15.6
f f * + f f
0.19 0.25 4.10 0.05 1.45 1.33
616 231.2” f. 48.5
“,hR~~ means without superscripts or bearing similar superscripts at P < 0.05 by the Student-Newman-Keuls method. ‘One testis sample was inadvertently lost.
25.0 Histometric data 616 216 0.16 * 2.06 + 1.87”.b f 6.20” *
0.02 0.21 0.18 1.2 1
3.22” f 0.40 3.63” * 0.17 2.95” f 0.38 3.52” * 0.14 39.25” f 5.56 57.2”? 5.62 0.92 f 0.02 0.98 ? 0.05 12.1 f 0.72 15.8 + 1.59 13.3 * 0.99 16.2 f 1.31 Homogenization-resistant elongated 515’ 616 195.0” f 48.7 256.1” f 22.6
0.16 1.91 1.42b 2.38h
? f * +
0.02 0.18 0.15 0.81
1.63b + 0.46 1.22b * 0.43 14.2h + 4.25 0.76 * 0.09 10.1 f 1.70 13.3 * 1.78 spermatids 616 51.7b f 20.5
0.22 f 2.14 f 2. 10a.b f 7.42” f
0.03 0.02 0.08 0.66
3.54”f0.19 3.46” f 0.27 34.8” f 7.72 0.98 f 0.02 9.8 ? 1.67 9.95 + 1.47 717 67.0b + 36.8
did not differ (I’ > 0.05). Row means that do not bear a similar superscript
sarily based on two seemingly unaffected individuals from a group within which all five others were severely affected. Among animals in the remaining groups, no effect (P > 0.05) of EGME was observed on the numbers of spermatogonia or young primary spermatocytes per Sertoli cell, while the numbers of old primary spermatocytes and round spermatids were depressed to 58.4 and 26.9% of control values, respectively, by the 37.5 mg/kg dosage. By contrast, disruption of the seminiferous epithelium does not preclude assessment of sperm production rates via the homogenization method, by which it was determined that the mean numbers of elongated spermatids were reduced (P < 0.05) to approximately one-fourth of control values by the 37.5 and 50.0 mg/kg doses (Table 1). It should be noted that the similarity in the mean numbers of elongated spermatids between the 37.5 and 50.0 mg/kg groups was attributable to the presence of the two unaffected outliers within the 50.0 mg/kg group. Indeed, numbers of elongated spermatids per testis in the five affected animals from the 50.0 mg/kg group equalled only 6.4, 7.0, 15.2, 7.7, and 13.5 million, respectively, and averaged only 4.3% of the corresponding mean for the untreated controls. Because the numbers of germ cells of each type were quantified, it was possible to estimate spermatogenie efficiency, as each division should result theoretically in a doubling in the number of progeny cells. By comparing the actual cell yield against the theoretical expectation in control and treated animals, the efficiency and sensitivity of a given step or steps of spermatogenesis may be estimated. In five of the seven animals receiving the highest dose of EGME, there were few cell types remaining. Thus, the two higher doses of EGME
severely depressed sperm formation. The present data available for each of the remaining dosages (Table 1) reveal a primary influence of EGME on the more developmentally advanced germ cells of the rabbit. Specifically, among all possible germ cell:germ cell ratios, those between round spermatids and their precursor types were depressed (P < 0.05) most severely by treatment. For the group receiving 37.5 mg/kg, the yield of round spermatids per old primary spermatocyte, per young primary spermatocyte, or per spermatogonium averaged only 44.4, 34.8, and 26.3% of control means, respectively. Also, the number of old primary spermatocytes per Sertoli cell in the 37.5 mg/kg group was only 58.4% of the control. At the lower doses studied, yields (ratios) of old primary spermatocytes per young primary spermatocyte or of either old or young primary spermatocytes per spermatogonium (Table 1) did not differ (P > 0.05). Relationships among testicular parameters The correlation and regression relationships among selected testicular parameters are summarized in Table 2. In Table 2, histometric data for the 50 mg/kg group were excluded from most regression analyses, because the data for the two available rabbits were clearly unrepresentative. As anticipated, most measures of spermatogenesis were moderately or highly correlated with each other (Table 2). For example, the predicted sperm production based upon numbers of homogenization-resistant elongated spermatids was highly correlated (r = 0.92, P < 0.05) with estimates based on either sperm output (7) during the last week of the experiment (7) or on numbers of round spermatids per Sertoli cell (r = 0.84, P < 0.001). But, because germ cell losses throughout sper-
Table 2. Relationship between selected reproductive variables in EGME-treated rabbits Variable Dependent
Testis weight (for histology) Spermatids/testis Spermatids/testis Spermatids/testis Spermatids/testis Spermatids/testis
Testis weight (for homogenization) Spermatogonia/Sertoli Young 1” spermatocytes/Sertoli Old 1” spermatocytes/Sertoli Round spermatids/Sertoli Testis weight (for homogenization)
Spermatogonia/Sertoli Young 1” spermatocytes/Sertoli Old lo spermatocyte/Sertoli Round spermatids/Sertoli Round spermatids/spemratogonia Round spermatids/young 1’ spermatocyte Round spermatids/old 1’ spermatocyte Old lo spermatocytes/young lo spermatocyte Old 1” spermatocytes/spermatogonia Young lo spermatocytes/spennatogonia
Testis Testis Testis Testis Testis Testis Testis Testis Testis Testis
weight weight weight weight weight weight weight weight weight weight
(for (for (for (for (for (for (for (for (for (for
histology) histology) histology) histology) histology) histology) histology) histology) histology) histology)
Regression y y y y y y y y y y y y y y y y y
= = = = = = = = = = = = = = = = =
0.088 + 0.904 x 63.30 + 744.~52 135.75*x x -125.17 + .: -111.92 + 143.25 -14.14 + 29.33 x* -112.60 + 170.53 x -124.95 + 178.40 x* 0.135 + 0.017 x* 1.58 + 0.41 x* 0.84 + 0.73 x* -1.01 + 4.58 x* -4.27 + 27.40 x* 0.019 + 1.677 x* 0.78 + 1.36 x* 0.623 + 0.169 x* 7.57 + 3.43 x* 13.031 + 0.720 x*
0.962 0.278 0.566 0.739 0.839 0.933 0.934 0.229 0.469 0.680 0.756 0.726 0.761 0.656 0590 0.472 0.1 I I
0.922 0.037 0.287 0.525 0.690 0.866 0.866 0.013 0.184 0.438 0.553 0.506 0.560 0.404 0.319 0.187 -0.029
Analyses with data for rabbits 103D and 104D (from the 50 mg/kg dosage group) deleted. In many instances, the deletion of these two rabbits resulted in the total absence of any data for the SO mg/kg dosage group. Such analyses are denoted by an asterisk (*) in the regression equation column.
EGME and Spematogenesis
matogenesis are cumulative, estimates of sperm production were less reliable when based on the numbers of less advanced germ cell types. For example, whereas the correlation between the number of homogenization-resistant elongated spermatids and the number of round spermatids per Sertoli cell was 0.84, the corresponding correlations with the numbers of old primary spermatocytes, young primary spermatocytes, and spermatogonia per Sertoli cell, were only 0.74, 0.57, or nonsignificant, respectively.
DISCUSSION Spermatogenesis would appear to be among the most highly sensitive and selective targets of EGME in the rabbit. As presented previously (1 l), EGME did not alter the general health of these rabbits, as evidenced by body weight, general appearance, appetite, or the weight and histologic appearance of vital organs. Leydig cells appeared to be unaffected as judged by normal libido and accessory sex gland weight (11). However, testicular weight, the number of sperm obtainable with an artificial vagina, and the gross appearance of spermatogenic tissue were severely affected in most individuals receiving the higher dosages of EGME. These changes were consistent with the profound quantitative changes in estimated sperm production reported herein. An increased incidence of oligospermia or azoospermia was associated with the higher doses (11). Interestingly, EGME effects on spermatogenesis would appear to be largely quantitative rather than qualitative in terms of functional integrity of the sperm produced. This conclusion is based on the observation that despite marked reductions in sperm production, fertility of treated males was not reduced below that of the controls when artificial insemination was performed with equal but low numbers of spermatozoa (11).
Quantitative histology: EGME effects on cell types By quantifying germ cells of different types, it is possible to assess their relative sensitivity and/or to identify the points in spermatogenesis that are disrupted by a given treatment. The testicular tissues we evaluated had presumably reached a steady state by 12 weeks, as the total duration of spermatogenesis in the rabbit is 52 days (28). In addition, spermatogonia had an opportunity to initiate six cycles of the seminiferous epithelium during exposure, as one cycle is 10.9 d. The quantitative histologic data from the present investigation provided evidence that the more developmentally advanced germ cells were affected more severely by EGME than their less developed precursors. In that regard, among rabbits in the 0.0 to 37.5 mg/kg dos-
W. E. BERNDTSONand R.H. FOOTE
age groups, the only statistically significant reductions in germ cell numbers involved old primary spermatocytes and round spe~atids at the 37.5 mg/kg dosage (Table l), for which cell numbers were reduced to 58% (from 2.43 to 1.42) and 27% (8.86 to 2.38) of control values, respectively. In the highest treatment group, spermatogenesis was so disrupted that histometric analysis was not possible for five males, Spe~atogonia were found to be relatively resistant, as there was no reduction in spermatogonia per Sertoli cell with increasing doses of EGME. However, the progressive (although not statistically significant) decrease in young primary spermatocyte numbers as dosages increased to 25.0 and 37.5 mglkg provides reason to suspect that EGME might be somewhat toxic to these cells. Also, the progression from the least to most advanced germ cell types resulted in only 1.91 young primary spermatocytes, 1.42 old primary spermatocytes, and 2.38 round spermatids, respectively. in the 37.5 m&g group (Table 1). These numbers represented 75,58, and 27% of control values, suggesting sensitivity of all three types of cells resulting in a cumulative effect. Additional replication would be needed to increase the power and sensitivity of the experiment for detecting an actual response (19,20). Regardless of whether numbers of young primary spermatocytes were or were not decreased by EGME, cellular attrition would have been necessary during the subsequent cycle of the seminiferous epithelium to reduce the number of old primary spermatocytes as noted (Table 1). In that regard, at the 37.5 mg/kg dosage, the ratio of old primary spermatocytes to their one-cycleyounger counterparts (1.42/1.91) averaged 0.74. This is indicative of a 26% loss via cellular attrition, because the old primary spermatocytes evolve from young primary spermatocytes without cell division and, in the absence of attrition, would be present in equal numbers. Some attrition is to be expected even among normal, untreated subjects, but the corresponding loss among control rabbits was only 4%. The two meiotic divisions of spermatogenesis involve the division of primary spermatocytes to yield secondary spermatocytes, which in turn, divide to yield round spermatids. Thus, the theoretical yield of spermatids per primary spermatocyte would be 4.0 in the absence of cellular attrition. For the control rabbits, 3.49 round spermatids were observed per young primary spermatocyte (i.e., 8.86 spermatids per 2.54 young primary spermatocytes, Table 1), which represents a 13% loss of cells from the theoretical maximum. These losses could be further attributed to a 4% attrition during the progression of young primary spermatocytes to old primary spermatocytes (young and old primary spermatocyte
numbers equalled 2.54 and 2.43, respectively, Table 1), and an additional 9% loss during the progression from old primary spermatocytes to round spermatids [i.e., 8.86 round spermatids were observed per 2.43 old primary spermatocytes (Table l), representing a yield of 3.64 vs. the theoretical yield of 4.01. By comparison, the corresponding overall attrition rate for rabbits in the 37.5 mg/ kg group was 69%, and consisted of a 26% loss during the progression from young to old primary spermatocytes and 58% loss during progression of the remaining old primary spermatocytes to round spermatids. The losses attributable to the two successive component phases of development are not simply additive to 69% because of their compounding nature. The overall attrition rate of 69% reflects the yield of 2.38 spermatids per 1.91 young primary spermatocytes, providing a ratio of 1.25, or 3 1% of the theoretical yield of 4.0. Contributing to this attrition was the decrease from 1.91 to 1.42 primary spermatocytes, which represented a 26% loss, while the ratio of 1.68 round spermatids per old primary spermatocytes (2.38/l .42) equalled 42% of the theoretical ratio of 4.0. Because only one stage of the cycle of the seminiferous epithelium was chosen for quantitative histometric evaluation, it is impossible to ascribe the latter losses more specifically to either or both of these meiotic divisions or to other intervening events in germ cell development. Nonetheless, it is clear that the 37.5 mg/kg dosage became progressively more toxic to germ cells during successive phases of development. Based upon the duration of the cycle of the seminiferous epithelium, the timing of the appearance of specific cell types during this cycle, and the duration of epididymal transit, it is possible to predict when adverse effects of an agent on specific testicular germ cells would first be manifested in ejaculated semen (2). Sperm output decreased in the 50 mg/kg group 5 weeks after initial exposure and in the 37.5 mg/kg groups 6 to 7 weeks after initial exposure (11). With a 10.9-d-cycle of the seminiferous epithelium, plus 9 d of transport through the epididymis, this would point to attrition of young primary spermatocytes as the initial cause of reduced sperm output.
Comparison with rats and mice Rats have received EGME orally over a wide range of doses for different lengths of time (8-10, 21-27). In general, doses greater than 50 mg/kg were required to produce any detectable effect upon rat testes. In rats, the pachytene spermatocytes were most sensitive, with the preleptotene, leptotene, and zygotene spermatocytes less sensitive. At higher doses of EGME the round spermatids and spermatogonia were reduced. Thus, with the
1 I, Number
higher doses of EGME there were direct effects on different types of spermatogenic cells and a cumulative depletion of more mature cell types as the progenitor cells were destroyed. Chapin et al. (25) reported differences among three strains of mice in testicular sensitivity to EGME. What was even more striking was the relative insensitivity of mice to this agent. When the three strains consumed 170 to 219 mg EGME/kg/d for 14 weeks there was no effect on the testis, whereas groups consuming 505 to 636 mg EGME/kg/d had testis weights reduced by 17 to 51%. These doses are 10 to 20 times the 225 mg EGME/kg/d that produced an effect on round spermatids in the current study and on testis weights of these animals (11). These collective data indicate that the relative sensitivity of different spermatogenic cell types is similar between rabbits and rats, but that the doses of EGME required to produce an effect are much greater in rats and mice than in rabbits. Much of the toxic effect of EGME on the testis of rats has been demonstrated to be due to the metabolite, methoxyacetic acid (24,26,29). No published reports were found on the metabolism of EGME in the rabbit, but apparently it is similar to the rat (Dr. E. Camey, personal communication, 1996). So, a difference in metabolism does not explain the difference in relative sensitivity to EGME between rabbits and rats.
Advantages and limitations of different assessment procedures The rabbit experiment was designed to provide each group with rabbits equal in average size, age, and reproductive potential, and with considerable, but equal range within groups. This was done to provide data with more general application. We also used two methods to evaluate testicular function in addition to the previously reported semen studies. These three measures, sperm output (11) histometric determination of the number of round spermatids per Sertoli cell, and numbers of homogenization resistant elongated spermatids were highly correlated and revealed treatment effects of generally similar magnitude. These results support the general validity of these procedures. However, advantages and limitations of each were also apparent. The histometric method permitted the quantification of spermatogenesis for most animals and the elucidation of the specific cell types or phases of development that were susceptible to disruption. Unfortunately, when there was serious disruption of spermatogenesis, this damage was readily apparent from gross histologic observation, but staging was precluded. The need to exclude data for these most severely affected individuals resulted in underestimation of the magnitude of the treatment effect. This limitation was
EGME and Spermatogenesis
partially offset by noting the need for such exclusions and their basis. In contrast, the homogenization method did not allow one to establish the relative sensitivity of specific germ cells or points in spermatogenesis that were affected by the toxicant. But, it permitted an evaluation of all testes, regardless of the dose of EGME, and the spermatogenic potential, based upon elongated spermatids, was estimated on all animals. The latter estimate, when correlated with sperm output measured by frequent semen collection, was r = 0.92 (11). Finally, the sperm output data (11) allowed an assessment of both sperm production and seminal quality, and rendered sperm available for use in artificial insemination. The latter revealed the important finding that effects of EGME on developing sperm were primarily of a quantitative rather than a qualitative nature. Not all methods can be applied in every experimental situation, but the collective results serve to emphasize the advantages to the use of two or more complementary approaches for examining effects of toxicants upon spermatogenesis. In that regard, it has been demonstrated that the power and sensitivity of experiments may be increased by using both testes of each animal for assessing treatment effects by a given method (18). But, as a practical matter and given the value of information gained by complementary procedures, the high coefficient of correlation between the weights of testes used for the homogenization and histometric methods (0.96, P < 0.001, Table 2) supports the general acceptability of alternately using the left and right testes of each subject for different assessment methods. From these studies, it is concluded that administration of EGME in the drinking water to sexually mature rabbits resulted in a depression of spermatogenesis in a dose-dependent manner. Additionally, these data established that EGME is increasingly more detrimental to the more advanced germ cells of the seminiferous epithelium. The present data on estimated sperm production are consistent with sperm output data from the same study reported previously. When compared to published data for mice and rats, the results support the likelihood that the testes of the rabbit are at least 10 times more sensitive to EGME than the testes of rats and mice. These results may provide valuable information when extrapolating effects from animal models to humans. Acknowledgmenfs -This study was partially supported by the Reproductive Effects Assessment Group, U.S. Environmental Protection Agency (CR820692-10). The authors are grateful to Dr. Eric D. Clegg for helpful suggestions in planning the experiment and to Dr. Edward W. Camey for thoughtful review of the manuscript, Technical assistance with the quantification of homogenization-resistant spermatids was provided by Antonio Castro and by Charlotte Bemdtson and Jason Bemdtson.
0 W. E. BERNDTSONand R. H. FOOTE
REFERENCES 1. Amann RP. A critical review of methods for evaluation of spermatogenesis from seminal characteristics. J Androl. 1981;2:37-58. 2. Foote RH, Bemdtson WE. The germinal cells. In. Scialli AR, Clegg ED, ed. Reversibility in testicular toxicity assessment. Boca Raton: CRC Press, 1992; l-55. 3. Freund M. Effect of frequency of emission on semen output and an estimate of daily sperm production in man. J Reprod Fe&l. 1963; 6:269-86. 4. Johnson L, Petty CS, Neaves WB. Influence of age on sperm production and testicular weights in men. J Reprod Fertil. 1984; IO:21 l-8. 5. Carlsen E, Giwercman A, Keiding N, Skakkebaek NE. Evidence of decreasing quality of semen during past 50 years. Br Med J. 1992; 305:609-13. 6. Clegg ED, Sakai CS, Voytek PE. Assessment Biol Reprod. 1986;34:5-16.
7. Environmental Protection Agency (EPA). Proposed guidelines assessing male reproductive risk. Fed Reg. 1988;53:24850-69.
8. Seed J, Chapin RE, Clegg ED, Dostal LA. Foote RH, Hurtt ME, Klinefelter GR. Makris SL, Perreault SD, Schrader S, Seyler D, Sprando R, Treinen KA, Rao Veermachaneni DN, Wise DL. Methods for assessing sperm motility, morphology, and counts in the rat, rabbit, and dog: A consensus report. Reprod Toxicol. 1996;lO: 23144. 9. Miller RR, Ayres JA, Young JT, McKenna MJ. Ethylene glycol monomethyl ether. I. Subchronic vapor inhalation study with rats and rabbits. Fundam Appl Toxicol. 1983;3:49-54. 10. Miller RR, Hermann EA, Young JT, Landry TD, Calhoun LL. Ethylene glycol monomethyl ether and propylene glycol monomethyl ether: Metabolism, disposition, and subchronic inhalation studies. Environ Health Perspect.
11. Foote RH, Farrell PB, Schlafer DH, McArdle MM, Trouem-Trend V, Simkin ME, Brockett CC, Giles JR, Li J. Ethylene glycol monomethyl ether effects on health and reproduction in male rabbits. Reprod Toxicol. 1995;9:527739. 12. Foote RH, Schermerhom EC, Simkin ME. Measurement of semen quality, fertility, and reproductive hormones to assess dibromochloropropane (DBCP) effects in live rabbits. Fundam Appl Toxicol. 1986;6:628-37. 13. Amann RP, Lambiase JT Jr. The male rabbit. III. Determination daily sperm production by means of testicular homogenates. Amin Sci. 1969;28:369-74.
14. Swierstra EE, Foote RH. Cytology and kinetics of spermatogenesis in the rabbit. J Reprod Fertil. 1963;5:309-22. 15. Bemdtson WE. Methods for quantifying mammalian esis: A review. J Anim Sci. 1977;44:818-33. 16. Abercrombie M. Estimation of nuclear population sections. Anat Rec. 1946;94:23948.
17. de Jong FH, Sharpe RM. The onset and establishment of spermatogenesis in rats in relation to gonadotrophin and testosterone levels. J Endocrinol. 1977;75:197-207. 18. Bemdtson WE. Sampling intensities and replication requirements for detection of treatment effects on testicular function in bulls and stallions. A statistical assessment. J Anim Sci. 1989;67:213-25. 19. Bemdtson WE, Neefus C, Foote RH, Amann RP. Optimal replication for histometric analyses of testicular function in rats or rabbits. Fun&am Appl Toxicol. 1989; 12:29 I-302. 20. Bemdtson
WE. A simple, rapid and reliable method for assessing
the number of replicates for animal experiments. J Anim Sci. 199 I ; 69:67-76. Chapin RE, Lamb JC IV. Effects of ethylene glycol monomethyl ether on various parameters of testicular function in the F344 rat. Environ Health Perspect. 1984;57:219-24. Chapin RE, Dutton SA, Ross MD, Sumrell BM, Lamb JC IV. The effects of ethylene glycol monomethyl ether on testicular histology in F344 rats. J Androl. 1984;5:369-80. Chapin RE, Dutton SA, Ross MD, Swaisgood RR, Lamb JC IV. The recovery of testis over 8 weeks after short-term dosing with ethylene glycol monomethyl ether: Histoiogy, celt-specific enzymes, and fete testis fluid protein. Fundam Appl Toxicol. 1985; 5515-25. Foster PMD. Creasy DM, Foster JR, Gray TJB. Testicular toxicity produced by ethylene glycol monomethyl and monoethyl ethers in the rat. Environ Health Perspect. 1984:57:207-17. Chapin RE, Morrissey RE, Gulati DK, Hope E. Barnes LH, Russell SA, Kennedy SR. Are mouse strains differentially susceptible to the reproductive toxicity of ethylene glycol monomethyl and
11,Number 1, I997
ethers in the rat. Toxic01 Appl Pharmacol.
26. Foster PMD, Greasy DM, Foster JR, Thomas LV, Cook MW, Gangolh SD. Testicular toxicity of ethylene giycoi monomethyl and monoethyl ethers in the rat. Toxicol Appl Pharmacoi. 1983; 69:385-99. 27. Holloway AJ, Moore HDM, Foster PMD. The use of rat in vitro fertilization to detect reductions in the fertility of spermatozoa from males exposed to ethylene glycol monomethyl ether. Reprod Toxicol. 1990;421-7. 28
Swierstra EE, Foote RH. Duration of spermatogenesis and spermatozoan transport in the rabbit based on cytological changes, DNA synthesis and labeling with tritiated thymidine. Am J Anat. 196S:ll6:401-12.
Bartlett JMS, Keer JB, Sharpe RM. The selective removal of pachytene spetmatocytes using methoxy acetic acid as an approach to the study i/z viva of paracrine interactions in the testis. J Androl. 1988;9:31&40.