The Effects of Feed Restriction during in Utero and Postnatal Development in Rats

E. W. Carney1, C. L. Zablotny, M. S. Marty, J. W. Crissman2, P. Anderson, M. Woolhiser and M. Holsapple3

Toxicology & Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan 48674

Received April 15, 2004; accepted June 24, 2004


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This study determined the effects of feed restriction (FR) during in utero and postnatal life on standard reproductive toxicity and developmental immunotoxicity end points. Groups of 26 time-mated CD rats were fed various amounts of Purina 5002 diet from gestation day 7 through lactation. Control rats were fed once per day in amounts based on historical control feed consumption data, while the amounts fed to the FR groups were reduced by 10% (10% FR), 30% (30% FR), or 50% (50% FR) relative to controls. Selected F1 weanlings were necropsied on postnatal day (PND) 22, assessed for immunotoxicity end points between PND 22 and 27 or PND 52 and 56, or maintained on FR through PND 70. Thereafter, half the remaining F1 rats in each group were fed ad lib (recovery subgroup), while the rest continued on FR. Both subgroups were necropsied at 21 weeks of age. In the 10% FR group, slight decreases in maternal body weight had no effect on F1 offspring body weights, but did decrease F1 liver weights. FR at the 30% level reduced maternal body weights by 10–20%, reduced F1 offspring body weights by as much as 21%, caused changes in numerous weanling organ weights, but did not affect reproductive or immune system function. Dams in the 50% FR group were 17–32% lighter than controls, resulting in F1 body weights that were 12–47% lower than controls. F1 estrous cycle length was increased, puberty was delayed by 6 days (males and females), and anogenital distance, epididymal sperm counts, and all organ weights were decreased in this group. Antibody responses were unaffected despite decreased spleen and thymus weights. Essentially all effects of feed restriction showed evidence of reversibility.

Key Words: reproductive toxicity; developmental toxicity; maternal toxicity; developmental immunotoxicity.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In recent years, regulatory guidelines for reproductive toxicity testing of agricultural and industrial chemicals (e.g., OPPTS 870.3800, OECD 416) have been extensively revised with the addition of many new end points of reproductive function and offspring development. The new end points include anogenital distance (a marker of altered sexual differentiation in utero), organ weights in weanlings, markers of puberty (age at preputial separation in males and age at vaginal opening in females), estrous cyclicity, oocyte quantitation, and sperm count, motility, and morphology. One reason these end points were added was to address concerns about the potential toxicity of hormonally-active chemicals. Although these end points are indeed sensitive to such chemicals, they can also be influenced by chemicals acting through nonendocrine mechanisms, as well as by extraneous physiological factors such as feed intake, body weight effects, or stress. It has also become increasingly common to incorporate additional parameters, such as immunotoxicity end points, to standard reproductive toxicity study designs in an attempt to comprehensively evaluate juvenile toxicity. Developmental immunotoxicity parameters may similarly be influenced by generalized effects associated with systemic toxicity.

Given that reproductive toxicity test guidelines require testing at a maximum tolerated dose level, one of the most prevalent and often vexing study confounders is reduced maternal and/or neonatal body weight. Considering the interrelated nature of neonatal growth and maturation, one would expect that the new maturational landmarks added to the test guidelines would be particularly sensitive to reduced rates of body weight gain. To examine the relationship between body weight and reproductive function, previous investigators have used feed restriction as a tool to manipulate body weight. While feed restriction in adult animals has been given a considerable amount of attention (Chapin et al., 1993Go; Cokelaere et al., 1998Go; Holehan and Merry, 1985aGo,bGo; Lederman and Rosso, 1980Go; Shaw, 1968Go), relatively little is known about the impact of feed restriction during in utero and postnatal development. Therefore, the effects of feed restriction and resultant maternal and offspring body weight decrement on reproductive and developmental immunotoxicity parameters were investigated in a one-generation design involving three different levels of feed restriction during gestation, lactation and development of the F1 offspring up to 21 weeks of age. Graded levels of feed restriction were utilized to develop what was essentially a dose-response for changes in body weight and resultant effects on reproductive and immunotoxicological parameters. The intent was to better understand the impact of reduced feed consumption and reduced maternal and neonatal growth on these parameters and, thus, improve the basis for interpreting data from conventional reproductive and/or juvenile toxicity studies.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Test animals. Sexually mature, time-mated CD (Sprague-Dawley derived; positive evidence of mating = gestation day 0) rats weighing 200–250 g each were purchased from Charles River Breeding Laboratories (Portage, MI) and acclimated to laboratory conditions for 5–6 days prior to feed restriction. The rats were housed singly in wire mesh, stainless steel cages in racks, except for shortly before and during lactation, when they were housed in plastic nesting boxes provided with ground corncob nesting material. All test animals were uniquely identified using subcutaneously implanted transponders (BioMedic Data Systems, Maywood, NJ). The animal room conditions were as follows: humidity (40–70%), temperature (22°C), photoperiod (12 h light:12 h dark), and air flow (12–15 changes/h). Certified Lab Diet # 5002 (PMI Feeds, St. Louis, MO) was ground and fed ad libitum during the acclimation period. Municipal tap water was available ad libitum for all animals throughout the study. All animal procedures were approved by the Institutional Animal Care and Use Committee. In addition, the study was conducted in an ALAAC accredited facility.

Experimental design. Groups of 26 time-mated rats were randomized, based on gestation day (GD) 0 body weights, into one of four groups fed different amounts of standard Purina 5002 diet. The control group was fed ground feed once per day, the amount determined from laboratory historical control feed consumption data for various stages of gestation, lactation, and postnatal development. The amount fed each day to the feed-restricted groups was reduced by 10% (10% FR), 30% (30% FR), or 50% (50% FR) relative to the control group. These graded levels of feed restriction were intended to mimic a variety of feed intake scenarios that we have encountered in past rat multigeneration reproductive toxicity studies.

Feed restriction regimens for the dams began on GD 7 and continued throughout gestation and lactation until weaning of the pups (Fig. 1). The dams were allowed to deliver naturally, with the day of delivery designated as postnatal day (PND) 0 for the F1 offspring and lactation day (LD) 0 for the dams. On PND 4, all litters were culled to 8 pups (4 males and 4 females whenever possible). Anogenital distance was determined on PND 4 for all males and females remaining in each litter after culling. On PND 22, one randomly selected F1 pup/sex/litter was subjected to a complete gross necropsy. A subgroup of approximately 20–23 pups/group were weaned on PND 21, put on the appropriate feed restriction regimen, and used for immunotoxicity evaluations between PND 22 and 27 or PND 52 and 56. Due to the small size (i.e., low body weight) of some FR pups, weaning of all remaining pups was delayed until PND 28 (typically done on PND 21). Beginning on PND 28, one randomly selected weanling/sex/litter was continued on the same feed restriction regimen until PND 70. To determine if any effects of feed restriction during early development were reversible, approximately one half of the F1 rats in each group were switched on PND 70 to ad libitum feeding (recovery subgroup), while the other half remained on feed restriction. Both the rats on continuous feed restriction and those in the recovery subgroup were terminated at 21 weeks of age and given a complete necropsy.



View larger version (19K):
[in this window]
[in a new window]
 
FIG. 1. Overview of experimental design and study schedule. Solid lines = feed restriction (FR); Dashed lines = ad lib feeding; Small circles = in utero/lactation development. GD = gestation day; LD = lactation day; PND = postnatal day.

 
General and reproductive toxicity end points. All animals were observed twice daily for alterations in behavior or demeanor. In addition, detailed clinical examinations were conducted weekly on each adult and weanling. Body weights of the dams were recorded on GD 0 (provided by supplier), 7, 14, and 21 and LD 1, 4, 7, 14, 21, and 28. The body weight of each pup was recorded on PND 1, 4, 7, 14, 21, and weekly thereafter. F1 offspring were examined for vaginal opening or preputial separation beginning on PND 28 or 35, respectively, continuing daily until these markers of puberty were attained. In addition, body weights were taken on the day of vaginal opening or preputial separation. Estrous cycle length and normality were evaluated via daily vaginal lavage (Cooper et al., 1993Go) on one F1 female/litter from 13–16 weeks of age.

Following weaning of their pups on PND 28, maternal animals were euthanized without any further collection of data. On PND 22, one randomly selected F1 pup/sex/litter was anesthetized, euthanized, and given a complete necropsy. The following organs were weighed: adrenals, brain, kidneys, liver, spleen, testes, thymus, and uterus. Males and females chosen for evaluation of developmental landmarks were euthanized in a similar manner at 21 weeks of age. Organs weighed in the adults were: adrenals, brain, epididymis, kidneys, liver, prostate, seminal vesicles, spleen, testes, thymus, ovaries, and uterus.

Immediately after euthanasia of the 21-week-old F1 males, a small sample of sperm from the right cauda epididymis was expressed into a dish containing Medium 199 (with bovine serum albumin) and was incubated at room temperature for approximately 2–3 min. An aliquot of the incubated sperm suspension was placed in a chamber of an HTM Integrated Visual Optical System semen analyzer (IVOS; Hamilton-Thorn Research, Beverly, MA) for the determination of total percent motile and percent progressively motile (showing net forward motion) sperm. A second aliquot of sperm suspension was transferred to a slide, and a smear was prepared for evaluation of sperm morphology. At least 200 sperm per male were evaluated and classified as normal or abnormal using criteria described by Filler (1993)Go. The left cauda epididymis was weighed and then frozen for subsequent determination of sperm count. After thawing, the left cauda epididymis was minced, diluted and stained with a fluorescent DNA-binding dye (HTM-IDENT, Hamilton-Thorn Research, Beverly, MA), and sperm count was determined using the IVOS sperm analyzer as described by Stradler et al. (1996)Go.

Immunotoxicity evaluations. On PND 22, 23, and 52, selected excess F1 pups were evaluated for immune function using the antibody plaque forming cell (AFC) assay (Ladics et al., 2000Go). On the aforementioned days, pups were immunized with a single intravenous injection in a lateral tail vein using approximately 2 x 108 sheep red blood cells (SRBC; Colorado Serum Company, Denver, CO). The rats were euthanized 4 days after immunization, and the spleens were removed and placed in HEPES buffered EBSS (Earle's Balanced Salt Solution). A single cell suspension was prepared by gentle disaggregation using frosted microscope slides. A dilution of the spleen cells was prepared and the number of cells was determined using a Coulter Z1® cell counter (Coulter, Miami, Florida). Additional cells were diluted and mixed with washed SRBC, guinea pig complement, and an agar solution. The contents were plated into the middle of a Petri dish, and a breath-humidified cover slip was placed onto the mixture. After several minutes the dish was placed into a dry incubator set at 37°C. Clear plaques that formed as a result of IgM and complement-mediated lysis of SRBC were counted 3 h later.

Statistical analysis. Body weights, body weight gains, anogenital distance, sperm count, percent total and progressively motile sperm, percent abnormal sperm, number of days per estrous cycle, number of estrous cycles per female, organ weights, and AFC data were evaluated by Bartlett's test for equality of variances ({alpha} = 0.01). Based on the outcome of Bartlett's test, a parametric or nonparametric analysis of variance (ANOVA) was performed. If the parametric or nonparametric ANOVA was significant, differences between controls and the various feed restriction groups were analyzed by Dunnett's test or the Wilcoxon Rank-Sum test with Bonferroni's correction, respectively. Litter size, gestation length, age at vaginal opening, and age at preputial separation were evaluated by nonparametric ANOVA and, if significant, were followed by the Wilcoxon Rank-Sum test with Bonferroni's correction. Pup survival indices were analyzed using the litter as the experimental unit by the Wilcoxon test as modified by Haseman and Hoel (1974)Go. Statistical outliers were identified using a sequential method ({alpha} = 0.02), but only values for feed consumption were routinely excluded unless justified by sound scientific reasons unrelated to treatment. The nominal alpha level was {alpha} = 0.05 unless noted otherwise.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
General Toxicity
There were no treatment-related clinical observations in any of the maternal rats. Relative to controls, maternal body weight was decreased by 1–5%, 10–20%, and 17–32% in the 10% FR, 30% FR, and 50% FR groups, respectively (Fig. 2A). Maternal body weight gains in the 30% and 50% FR groups were significantly decreased throughout gestation, whereas body weight gains in the 10% FR group were significantly decreased only on GD 7–14. Maternal body weight gains across the entire lactation period were significantly decreased in all FR groups relative to controls. The period most sensitive to feed restriction effects on maternal body weight appeared to be the first few days postpartum, as all feed-restricted dams lost weight, whereas the control dams gained weight during this period.



View larger version (25K):
[in this window]
[in a new window]
 
FIG. 2. Effect of various levels of feed restriction (FR) on body weights. (A) Maternal body weights during gestation and lactation. (B) F1 male body weights. (C) F1 female body weights. Dark lines indicate continuous feed restriction. Lighter lines indicate subgroup of animals switched to ad lib feeding on postnatal day 70. n = 25–26 animals/sex/group (split into 11–13 animals/sex/subgroup after postnatal day 70).

 
Of the 104 time-mated rats in the study, only one 30% FR group rat was not pregnant, and all pregnant rats produced viable litters (Table 1). There were no significant effects of feed restriction on offspring viability, litter size, or sex ratio. Although litter sizes on PND 7, 14, and 21 appeared to be slightly decreased in the 50% FR group, all values were well within historical control ranges. Mean gestation length was significantly increased in the 30% and 50% FR groups. However, all values for mean gestation length were within the laboratory historical control range.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Reproductive Performance of Feed-Restricted Dams

 
No treatment-related clinical observations or physical alterations were observed for the F1 pups in any of the feed-restriction groups during the lactation or postweaning periods. Statistically significant decreases in pup body weights occurred in the 50% FR (PND 1, 4, 7, 14, and 21) and 30% FR groups (PND 7, 14, and 21), but the 10% FR group pup body weights were very similar to those of the control pups (Table 2). Pup body weights in the 50% FR group were reduced by 12–13% at the time of birth, but by PND 21 these pups were almost 50% lighter than controls. In the 30% FR group, pup body weights remained almost identical to controls during the first few days of postnatal life, but began to fall below the control rate of growth starting around PND 7. A decrement of 21% relative to controls was observed by PND 21. Body weights of the 30% and 50% FR group male and female weanlings from PND 29–70 and those of the 30% and 50% FR subgroups maintained on continuous feed restriction were significantly decreased at all time points (Figs. 2B and 2C). The degree of body weight decrement during the postweaning period ranged from 6 to 19% and 30 to 43% for the 30% and 50% FR group animals, respectively. In the 30% and 50% FR F1 rats switched to ad libitum feeding from PND 70 (10 weeks) to 21 weeks of age, body weights showed evidence of recovery, but remained significantly lower than controls for all time points. Although the body weights of the 30% and 50% FR recovery subgroup animals did not achieve full recovery by termination at 21 weeks of age, their rate of body weight gain was comparable to or significantly higher than controls once ad libitum feeding began. Body weights and body weight gains of the 10% FR group F1 offspring were not significantly affected.


View this table:
[in this window]
[in a new window]
 
TABLE 2 Mean Pup Body Weights (g)

 
Sexual Differentiation and Development
Absolute anogenital distance (Table 3) was unaffected in the 10% and 30% FR groups, but was significantly decreased in the 50% FR group. However, anogenital distance:body weight ratios were essentially identical across treatment groups. Both age at vaginal opening and age at preputial separation were delayed by about 6 days in the 50% FR group. Animals in this group tended to acquire these markers of puberty at a lower body weight than in controls. No significant differences in age at vaginal opening or age at preputial separation were noted in the 30% or 10% FR groups. To explore the relationship between body weight and age at puberty onset, linear regression analysis was conducted post hoc. All feed-restricted groups showed very high individual animal correlations between body weight and age at preputial separation, whereas the correlation was much lower in controls. In contrast, the correlation between individual animal body weight and age at vaginal opening in female rats varied with level of feed restriction, showing only a moderate correlation in controls and mildly feed-restricted animals, but a very high correlation in the 50% FR group.


View this table:
[in this window]
[in a new window]
 
TABLE 3 F1 Offspring Developmental and Functional Parameters

 
Estrous Cyclicity
Monitoring of the estrous cycle of F1 females revealed a slight increase in the mean number of days per cycle and a significant decrease in the mean number of cycles completed in F1 females maintained on 50% FR throughout the study (Table 3). No effects on mean number of days per cycle or number of cycles per dam were noted for the 50% FR recovery subgroup, or any of the 10% FR or 30% FR group females.

Organ Weights
The absolute weights of all organs evaluated (see Materials and Methods) were significantly decreased in the 50% FR PND 22 weanlings compared to control weanlings (Table 4, only selected organ weights shown). In the 30% FR group, most absolute organ weights also were significantly decreased, with the exception of the ovary and adrenal weights of female weanlings and brain weight in males. When organ weights were expressed relative to terminal body weight, a clear increase in relative brain weight was noted in the 30% and 50% FR male and female weanlings. Conversely, relative weights of liver, spleen, and thymus tended to decrease in response to feed restriction. Relative liver weights of male weanlings appeared to be highly sensitive to feed restriction, with a significant decrease noted even at the 10% FR level.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Mean Absolute and Relative Organ Weights (g) — F1 Weanlings

 
At the 21-week F1 necropsy, the absolute weight of all organs was decreased in the 50% FR group rats on continuous feed restriction (Tables 5 and 6). Among the 30% FR rats on continuous feed restriction, there was a general tendency for most organs to be lighter than those of controls, but only the decreases in the absolute weights of the brain (males only), kidney (data not shown), liver, prostate, and spleen were statistically identified. Organs that exhibited an increase in weight relative to terminal body weight included the brain, adrenals (males only; data not shown), epididymides, and testes of 30% and 50% FR group rats, while relative liver weights were decreased in 30% and 50% FR group males and females maintained on continuous feed restriction. The only organ weight effect seen in the 10% FR group was a statistically significant decrease in female absolute liver weight.


View this table:
[in this window]
[in a new window]
 
TABLE 5 Mean Absolute and Relative Organ Weights (g) — F1 Adult Males

 

View this table:
[in this window]
[in a new window]
 
TABLE 6 Mean Absolute and Relative Organ Weights (g) — F1 Adult Females

 
Among the rats in recovery subgroups switched to ad libitum feeding on PND 70, absolute brain, epididymis, kidney (data not shown), liver, and testes weights remained significantly decreased in the 50% FR recovery subgroup rats, while only absolute liver weights (males only) remained decreased in the 30% FR recovery subgroup animals. Relative brain and epididymis weights were increased in males of the 30% and 50% FR recovery subgroup rats. A statistically significant increase in relative prostate weight among the 50% FR recovery subgroup rats was considered spurious, as a change in relative prostate weight was not seen for the 50% FR males which stayed on continuous feed restriction.

Sperm Analysis
The total number of sperm per cauda epididymis was significantly decreased in 50% FR rats on continuous feed restriction when compared to control rats (Fig. 3A). However, sperm counts were comparable to controls when expressed on per gram of tissue basis. No effects on epididymal or testis sperm counts were noted in the 50% FR recovery subgroup, nor in any of the other treatment groups (sperm counts were not determined in the 10% and 30% FR recovery subgroup animals). There were no effects of feed restriction on sperm motility or morphology (Fig. 3B).



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 3. Effect of various levels of feed restriction (FR) on sperm parameters in 21-week-old F1 males. (A) Sperm counts: total sperm per cauda epididymis; sperm per gram cauda epididymis. (B) Sperm motility (percent motile and percent progressively motile) and morphology. Recovery subgroup was switched to ad lib feeding on postnatal day 70. *Statistically different from control, {alpha} = 0.05. Sperm morphology evaluated only in the control and 50% FR groups. (n = 11–13 males/group).

 
Immunotoxicity
As mentioned above, F1 offspring in the 30% FR and 50% FR groups demonstrated reduced thymus and spleen weights at PND 22 (Table 4). Consistent with their reduced spleen weights, 30% FR (male) and 50% FR (male and female) pups evaluated on PND 22 or 23 also had reduced numbers of spleen cells (Table 7). These same pups were evaluated for their ability to generate an antibody response toward SRBC. Male F1 offspring from 50% FR dams exhibited a significant decrease in AFC/spleen. Although more variable, feed-restricted female pups also showed a similar reduction in AFC/spleen, but these differences were not statistically significant. Once normalized for spleen cell number (AFC/106 splenocytes), the ability of feed-restricted male offspring to generate a T-dependent antibody response was deemed to be normal, as AFC/106 splenocytes were similar to controls. However, female pups born to FR dams still demonstrated slight evidence of a reduced SRBC antibody response, as the number of AFC/106 spleen cells was consistently below that of controls and the 50% FR pups approached a 25% reduction.


View this table:
[in this window]
[in a new window]
 
TABLE 7 Summary of Total Spleen Cells and Antibody Forming Cells

 
Litter-matched pups (n = 10–12/sex) were provided normal, control diets for an additional 30 days (PND 52) prior to a second evaluation using the AFC assay on PND 52. FR effects on spleen cell number were largely recovered, but still demonstrated a slight (nonstatistically significant) reduction by F1 offspring born to 50% FR dams. Similar to PND 22/23, those offspring with a more prominent reduction in spleen cell numbers (i.e., pups from 50% FR) exhibited a decrease in AFC per spleen. The degree of suppression by male offspring (46%) and female offspring (22%) was considerably less as compared with PND 22/23 results (58 and 66%, respectively). As with weanling pup results, once normalized for cell counts the AFC response failed to demonstrate strong evidence of a treatment-related effect due to feed restriction. Female offspring immunized on PND 52 demonstrated a number of AFC/106 splenocytes that was similar to controls, while the male AFC/106 response was consistently below controls (29–34%), but did not show any relationship with the degree of FR.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Reproductive toxicity studies conducted according to regulatory guidelines require that the highest dose produce evidence of systemic toxicity, usually manifest as decreased body weight, and often associated with reduced feed consumption. Therefore, one of the most frequent challenges facing the reproductive toxicologist is to judge whether effects on various reproductive end points are secondary consequences of reduced feed intake and/or decreased body weight gain, or instead, represent direct reproductive effects of the toxicant (e.g., testicular toxicant). Such distinctions can be quite important for extrapolating animal data to estimate human risk, for the classification and labeling of chemicals, and to help understand mode of action.

To examine the relationship between body weight and reproductive function, previous investigators have used feed restriction as a tool to manipulate body weight. However, many of these studies have involved single levels of feed restriction (often quite severe) and/or have examined a limited set of end points (Cokelaere et al., 1998Go; Holehan and Merry, 1985aGo, bGo; Lederman and Rosso, 1980Go; Shaw, 1968Go). The design of this study parallels the work of Chapin et al. (1993)Go, who utilized graded levels of feed restriction in order to enhance relevance for typical reproductive toxicity studies. However, the present study differed from previous work in that it involved feed restriction during gestation, lactation, and adolescence, and it focused on an integrated set of new reproductive toxicity end points, including developmental markers, estrous cycling, sperm parameters, and weanling organ weights, as well as immune function parameters. Another departure from typical feed restriction studies was the fact that all groups, including controls, were fed measured amounts of feed once per day. Although it has been common practice in feed restriction studies to use ad libitum–fed controls, often fed once per week, such designs are potentially confounded by differences in feeding patterns, which are not controlled for. This point is not trivial, as daily meal feeding and ad libitum feeding had different effects on body weight gain, circadian patterns of corticoid secretion, and maternal behavioral patterns, even when daily feed consumption amounts were similar (Gallo et al., 1981; Rider and Chow, 1971Go). The amount of diet fed to controls in this study was based on extensive historical control data from multigeneration studies conducted in our laboratory. Finally, this study also included a recovery phase to determine the extent to which effects observed during development were reversible in adulthood.

A wide range of body weight effects was produced in the various feed restriction groups. In the 10% FR group, maternal body weights were decreased by 1–5%, which was considered to model a scenario of very mild general toxicity. The decreases in dam body weights reached the level of statistical significance during lactation, but not gestation, suggesting maximal maternal sensitivity to reduced feed intake associated with the demands of lactation. There were no statistically significant effects on F1 pup body weights in this group, indicating that the dams were able to maintain sufficient milk production despite the effects on their body weight. In the 30% FR group, maternal body weights were reduced by about 10–20%, resulting in F1 body weights that were 0–21% lower than controls. This degree of body weight decrement represents a typical scenario in which a significant, but not excessive level of general toxicity is reached. Under such conditions, the dam is unable to support normal growth of her pups. More extensive maternal and Fl body weight effects were noted in the 50% FR group, such that this group mimics a situation in which the targeted level of general toxicity specified in the testing guidelines was exceeded. While these situations are not ideal, dose selection is an imperfect science and such instances can and do occur. Regardless of the level of feed restriction, the pups generally appeared healthy, and litter sizes and pup survival rates were within the range of historical controls.

One unexpected result was a statistically significant increase in gestation length in dams of the 30% and 50% FR groups. None of the dams in this group delivered before GD 22, as opposed to the control and 10% FR group dams, which exhibited a nearly even distribution between GD 21 and 22 deliveries. However, the mean gestation length observed in the 30% and 50% FR groups was equal to the upper limit of the laboratory historical control range, making this finding of equivocal significance. Examination of earlier studies (Cokelaere et al., 1998Go; Lederman and Rosso, 1980Go; Shaw, 1968Go) involving feed restriction during gestation did not report this effect.

Anogenital distance, a marker of sexual differentiation, was decreased in both males and females of the 50% FR group, but was unaffected in the other feed restriction groups. However, ratios of anogenital distance to body weight were nearly identical across groups, indicating that the changes in absolute anogenital distance were merely due to a smaller size of the pups. This finding points to the critical importance of normalizing anogenital distance to a measure of body size in order to avoid "false positives" due to reduced pup size. In the present study, normalization to absolute body weight on PND 4 appeared to sufficiently correct for changes in body size, even when differences in pup body weight were large. Other authors have recommended normalizing to the cube root of body weight, as this ratio increases in a linear fashion during the early postnatal period (Gallavan et al., 1999Go).

The interpretation of delays in puberty onset in conjunction with decreased body weight is an extremely common problem for practitioners of reproductive toxicology studies. This is related to the fact that puberty onset is an apical end point which is not only influenced by sex steroid hormones, but also by body weight, body composition, and other general factors (Frisch, 1980Go; Frisch et al., 1975Go). An examination of the impact of body weight decrement on puberty onset indicated that it is greatly dependent on the magnitude of body weight decrement. For example, for rats in the 10% and 30% FR groups, in which body weights at puberty onset were reduced just slightly (≤3% in females; ≤11% in males), mean age of puberty onset was delayed by no more than 1.2 days. However, in the 50% FR group, in which body weight at puberty onset was decreased by 15% (females) or 23% (males), mean age of puberty onset was delayed by approximately 6 days (males and females). In a feed restriction study using the male pubertal onset assay, a short-term endocrine screen based on assessment of preputial separation, Marty et al. (2003)Go found that a feed restriction regimen that kept body weight at 5.4 to 9.9% below ad libitum–fed controls produced slight (0.6–1.8 day), nonsignificant delays in pubertal onset. However, Stoker et al. (2000)Go observed a significant 2.1-day delay in age at preputial separation in feed-restricted animals whose body weights were reduced by approximately 13%. The combined weight of evidence from these studies suggests that body weight reductions of approximately 10–15% at the time of puberty onset may represent an important transition point at which body weight can have a large impact on mean age of puberty onset (i.e., delays of several days).

The relationship between body weight and age at puberty onset in this study was further examined on an individual animal basis using linear regression analysis. In control group males and females, puberty onset was only slightly correlated with individual body weight (r2 = 0.55 and 0.48, respectively). However, these correlations were very high (r2 = 0.90–0.96) in males at all levels of feed restriction, while in females, the correlation increased steadily with the degree of feed restriction (Table 3). These increased individual animal correlations in the FR groups suggest an underlying influence of body weight, even when body weight decrements are minimal, and perhaps could explain the small (i.e., 1–2 day) delays which are so commonly encountered when evaluating puberty onset. Clearly, more needs to be learned about the complex relationship between body weight and puberty onset. In the meantime, it is recommended that sufficient body weight data be collected at and around the time of puberty onset, and that these data be carefully considered in a weight-of-evidence approach for interpreting delays in puberty onset.

Effects on estrous cyclicity were limited to a lengthening of the cycle only in those females maintained on continuous 50% FR. Interestingly, females which were on 50% FR until PND 70 showed no alteration of the estrous cycle when evaluated just three weeks later (PND 91–111), indicating a rapid recovery. Previous studies also have reported increased estrous cycle length in young adult female rats fed to maintain body weight at 30% less than controls (Chapin et al., 1993Go), while lesser degrees of feed restriction had no apparent effects on the estrous cycle.

Of all the end points evaluated in this study, weanling and adult organ weights were among the most sensitive to feed restriction (Tables 4Go6). Even very mild feed restriction (10% FR) brought about statistically significant decreases in adult female absolute liver weights and weanling male relative liver weights, while an extensive number of organs exhibited significantly lower absolute weights in the 30% and 50% FR groups. Relative weights exhibited responses that were very much organ specific, and in some cases, age dependent as well. Relative liver weights decreased in response to feed restriction in both adults and weanlings. In contrast, relative weights of brain and testes increased in both adults and weanlings, consistent with the well known "sparing" of these organs in the face of nutritional stress (Scharer, 1977Go). It is important to point out that this sparing phenomenon is more a difference in the slope of the weight change response than an absolute sparing of these organ weights (Figs. 4A and 4B). Thus, even so-called "spared organs" can still exhibit decreases in absolute weight in response to decreased body weight.



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 4. Percent change in absolute organ weight as a function of percent change in body weight (PND 22 data).

 
Age-dependent changes in organ weights were evident in the spleen and thymus, as relative weights of these organs were decreased in weanlings, but showed a relatively constant relationship with body weight in adults. Another response unique to weanlings was the lack of change in relative testes weights, as opposed to the increase that occurred in F1 adults (Tables 4 and 5; Figs. 4A nd 4B). Needless to say, organ weight changes that occur in conjunction with overall body weight effects warrant careful interpretation, giving consideration to age, degree of body weight decrement, and the particular organ in question. The issue is particularly complicated for reproductive organ weights in young animals, as their growth is hormonally regulated and these organs often exhibit nonlinear growth rates during specific windows of development (Ashby and Lefevre, 2000Go; Marty et al., 2003Go).

While total sperm per cauda epididymis was decreased in the 50% FR group, tissue efficiency as determined by total sperm per gram of tissue was unchanged. This observation, coupled with the fact that sperm motility and morphology were unaffected at all levels of feed restriction, indicated that these males were functionally normal. Thus, lower sperm counts in the 50% FR males were merely a reflection of their smaller size. While fertility was not evaluated in this study, others have found that adult males which are up to 30% lighter than ad lib–fed controls showed no compromise of fertility as evidenced by litters sired per male, litter size, or pup weight (Chapin et al., 1993Go).

With the heightened interest in the evaluation of juvenile toxicity, as well as increasing concerns about reducing animal usage, opportunities will be sought to incorporate additional toxicological end points into developmental and reproductive toxicology studies. Thus, any extraneous effects on these added end points due to nonspecific nutritional or general body weight effects also need to be properly understood. In this regard, lower antibody production potential (AFC/spleen) in feed-restricted rats appeared to be secondary to reduced spleen weights and cell numbers, both interpreted as effects due to compromised nutritional status of the dams (Table 7). Nonetheless, the ability of splenocytes to mount a T-dependent antibody response (as indicated by AFC/106 cells) was not compromised. AFC/106 spleen cells failed to demonstrate consistent reductions in SRBC-specific IgM or any treatment-related dose responses. This observation becomes important, as lymphoid organ weights from offspring are standard endpoints in reproductive and developmental toxicity studies that can be influenced by maternal nutritional status, independent of a xenobiotic's mechanism of action. Thus, lymphoid organ weight changes in the absence of functional immune testing may be inappropriate as indicators of immunotoxicity. On the other hand, this study illustrates the ability to integrate into a reproductive toxicology study a functional immunotoxicological evaluation using the AFC assay. While this integration presents some logistical challenges, they can be managed to produce a more complete picture of a xenobiotic's toxicity potential.

It also was of interest to determine whether or not feed-restriction-induced effects were reversible in F1 offspring switched to ad lib feeding on PND 70. In essentially all cases, the effects of early feed restriction had either recovered to the point that they were not statistically different from controls, or at least showed clear evidence that recovery was in progress (e.g., certain organ weights). This ability to recover from developmental insult demonstrates that not all changes incurred during developmental necessarily result in permanent effects. Furthermore, this pattern of reversibility may help distinguish agents that specifically target developmental processes from agents that affect certain testing end points through secondary, nonspecific mechanisms and, hence, may be of lesser concern.

In conclusion, graded feed restriction of CD rats during gestation, lactation, and up to 21 weeks of postnatal life in F1 offspring resulted in a variety of effects on several new reproductive and developmental end points (summarized in Table 8). Of all the end points examined, F1 liver weights were the most sensitive to decreases in feed intake/body weight, being affected by even very mild feed restriction. Certain other reproductive and developmental end points were affected to various degrees depending on the specific end point and degree of feed restriction. Furthermore, the reproductive and developmental changes induced by feed restriction during early development appeared to be reversible when assessed in animals returned to ad libitum feeding as adults. These data can be used as an aid in the interpretation of reproductive and/or juvenile toxicity tests confounded by decreased feed consumption and/or body weight.


View this table:
[in this window]
[in a new window]
 
TABLE 8 Summary of Responses in Relation to Degree of Body Weight Decrement

 

    ACKNOWLEDGMENTS
 
The authors gratefully acknowledge the following individuals for their important contributions to this study: K. Gallagher, B. Scortichini, A. Liberacki, K. Mikula, S. Eddy, and the TERC Toxicology Technologists and Animal Care staff. The authors thank Dr. J. Pitt for providing helpful comments on the manuscript.


    NOTES
 
Reprint Requests: Product Regulatory Management, 1803 Building, The Dow Chemical Company, Midland, MI, 48674.

2 Current address: Dow Corning Corp., Midland, MI. Back

3 Current address: International Life Sciences Institute, Washington, DC. Back

1 To whom correspondence should be addressed at 1803 Building, The Dow Chemical Company, Midland, MI 48674. Fax: (989) 638-9863. E-mail: ecarney{at}dow.com.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ashby, J., and Lefevre, P. A. (2000). The peripubertal male rat assay as an alternative to the Hershberger castrated male rat assay for the detection of anti-androgens, oestrogens and metabolic modulators. J. Appl. Toxicol. 20, 35–47.[ISI][Medline]

Chapin, R. E., Gulanti, D. K., Barnes, L. H., and Teague, J. L. (1993). The effects of feed restriction on reproductive function in Sprague-Dawley rats. Fundam. Appl. Toxicol. 20, 23–29.[CrossRef][ISI][Medline]

Cokelaere, M., Daenens, P., Decuypere, E., Flo, G., Kuhn, E., van Boven, M., and Vermaut, S. (1998). Reproductive performance of rats treated with defatted jojoba meal or simmondsin before or during gestation. Food Chem. Toxicol. 36, 13–19.[CrossRef][ISI][Medline]

Cooper, R. L., Goldman, J., and Vandenbergh, J. G. (1993). Monitoring of estrous cyclicity in the laboratory rodent by vaginal lavage. In Methods in Reproductive Toxicology (R. E. Chapin and J. Heindel, Eds.), pp. 45–56. Academic Press, Orlando, FL.

Filler, R. (1993). Methods for evaluation of rat epididymal sperm morphology. In: Male Reproductive Toxicology (R. E. Chapin and J. Heindel, Eds.), pp. 334–343. Academic Press, San Diego, CA.

Frisch, R. E. (1980). Pubertal adipose tissue: is it necessary for normal sexual maturation? Evidence from the rat and human female. Fed. Proc. 39, 2395–2400.[ISI][Medline]

Frisch, R. E., Hegsted, D. M., and Yoshinaga, K. (1975). Body weight and food intake at early estrus of rats on a high-fat diet. Proc. Natl. Acad. Sci. U.S.A. 72, 4172–4176.[Abstract]

Gallavan, R. H., Holson, J. F., Stump, D. G., Knapp, J. F., and Reynolds, V. L. (1999). Interpreting the toxicological significance of alterations in anogenital distance: Potential for confounding effects of progeny body weight. Reprod. Toxicol. 13, 383–390.[CrossRef][ISI][Medline]

Gallo, P. V., and Weinberg, J. (1981). Corticosterone rhythmicity in the rat: Interactive effects of dietary restriction and schedule of feeding. J. Nutr. 111, 208–218.[ISI][Medline]

Haseman, J. K., and Hoel, D. G. (1974). Tables of Gehan's generalized Wilcoxon test with fixed point censoring. J. Statis. Comput. Simul. 3, 117–135.

Holehan, A. M., and Merry, B. J. (1985a). Modification of the oestrous cycle hormonal profile by dietary restriction. Mech. Ageing Dev. 32, 63–76.[CrossRef][ISI][Medline]

Holehan, A. M., and Merry, B. J. (1985b). The control of puberty in the dietary restricted female rat. Mech. Ageing Dev. 32, 179–191.[CrossRef][ISI][Medline]

Ladics, G. S., Smith, C., Bunn, T. L., Dietert, R. R., Anderson, P. K., Wiescinski, C. M., and Holsapple, M. P. (2000). Characterization of an approach to developmental immunotoxicology assessment in the rat using SRBC as the antigen. Toxicol. Methods 10, 283–311.[CrossRef][ISI]

Lederman, S. A., and Rosso, P. (1980). Effects of food restriction on fetal and placental growth and maternal body composition. Growth 44, 77–88.[ISI][Medline]

Marty, M.S., Johnson, K.A., and Carney, E.W. (2003). Effect of feed restriction on Hershberger and pubertal male assay endpoints. Birth Defects Res. Part B Dev. Reprod. Toxicol. 68, 363–374.[CrossRef][ISI][Medline]

Rider, A. A., and Chow, B. F. (1971). Body weight changes in rats rationed to their ad libitum feed intake. Nutr. Rep. Int. 3, 21–29.

Scharer, K. (1977). The effect of chronic underfeeding on organ weights of rats. Toxicology 7, 45–56.[CrossRef][ISI][Medline]

Shaw, J. H. (1968). Mild caloric restriction during the reproductive cycle in rats: Its influence on the offspring. J. Dental Res. 47, 492–494.[ISI][Medline]

Stoker, T. E., Laws, S. C., Guidici, D. L., and Cooper, R. L. (2000). The effect of atrazine on puberty in male Wistar rats: An evaluation in the protocol for the assessment of pubertal development and thyroid function. Toxicol. Sci. 58, 50–59.[Abstract/Free Full Text]

Stradler, L. F., Linder, R. E., and Perreault, S. D. (1996). Comparison of rat epididymal sperm counts by IVOS HTM-IDENT and hemocytometer. Reprod. Toxicol. 10, 529–533.[CrossRef][ISI][Medline]