Abstract
Histopathologic examination of the immature ovary is a required end point on juvenile toxicity studies and female pubertal and thyroid function assays. To aid in this evaluation and interpretation of the immature ovary, the characteristic histologic features of rat ovary through the developmental periods are described. These histologic features are correlated with published changes in neuroendocrine profiles as the hypothalamic–pituitary–gonadal axis matures. During the neonatal stage (postnatal day [PND] 0–7), ovarian follicle development is independent of pituitary gonadotropins (luteinizing hormone [LH] or follicle-stimulating hormone [FSH]), and follicles remain preantral. Antral development of “atypical” follicles occurs in the early infantile period (PND 8–14) when the ovary becomes responsive to pituitary gonadotropins. In the late infantile period (PND 15–20), the zona pellucida appears, the hilus forms, and antral follicles mature by losing their “atypical” appearance. The juvenile stage (PND 21–32) is the stage when atresia of medullary follicles occurs corresponding to a nadir in FSH levels. In the peripubertal period (PND 33–37), atresia subsides as FSH levels rebound, and LH begins its bimodal surge pattern leading to ovulation. This report will provide pathologists with baseline morphologic and endocrinologic information to aid in identification and interpretation of xenobiotic effects in the ovary of the prepubertal rat.
Keywords
Introduction
Examination of immature rat tissues may be required in prenatal and postnatal development (segment III) and 1- and 2-generation developmental and reproductive toxicity (DART) studies, pubertal and thyroid function assays, and in juvenile toxicity studies. In DART studies, microscopic examination of tissues is typically limited to unscheduled deaths, since end points generally include only functional, gross morphology, and developmental landmarks. Evaluation of ovaries from immature rats is routinely required in pubertal assays since microscopic examination of ovaries at PND 42/43 is an end point (US EPA 2009). However, evaluation is limited to this peripubertal time frame, unless there are unscheduled deaths requiring examination of ovaries from the juvenile period (PND 21–32).
The study type in which pathologists will most often evaluate immature tissues is the juvenile toxicity study. Juvenile toxicity studies have become a higher priority with the changes in the regulatory environment. In 1998, the Food and Drug Administration (FDA) issued the Pediatric Rule requiring the pharmaceutical industry to assess safety of all new drugs likely to be of therapeutic value in children, and this rule immediately prompted the design of the juvenile toxicity study that is used to assess general organ toxicity and adverse effects on postnatal development. The FDA and the European Medicines Agency (EMA) have since issued guidance documents addressing when, where, and how to conduct juvenile toxicity studies. (EMA 2008; FDA 2006; Bailey and Marien 2011), and these studies are now part of a pediatric assessment that is required for each new drug application (NDA) in North America and for every marketing authorization application (MAA) in Europe, unless the sponsor is given a waiver for adult drugs with no therapeutic use in children (Barrow, Barbellion, and Stadler 2011; Tassinari et al. 2011). In these guidance documents, the reproductive system is one of the enumerated organ systems that is considered at high risk for drug toxicity since this system undergoes significant postnatal developmental changes (Tassinari et al. 2011). Based on a survey of outcomes of juvenile toxicity studies, there is encouragement for industry to conduct more targeted studies to identify juvenile toxicity (Bailey and Marien 2011). For more targeted studies to be appropriately designed and evaluated, the pathologist must understand the normal development of the ovary, be able to identify time-sensitive points in development, and differentiate delayed development from direct toxicity.
Conventional study designs provide age- and gender-matched control animals for comparison to xenobiotic-treated animals. Comparison of histologic features in the control versus xenobiotic-treated animals is commonly necessary to distinguish xenobiotic-associated histologic alterations from normal organ or tissue development. This direct comparison with control animals is not possible in the histopathologic evaluation of unscheduled deaths, which may occur at any time during the course of the study. Unless the study protocol makes some provision for collection of specimens from control animals in synchrony with unscheduled deaths, the pathologist must rely on personal experience and published information as histologic changes in xenobiotic-treated animals are detected and interpreted.
We have published the histology of the normal developing rat ovary from PND 20 through PND 50 to provide guidance for pathologists evaluating and interpreting findings in the female pubertal assay (Picut et al. 2014). With the increasing emphasis on juvenile toxicity studies, the present work was expanded to include the histologic features of rat ovaries from PND 3 until puberty in the first part of this article. The second part of this article is a summary of the pertinent neuroendocrine parameters during reproductive development of the female rat based on an extensive review of published literature. In this second part, temporal correlations are made between the neuroendocrine profiles and the morphologic appearance of the prepubertal ovary.
Part I: Morphologic Study of the Developing Rat Ovary
Materials and Methods
Animals
Female Sprague-Dawley (SD) rats (Charles River Laboratories, Inc., Crl:CD [SD] rats) ranging from PND 3 to 46 were used. The rats were randomly selected from the pool of clinically normal rats available in the WIL Research stock colony and did not undergo any treatment regimen. All study procedures and animal handling techniques were performed in compliance with WIL Research’s Institutional Animal Care and Use Committee (IACUC) and the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals (NIH 2002).
Necropsy and Histology
Each day, 3 female SD rats from separate litters, beginning at PND 3 and progressing through PND 46 for a total of 132 rats, were selected for scheduled necropsy. Prior to necropsy, final body weight and clinical observations for each animal were recorded. Animals were euthanized by decapitation without anesthesia on PND 3 through PND 12 or by carbon dioxide inhalation on PND 13 through PND 46 and subjected to gross necropsy examination. At the time of necropsy, fat was trimmed away from ovaries and then ovaries (paired) weights were measured and recorded for all animals which were PND 11 and older. The ovaries were collected and placed in 10% neutral buffered formalin. After approximately 24 hr, the ovaries were rinsed and stored in 70% ethanol until histologic processing.
The fixed tissues were trimmed, subjected to routine histologic processing, and paraffin-embedded. Ovaries were embedded in paraffin uniformly with the longest axis face down and any identifiable hilus oriented in a manner to achieve sagittal sections. Five sections of 5-μm thickness were collected from each ovary at 25-µm intervals for animals ≤ PND 19 or at 50-μm intervals for animals ≥ PND 20 beginning from a random starting point. Sections of ovary were prepared using a modification of the systematic uniform random sampling technique (Boyce et al. 2010), as follows. Ovarian tissue blocks were faced on a microtome until ovarian tissue was reached, a randomly generated number between 26 and 50 for animals ≤ PND 19 or 51 and 100 for animals ≥ PND 20 was identified, and that random number of sections were cut from the block and discarded prior to collection of the first section for histologic examination. After collection of the first section, 4 additional ovarian sections were obtained at the predetermined interval (either 25-µm or 50-μm intervals depending on the age of the animal). Routine hematoxylin and eosin (H&E) staining was performed on the obtained tissue sections.
Pathology
Detailed microscopic examination of the prepared ovarian sections was performed by a board-certified veterinary pathologist.
Developmental Time Periods
Five developmental time periods, indicated in PNDs, are recognized in the rat and were referred to when characterizing microscopic changes. These stages include the neonatal (PND 0–7), infantile (PND 8–20), juvenile (PND 21–32), peripubertal (PND 33–37), and pubertal (PND 38–46) periods (Ojeda, Advis, and Andrews 1980; Ojeda and Skinner 2006). These periods loosely correlate with similar developmental stages in humans (Barrow, Barbellion, and Stadler 2011; FDA 2000). An age comparison of developmental stages in rats and humans is summarized in Table 1. The infantile period was divided into early (PND 8–14) and late (PND 15–20) infantile periods for the purposes of this article, since this time period is fraught with dramatic morphologic and endocrinologic changes, many of which occur in the middle of the time period. The division therefore permits more effective correlation between morphologic and neuroendocrine dynamics.
Age comparison of developmental stages in rat and human.
Note. PND = postnatal day; ND = stage not defined.
cThe infantile period in the rat was divided into early (PND 8–14) and late (PND 15–20) infantile periods for purposes of this article to more effectively correlate important morphologic changes in the ovary (i.e., loss of atypical follicle morphology) with abrupt changes in neuroendocrine parameters (loss of inhibitory GABA; decrease in FSH) that occur at or about the middle of the infantile period.
These developmental stages depend on physiologic developmental changes and are not defined solely by reproductive changes, except for the peripubertal and pubertal periods. The peripubertal period is defined as the 3 to 5 days prior to first ovulation when luteinizing hormone (LH) exhibits differences in the morning and afternoon surges and when intrauterine fluid first appears. While PND 33 is typical for the onset of the peripubertal period, there can be considerable individual animal variation in onset and duration. Puberty is defined as the period when vaginal opening and first ovulation occurs. In the WIL Research historical control database, the mean age of vaginal opening for SD rats is PND 33.6, which is earlier than the published pubertal age of PND 38 to 46.
Definitions
The following definitions were used in the morphologic descriptions of follicles in the developing ovary (Pedersen and Peters 1968): Primordial follicle: primary oocyte surrounded by a single layer of flattened pregranulosa cells; types 1 and 2. Primary follicle: primary oocyte surrounded by a single layer of cuboidal granulosa cells; type 3b. Secondary follicle: primary oocyte surrounded by more than one layer of granulosa cells; types 4, 5a, and 5b. Tertiary follicle (also known as antral follicle): primary oocyte surrounded by more than one layer of granulosa cells and having antral formation (fluid filled space); types 6 and 7. Graafian follicle (also known as, and referred to hereafter as, ovulatory follicle): late stage (or large) antral follicle (approximately 0.9–1.0 mm in diameter) where the primary oocyte is surrounded by a cumulus oophorus, and the follicle will respond to surges in LH level; type 8.
Results
The salient features of the morphologic parameters during the neonatal, early and late infantile, juvenile, peripubertal, and pubertal periods are summarized in Table 2.
Features of prepubertal ovarian development.
Note. DHT = dihydrotestosterone; FSH = follicle-stimulating hormone; GABA = gamma aminobutyric acid; GH = growth hormone; HPG = hypothalamic–pituitary–gonadal; LH = luteinizing hormone; LHRH = luteinizing hormone releasing hormone; PND = postnatal day.
a Ojeda and Skinner (2006); Ojeda and Ramirez (1972); Raynaud, Mercier-Bodard, and Baulieu (1971); Hirshfield and Desanti (1995).
Neonatal Period: Birth to PND 7
At PND 3, the primordial, primary, and secondary follicles are apparent, with a thick peripheral cortex of primordial follicles and a central core of primary and secondary follicles (Figure 1). Mesenchymal stromal cells encapsulate these follicles into packets or remnants of ovigerous nests (Figure 2). At no other time in the development of the ovary is the primordial follicle the most prominent feature.
PND 6 (neonatal). Note the predominance of primordial follicles near periphery (arrow) while primary and secondary follicles form in core (arrowhead).
Infantile Period: PND 8 to PND 20
Early infantile period (PND 8–14)
During the early infantile period, the central core of the ovary expands with secondary follicles. Primordial follicles are limited to a thinning peripheral rim. By PND 10, early antral development occurs in the ovarian core. The secondary and early antral follicles at this stage are immature compared to follicles of the cycling adult. The immature secondary follicles are characterized by having plump granulosa cells, plump thecal cells, and little discernment between these 2 layers (Figures 3 and 4). Other features of the immature follicles are poor adhesion between granulosa cells and the primary oocyte and absence of a zona pellucida (Figures 4 and 5). Throughout the early infantile period, the secondary and early antral follicles in the medulla gradually “mature” to become similar to adult secondary follicles by PND 14. As part of this maturation process, the granulosa cells become tightly adhered to the oocyte, the granulosa cells become smaller with large basophilic nuclei, and the thecal cells elongate and emerge as a defined layer by PND 14.
Throughout the early infantile period, the interstitial stroma expands, allowing the secondary follicles to physically separate from one another. However, there is a notable absence of interstitial glands.
Late infantile period (PND 15–20)
From PND 15 to 20, during the late infantile period, follicular immaturity no longer exists. By PND 15, the zona pellucida appears, and there are subjectively similar quantities of secondary and early antral follicles (Figures 6 and 7). By PND 18, there is clear discernment of a hilus (Figure 8). By PND 21, large antral follicles can be seen on occasion within the medulla. These large antral follicles in the medulla of the infantile ovary can be distinguished from the ovulatory follicles of the peripubertal and mature ovary since ovulatory follicles that eventually ovulate are more commonly located in the outer ovarian cortex.
PND 15 (late infantile). Higher magnification of Figure 6. Note the formation of zona pellucida (arrow) and distinct separation between granulosa cells (G) and thecal cells (T).
Ovaries from early and late infantile rats occasionally have a histoarchitectural feature reminiscent of embryonic development. This feature is characterized by preantral follicles arranged in a radiating linear pattern perpendicular to the cortical surface, with the more developed follicles near the center (Figure 8). This pattern of follicle alignment is not apparent in an adult ovary.
Juvenile Period: PND 21 to PND 32
The most prominent feature of the juvenile period is apoptosis of granulosa cells and atresia of follicles. Apoptotic granulosa cells (Figure 9) begin as small clusters of pyknotic nuclei in secondary and early antral follicles predominantly in the medulla, with continual development of midsized antral follicles in the cortex. As the stage progresses, there is more widespread apoptosis of secondary and small antral follicles, and a gradual increase in interstitial glands. By PND 21, remnants of necrotic oocytes (i.e., the eosinophilic zona pellucida) deposit in the medulla. Follicular atresia is especially prominent at PND 26 to PND 28, leaving behind a subjectively maximal number of zona pellucida remnants in the medulla (Figure 10).
Peripubertal period: PND 33 to PND 37
The hallmark morphologic feature of the peripubertal period is the appearance of the ovulatory follicle (0.9–1.0 mm in diameter) in the outer cortex. These follicles are of sufficient size for ovulation and have a distended antrum and a primary oocyte lined by cumulus oophorus. In the peripubertal stage, the subjectively large number of atretic follicles, necrotic oocytes, and zona pellucida remnants characteristic of the juvenile period are no longer present, and the ovary matures with a well-developed hilus, interstitium with interstitial glands, and a balanced distribution of various sizes of secondary and tertiary follicles (Figure 11).
Pubertal period: PND 38 to PND 46
Puberty occurs with ovulation of ovulatory follicles, 0.9 to 1.0 mm in diameter. Corpora lutea appear after ovulation and “current” corpora lutea from the recent ovulation are the only type of corpora lutea initially present. Approximately 2 to 5 corpora lutea are present in the ovary during the first cycle (Figure 12). Thereafter, increasing numbers of corpora lutea appear, and current corpora lutea become “previous” corpora lutea in various stages of regression.
Part II: Correlation of Morphology to Neuroendocrine Parameters
The salient features of pertinent published neuroendocrine parameters and dynamics through the stages of development are depicted in Figure 13 and Table 2.
Prepubertal and pubertal endocrinology in the rat. Graphical presentation of endocrinologic changes occurring during normal development of the female rat from PND 0 through PND 47. Hormonal, physiological, and structural parameters of the HPG axis are listed along the left, and their changes are portrayed horizontally in a semiquantitative manner from the neonatal stage through puberty. The interaction of these changes within each stage are described in the text. Serum LH levels are high and increase to a maximum by the early infantile period; thereafter, LH level declines rapidly and starts diurnal pulsing by the mid-juvenile period. In the peripubertal period, afternoon pulses are greater than morning surges, and a superimposed mini-surge at puberty leads to ovulation. Serum FSH levels are high and peak in the early infantile period. Thereafter, FSH level declines to a minimum by PND 30, and then increases to adult levels by puberty. Centrally acting neurotransmitters (central neurotransm) play a significant role at around PND 25 when they promote LH diurnal pulsing. The E2 positive feedback loop becomes apparent by PND 25 and helps accomplish the LH mini-surge required for ovulation. The E2 negative feedback loop becomes apparent at PND 14 and contributes to the decline in the levels of FSH/LH. GABA acts centrally in the hypothalamus, first as an excitatory neurotransmitter promoting LHRH release during the first 2 weeks of postnatal life, and then switches to an inhibitory neurotransmitter. This switch coincides with reducing levels of FSH/LH during the juvenile period. At PND 30, there is loss (or decline) in the amount of inhibitory GABA, resulting in increased LH/FSH. Prolactin production at around PND 16 acts at the ovary and hypothalamus to promote the effects of LH during the peripubertal time. Apoptosis occurs in 2 waves: the first is that of oogonia during early infantile period and the second is that of secondary and midsized antral follicles in the medulla during the juvenile stage. Pituitary dependence of the ovary occurs at the infantile stage and is characterized by formation of antral follicles. Growth of antral follicles coincides with increased density of LH/FSH receptors until receptors reach maximum density at the peripubertal stage. Inhibin production by granulosa cells starts at PND 16, and estrogen production starts at PND 12, both of which are partly responsible for waning FSH/LH levels at this time. Aromatase activity by granulosa cells is detectable by the second week of postnatal development, and its increase to adult levels by puberty is responsible for the heightened production of E2 required for ovulation.
Neonatal Period (PND 0–PND 7)
The neonatal period is a time when the ovary, for the most part, is independent of the hypothalamus and pituitary, and development of primordial through secondary follicles is controlled by paracrine and autocrine growth factors produced locally by the oocyte, the granulosa cells, or stromal cells working in complex synchrony.
Factors required for primordial follicles to develop into primary follicles include members of the transforming growth factor beta (TGFβ) class of growth factors, and other cytokines and factors such as kit ligand, basic fibroblast growth factor–2 (FGF-2), bone morphogenetic protein 4 (BMP-4), leukemia inhibitory factor (LIF), and keratinocyte growth factor–7 (FGF-7; Skinner 2005), as well as blood insulin that activates recruitment of primordial follicles (Williams and Erickson 2012).
Factors that control the development of primary to secondary follicles include growth differentiation factor 9 (GDF-9; produced by the oocyte), and neurotransmitters, such as nerve growth factor (NGF), that arrive to the ovary by extrinsic innervation (Ojeda and Skinner 2006). While the ovary is considered pituitary independent during this neonatal stage, there is a growing body of evidence to suggest that the pituitary plays some role in the recruitment of primordial follicles and the transition from primary to secondary follicles. The granulosa cells of primary follicles begin to express LH and FSH receptors by the late neonatal period (PND 5–PND 7), and this low complement of receptors along with high serum FSH level is required for complete development of secondary follicles (Ojeda and Skinner 2006; White and Ojeda 1981). The developing granulosa cells of primary follicles regulate recruitment of additional primordial follicles by secreting antimullerian hormone that turns off recruitment (Williams and Erickson 2012), so an “internal” negative feedback loop within the ovarian parenchyma is operative.
During the neonatal period, steroidogenesis is stimulated by neurotransmitters vasoactive intestinal polypeptide (VIP) and catecholamines that reach the ovary by extrinsic innervation and work through peptidergic (VIP) and β2-adrenergic (catecholamine) receptors on theca and granulosa cells (Ojeda and Skinner 2006).
Despite the pituitary independence, there are relatively high levels of LH and FSH production by the pars distalis, and these gonadotropins increase dramatically during the neonatal period (reaching a maximum at PND 12 into the early infantile period). LH levels rise 50% and FSH levels rise 300% by PND 12 (Ojeda and Ramirez 1972). Both hormones are controlled by gonad-independent discharges of luteinizing hormone–releasing hormone (LHRH) at the level of the hypothalamus. The reason the LH and FSH levels are high and increasing is due to gamma aminobutyric acid (GABA) exerting an excitatory stimulatory signal that increases LHRH levels, and eventually LH and FSH production. Another reason for high levels of LH and FSH is the absence of an operative E2 negative feedback loop. Serum estrogen levels are undetectable for 2 reasons: (1) the aromatase activity of granulosa cells (which is required to convert androgens into estrogens) is insufficient, and (2) there is circulating alpha fetoprotein (AFP) that binds any small amount of E2 produced and prevents it from triggering the negative feedback at the hypothalamus (Raynaud, Mercier-Bodard, and Baulieu 1971). Insensitivity of the cycling structures of the ovary to high serum LH/FSH levels is in large part due to very low levels of LH/FSH receptors on thecal and granulosa cells, respectively.
During the neonatal period, while the ovary is pituitary-independent, the primordial, primary, and secondary follicles develop predominantly in the medulla. Medullary follicles are first to form and first to develop, and granulosa cells of these medullary follicles have a high proliferative labeling index (Hirshfield and Desanti 1995). The earliest adrenergic nerve fibers are located in the medulla of the neonatal ovary, and this extrinsic innervation is important in directing folliculogenesis in the neonate (Ojeda and Skinner 2006).
Early Infantile Period (PND 8–PND 14)
By PND 10, early antral development occurs, and antrum formation is the hallmark morphologic feature marking the switch of the ovary from pituitary independence (neonatal period) to pituitary-dependence (early infantile period). The secondary and early antral follicles at this stage are immature compared to follicles of the cycling adult and have been described as “atypical” (Hirshfield and Desanti 1995).
During the early infantile period, there is an increase in serum LH and FSH levels, reaching a maximum at PND 12 to PND 14 (Advis and Ramirez 1977; Ojeda and Skinner 2006). Serum LH and FSH levels increase in large part due to the continued excitatory effect of GABA on the hypothalamus. LH and FSH receptor levels on theca and granulosa cells are low but are increasing throughout this period (White and Ojeda 1981). The balance between maximum serum LH/FSH levels and low, but increasing, receptor levels regulates the rate of proliferation of secondary follicles into early antral follicles during the early infantile period (Hirshfield and Desanti 1995; White and Ojeda 1981).
Late Infantile Period (PND 15–20)
During the late infantile period, at about PND 15, GABA switches its signaling to inhibit rather than excite LHRH production, and there ensues a sudden drop in both serum LH and FSH levels. Another cause for the decrease in serum LH and FSH levels is the onset of the E2 negative feedback loop by PND 16 (Ojeda and Skinner 2006). Since circulating AFP levels disappear by PND 12, free E2 is available to exert this negative feedback effect. E2 production and the negative feedback are also enhanced because increasing numbers of thecal cells producing androgen are supplying increased substrate to the granulosa cells for E2 production. Also granulosa cells begin producing significant levels of inhibin, which has a negative effect on FSH secretion. Inhibin production is detectable at PND 5 and reaches a maximum by PND 17 (Rivier and Vale 1987).
Antral follicle development proceeds, despite waning FSH levels during the later infantile period. This apparent paradox is attributed to (1) increased density of LH/FSH receptors on granulosa cells with resulting heightened sensitivity of the follicle to the waning levels of LH/FSH (Fortune 1994), and (2) prolactin and growth hormone (GH) production by the pituitary that facilitates effects of LH/FSH on the ovary.
Juvenile Period (PND 21–PND 32)
During the juvenile time period, serum FSH levels are continually decreasing, reaching their minimal level at PND 30 (serum FSH level reaches 150 ng/ml and serum LH level declines to about 300 ng/ml; Advis and Ramirez 1977). FSH level remains low for the same reasons as in the late infantile period, namely inhibitory signaling of GABA, inhibin production (Rivier and Vale 1987), and the estrogen negative feedback effect. The minimum level of FSH reached by PND 30 correlates fairly well with the wave of atresia noted at PND 26 to PND 28. The cycling structures of the ovary continue to develop despite the nadir in serum FSH levels because FSH/LH receptor levels reach the maximum adult level by PND 30.
LH level wanes during the first half of the juvenile period, but unlike FSH level, starts being produced in a pulsatile and bimodal pattern with a morning and afternoon small surge. The maturation of the pulsatile and bimodal LH production pattern is due to the onset of the positive feedback of E2 on LH production at about PND 24 (Davis, Travlos, and McShane 2001). This E2 positive feedback effect is further buttressed by increasing production of E2, which continues to rise during the juvenile period as the aromatase activity of granulosa cells increases to a maximum by PND 30. By PND 26, LH secretion occurs in a bimodal pattern with the morning (a.m.) and afternoon (p.m.) levels equivalent at 40 ng/ml.
There is a striking difference between levels of gonadotropins in the rat and humans during the juvenile period. In the rat, LH/FSH levels significantly wane, but never reduce to zero. However, human gonadotropin levels approach zero during the juvenile period, and this is referred to as a “juvenile hiatus.” The lack of a legitimate juvenile hiatus of gonadotropin secretion in the rat should be kept in mind when evaluating the biological significance of gonadotropin-related morphologic changes in the rat compared to humans.
In the juvenile period, the continual development of cortical follicles into mid- and late-stage antral follicles, despite the nadir in serum FSH level and the wave of atresia of medullary follicles, seems contradictive. However, selected cortical follicles continue to develop due in part to the fact that FSH/LH receptors reach a maximum adult level by PND 30, resulting in heightened sensitivity of certain follicles to relatively low levels of serum FSH.
Peripubertal Period (PND 33–PND 37)
The peripubertal period lasts for 3 to 5 days and is defined as the time when the LH secretion occurs in a bimodal pattern with the afternoon surge greater than the morning surge. The LH level surges, which have been bimodal during the end of the juvenile period, are now characterized by a higher afternoon surge (approximately 80 ng/ml) when compared to a morning surge (40 ng/ml). This new surge pattern facilitates development of ovulatory follicles. LH level increases and develops this pattern due to a number of changes: (1) loss of inhibitory GABA; (2) onset of excitatory neurotransmitters and growth factors, such as IGF, FGF, EGF, TGFα, NPY, catecholamines, glutamate, neurokinin B, Kisspeptin, and PGE2, and their effect on LHRH release; (3) surges of prolactin that have a direct positive effect on the hypothalamus to increase LHRH production and eventual LH release; and (4) positive effect of estrogen on LHRH release.
Unlike LH levels, FSH levels are nonpulsatile and increase only after PND 35 (Advis and Ramirez 1977; Ojeda and Ramirez 1972). This blunted increase in FSH level likely occurs because estrogen and inhibin continue to be produced by the granulosa cells and exert a negative effect on FSH production.
Pubertal Period (PND 38–PND 46)
Puberty occurs with ovulation of ovulatory follicles, 0.9 to 1.0 mm in diameter, and this occurs when an LH level mini-surge occurs on top of the normal afternoon surge (Fortune 1994). The LH level mini-surge occurs around 2 to 4 p.m. and adds to the normal afternoon LH surge of 80 ng/ml. This additional mini-surge depends on maturation of the hypothalamic–pituitary–gonadal (HPG) axis, positive effects of neurotransmitters, and sufficient levels of serum estradiol sustained for a sufficient time. So even though the LH pulse pattern during the peripubertal time frame is independent of the ovary, the mini-surge of LH level required for ovulation is induced directly by sustained high levels of estrogen from the ovary. E2 acts in a positive feedback manner both at the hypothalamus to evoke LHRH release and at the anterior pituitary sensitizing LH secreting cells to the stimulatory effect of LHRH (Davis, Travlos, and McShane 2001). The LH level mini-surge causes contraction of the muscular theca externa cells, which have LH receptors. This results in a scalloped appearance to the ovulatory follicle, though this change can be nonspecific. During this late stage of follicle development, the primary oocytes (after obtaining 4N chromosomes) enters its first meiotic division and extrudes a polar body. Therefore at ovulation, a diploid secondary oocyte is extruded into the oviduct.
Discussion
In juvenile toxicity studies, examination of immature ovaries from rats is generally required depending on the study design. Histology of the rat ovary from birth to PND 3 has been reported by Rajah, Glaser, and Hirshfield (1992), limited histologic features of prepubertal ovaries have been published by Hirshfield and Desanti (1995) and Davis, Travlos, and McShane (2001), and histologic features of the rat ovary from PND 20 to PND 50 have been illustrated and described by Picut et al. (2014). In this study, we provide complete descriptions and illustrative examples of histologic features of the developing rat ovary from PND 3 through puberty in a manner useful to the toxicologic pathologist assigned the task of reading a juvenile toxicity study, and we further provide the pertinent published neuroendocrine dynamics during development to allow correlations between the morphologic changes and the neuroendocrine parameters.
At birth, the ovary consists of densely packed diploid oogonia clustered into packets or cords radiating from the hilus (Byskov 1974). These cords are called ovigerous nests and are separated by elongated mesenchymal cells (Schindler, Nilsson, and Skinner 2010). Early in the neonatal period, from PND 0 to PND 4, there is a marked apoptosis and loss of the majority of oogonia, while a minority of the oogonia (i.e., those that survive) enter the prophase stage of the first meiotic division (Fortune 1994), assemble into primordial follicles, and are called primary oocytes. Despite the wave of apoptosis from birth to PND 5, apoptotic oogonia are rarely visible in routine H&E sections (Peters 1969) and were not apparent in histologic sections from PND 3 to PND 5 in our study. Apoptosis of oogonia is dependent on TNFα and a germ cell–specific transcription factor FIGα and occurs in the early neonatal period when maternal progesterone and estrogen wane and no longer protect oogonia from demise (Skinner 2005; Soyal, Amleh, and Dean 2000).
Taking into consideration the dynamic morphologic and endocrinologic changes that are normally occurring during development, critical and sensitive time points for heightened susceptibility to test substances or other insults can be surmised. A sensitive time point in ovarian development is PND 12 to PND 14 when transitioning from early infantile to late infantile periods, and GABA switches from being an excitatory neurotransmitter to having an inhibitory effect on LHRH resulting in the physiologic waning of FSH/LH levels. Any disruption of this signaling change could conceivably result in precocious development of follicles leading to cystic unovulated follicles. Another sensitive time point could be PND 14 to PND 18 (late infantile period) when prolactin and growth hormone increase the sensitivity of the ovary to the waning levels of LH/FSH, allowing for follicular development in the cortex to continue. A xenobiotic that reduces prolactin levels (e.g., dopamine agonists) might result in delayed maturation of antral follicles and delayed puberty, while other xenobiotics that increase prolactin levels could lead to precocious puberty with increased ovarian weights and early onset of ovulation (Advis and Ramirez 1977).
The dynamic changes occurring at the end of the juvenile period (PND 30) provides yet another sensitive time point. At PND 30, there is loss of inhibitory GABA and an onset of prolactin surging that allows for FSH/LH levels to increase and promote late follicle stage development. If a xenobiotic results in diminished prolactin surging, delayed final development of the follicles would ensue, and there would be a predominance of large midsized follicles that likely undergo atresia. Puberty and ovulation depend on the concerted and complex intricacies of the HPG axis, and anything that interferes with this LH level mini-surge would delay ovulation and puberty, resulting in cystic follicles, increased atretic follicles, retained antral follicles, and decreased corpora lutea. The most common histopathologic findings with ovarian toxicity in 28-day studies include large-sized atretic follicles and decreased corpora lutea indicating a disturbance of ovulation and large follicle development (Sanbuissho et al. 2009).
Having an understanding of the relationship between basic histologic features of the developing ovary and changing neuroendocrine parameters will enable the pathologist to support appropriate design of targeted juvenile toxicity studies, identify endocrine disruptive events, and determine critical and sensitive time points for susceptibility to xenobiotics when performing histopathologic examination of ovaries as an end point in pubertal assays, juvenile toxicity studies, and other reproductive and developmental toxicity studies.
Footnotes
Acknowledgments
The authors wish to thank Jim Berkley, BS, RLATG, at WIL Research in Ashland, Ohio, for his contributions as the in-life project leader for this study; and Sirena L. Hudgins, HT(ASCP)CM, QIHC, Pamela S. Taylor, and Cynthia L. Swanson, MS, at WIL Research in Hillsborough, NC, for their technical expertise in the coordination of Histology efforts, preparation of tissues and slides, and preparation of the digital histopathology images used in this manuscript, respectively.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was fully funded by WIL Research and the NTP and National Institute of Environmental Health Sciences.