DEVELOPMENTAL DYNAMICS 194177-192 (1992)

The Changing Architecture of the Neonatal Rat Ovary During Histogenesis ROOPA RAJAH, EDMUND M. GLASER, AND ANNE N. HIRSHFIELD Departments of Anatomy (R.R., A.N.H.) and Physiology (E.M.G.), School of Medicine, University of Maryland at Baltimore, Baltimore, Maryland 21201

The purpose of this study was to ABSTRACT describe the changing histological organization of the rat ovary during postpartum days one through three (pl-p3). A PC-based image-combining microscope system was used to reconstruct the ovary in three dimensions. On pl, cylindrical pocket-like structures radiated from the core of the ovary that were open toward the surface epithelium. The walls of the pockets contained connective tissue cells and capillaries (stroma).By p2, these pockets had completely closed; each pocket enclosed a small nest of oocytes and a few presumptive granulosa cells. By p3, the pocket-like organization had disappeared. On pl, only one or two primordial follicle-like structures were observed in the core and toward the periphery of the ovary; most oocytes were not enclosed in follicles. By p3, very few naked oocytes remained; primordial follicles predominated in all the regions of the ovary and some of the follicles had multiple layers of granulosa cells. There were changes in location, area, and volume of the rete tubules during these postnatal days. The extraovarian rete was visible on all 3 days but changed its orientation relative to the ovary. The connecting rete was found beneath the epithelial layer of the ovary on all 3 days and showed dramatic increase in area on p2. The wide lumen of the intraovarian rete was in direct contact with some of the oocytes near by on all 3 days, but these “communication points” were most abundant on p2. Based on our observations of different cell-cell associations during this time period, we hypothesize (1)that the mesenchymalpresumptive granulosa cell association is essential for the completion of folliculogenesis, and (2) the rete ovarii may have an inductive role in follicle assembly. These observations suggest that the first 3 days postpartum are critically important for studying the heterogeneous cell interactions that lead to the assembly of primordial follicles. The regional differences in tissue organization during this formative period may have significant implications on later aspects of follicular development. o 1992 Wiley-Liss, Inc.

INTRODUCTION Very little is known about the histogenesis of the rat ovary. The histological organization of the embryonic (Arai, 1920; Torrey, 1945; Beaumont, 1962; Franchi and Mandl, 1963; Merchant, 1975; Merchant-Larios, 1979;Stein and Anderson, 1979;Prepin, 1984;Satoh, 1985;Paranko, 1987),immature (Jackson, 1913;Bradbury, 1940;Dawson and McCabe, 1951;Rennels, 1951; Bjorkman, 1962; Lawrence et al., 1979; Amsterdam and Rotmensch, 1987;Anderson et al., 1987;Carnegie et al., 1988; Ojeda and Urbanski, 1988;Ueno et al., 1989), and adult (Bjorkman, 1962; Lawrence et al., 1979; Wordinger et al., 1990) rat ovaries have been investigated in detail. However, the immediate postpartum days of development have been neglected. The purpose of this study was to investigate the changing architecture and different cellular associations of the ovarian tissues during postpartum days one through three (pl-p3). Particular attention was devoted to the rete, since it has been hypothesized that the rete is one of the sources for the follicular cells (Byskov and Moore, 1973;Byskov, 1975;Byskov et al., 1977;Costa Gudes and Miraglia, 1977;Stein and Anderson, 1979; Zamboni, 1979).

Key words: Rete, Oocyte, Presumptive granulosa Mesenchymal Stroma, mordial follicle

Received May 26, 1992; accepted July 17, 1992. Address reprint requestsicorrespondence to Roopa Rajah, Department of Anatomy, School of Medicine, University of Maryland a t Baltimore, Baltimore, MD 21201.

0 1992 WILEY-LISS, INC.

RESULTS Gross Morphology of Ovary and Rete Tubules On p l , the ovary hung from the caudal end of the kidney by a short connective tissue ligament that expanded into a triangular pouch enclosing the extra ovarian rete (ER) (Fig. l).The ER was located on the cranial end of the ovary. The pouch that enclosed the ER also extended caudally and enveloped the ovary. The ovary was a C-shaped structure similar in shape to a cashew nut; it had a thick cranial tip, a slightly thinner caudal tip, and its flanks (sides) faced the dorsolateral and ventrolateral surfaces of the body. The volume of the body of the C was much greater than the two tips. The ER was a bundle of convoluted tubules that were located at the cranial end of the ovary, partly on the

178

RAJAH ET AL

Figs. 1 and 2.

HISTOGENESIS OF THE NEONATAL RAT OVARY

179

dorsal flank and partly on the ventral flank. From this point, upon its entry into the ovary, it continued as the connecting rete (CR), which was restricted only to the core of the ovary (Fig. 2). On all 3 days, the CR was located just beneath the surface epithelium connecting the ER with the intra ovarian rete (IR). From p l through p2, it showed a gradual increase in area and specific pattern of organization beneath the surface epithelium. During these 48 hr, as the CR extended within the ovary more toward the caudal and less toward the cranial tip, its terminal ends graded into the IR. The CR and IR could be differentiated one from the other by their histological differences. Though the IR was found only in the core of the ovary on p l , by p2 it extended to the caudal and cranial tips of the ovary along with the CR.

Regional Differences Within the Ovary On p l and p2, there were significant regional differences in the organization of cell types within the ovary. This neonatal rat ovary showed distinct “medullary” and “cortical” regions. On p l , the two tips of the ovary were similar in their histology to that of the cortex; all these three regions differed from the medulla. The three major cell types (epithelial, stromal and germinal) were more uniformly distributed by p2, although the regional differences were still visible. By p3, the ovary showed a uniform distribution of primordial follicles in most of the regions of the ovary. Epithelial cells. There were two major types of epithelial cells found in the neonatal rat ovary: the surface epithelium and the epithelioid cells th at were presumably the future granulosa cells. On p l , these “presumptive granulosa cells” were found enveloping and invading the clusters of oocytes. The surface epithelium consisted of multiple cell layers on the surface of the flanks and the two tips of the ovary with discontinuous BM (Fig. 3). However on the back of the C, the surface epithelium was pseudostratified (Fig. 4). On p2, the surface epithelium was uniformly pseudostratified with a continuous BM beneath. By p3, the surface epithelium was organized into a single layered structure with a continuous BM in all the regions of the ovary. Stromal cells. The stromal cells were organized into

~

~

~~

~

~

~~

Fig. 1. Diagrammatic representation of the position of the left ovary relative to the left kidney on p l . (A) Ventral view. (B) Dorsal view. Od, oviduct; Ov, ovary; ER, extra ovarian rete. Fig. 2. Diagrammatic representation of a rat ovary on p l , showing the organization of the rete tubules and the ovarian tissues. The dash line indicates the back of the C-shaped ovary with the superior dorsal flank (D) and the inferior ventral flank (V).ER is found on both the dorsal and the ventral flanks at the cranial tip (R) and also extends on the dorsal flank toward the core of the ovary. In the midway, it enters the surface epithelium to become connecting rete (CR). The cranial tip and the middie region of the ovary are cut open to show the organization of the pockets. (L) ligament; (P) pocket; (CU) caudal end; (S) collagen bundle forming the stalk of the ovary.

Fig. 3. p l ovary: oocyte clusters at the cranial and caudal tips. The two tips are separated by a notch (arrowhead). Elongated funnel like pockets (double arrows) extend from the core (C) toward the periphery (P). Surface epithelium (arrow) is multilayered and not separated from the clusters. Original mag. x 800.

cylindrical pocket-like structures that radiated from the core of the medullary ovary toward the periphery on p l (Fig. 6). The walls of these pocket-like structures were primarily composed of connective tissue cells (mesenchyme) and endothelial cells. These stromal cells that formed the walls of the pockets were not continuous with the patches of stromal cells beneath the surface epithelium (subepithelial cells). Therefore, the pockets remained open toward the periphery (Fig. 5). The subepithelial tissue was discontinuous on the surface of the flanks and tips of the ovary, but on the back of the C they formed a continuous layer with a distinct BM separating the surface epithelium from them. There were no stromal cells among the clusters of oocytes and presumptive granulosa cells. On p2, the opened regions of the pockets, toward the surface, were completely closed (Fig. 7A) and were continuous with the subepithelial tissue that was now a continuous layer with the distinct and continuous BM separating them from the surface epithelium. By p2, the stromal cells from the walls of the pockets appeared to be invading the clusters of oocytes/presumptive granulosa cells. Wherever the stromal cells opposed the presump-

180

RAJAH ET AL.

Fig. 4. p l ovary: oocyte clusters in the middle region of the ovary. The clusters of oocytes are separated by the wall of the pockets (arrows). Rete tubules (R) are entering the ovary at the core (double arrows). At this point the rete tubules are in continuity with the base of the pockets (large arrow-

head). The degenerating oocytes (small arrowheads) are restricted to the medullary region and the surface epithelium (large arrow) is pseudostratified. Original mag. x 4,000.

tive granulosa cells, there was either a complete or forming BM. By p3, the pocket-like organization of the stromal tissue was replaced by a honeycomb-like organization, each circle representing follicular BM around individual primordial follicles. Presumptive granulosa cells. Compared with the stromal cells, the presumptive granulosa cells differed in their cellular structure and their affinity toward toluidine blue stain. The presumptive granulosa cells were large and irregular in shape with fine cytoplasmic strands that intervened between the oocytes; their nuclei stained light with toluidine blue. In contrast, the stromal cells were spindle shaped with darkly stained ellipsoidal nuclei. On p l , the presumptive granulosa cells were mostly found close to the walls of the stromal pockets with their cytoplasmic strands extending between the oocytes on the periphery of the clusters and the stromal cells in the pockets. A few of the presump-

tive granulosa cells were found in the center of the clusters physically separating some of the oocytes. This condition was more prominent in the pockets that were located in the core of the medulla than those in the

Fig. 5. p l ovary: oocyte clusters close to the periphery. Oocytes are directly in contact with the surface epithelium (arrow). There are no stroma1 cells separating the surface epithelium from the oocytes. The two big clusters of oocytes are partially separated by a few stromal cells (arrowheads) forming the wall of the pocket. Original mag. x 7,875. Fig. 6. p l ovary: oocyte clusters in the core of the ovary. Stromal tissue (S) extends from the core toward the periphery enclosing the oocytes (arrows) along with the presumptive granulosa cells (arrowheads) into cylindrical pockets (double arrows). The wall of the pocket (W) has spindle shaped mesenchymal cells with elongated nuclei and blood capillaries (C). Four of the degenerating oocytes (*) are in contact with the stroma. Original mag. x 500.

HISTOGENESIS OF THE NEONATAL RAT OVARY

Figs. 5 and 6.

181

182

RAJAH ET AL.

Fig. 7.

HISTOGENESIS OF THE NEONATAL RAT OVARY

cortex or in the tips of the ovary. By p2, most of the oocytes within the clusters in the core of the ovary were physically separated from each other by the “intervening” presumptive granulosa cells (Fig. 7B). The distribution of the presumptive granulosa cells within each pocket in the core of the ovary was more uniform than that found on p l . However, there were tight clusters of oocytes found near the periphery even on p2 (Fig. 7A). By p3, most of the presumptive granulosa cells had organized into follicles. The remaining naked oocyte clusters were restricted to the periphery and the two tips of the ovary. Oocytes. The oocytes were in interconnected large clusters on p l . By p2, the clusters were much smaller and all the oocytes in each cluster were bound by a BM that enclosed the entire cluster, along with their presumptive granulosa cells. There were a few true primordial follicles in the core of the ovary with distinct follicular BM (Fig. 8). By p3, the mesenchymal cells had organized themselves around each primordial follicle so that the pocket like mesenchymal organization was completely lost. Even at this early age, there were some follicles in the core of the ovary that had more than one layer of granulosa cells. The cellular association of the oocytes showed a gradual change from p l through p3. On p l , most of the oocytes were in contact with each other while some of the oocytes on the periphery of the clusters were in contact with the presumptive granulosa cells. Some of the oocytes in the middle of the clusters were in contact with the cytoplasmic strands of the presumptive granulosa cells that intervened between the oocytes. There were some oocytes that were in contact with the mesenchymal cells; all of these oocytes showed signs of degeneration (Fig. 6 ) . There was an increasing gradient in the distribution of the degenerating oocytes from the periphery to the core of the ovary (Fig. 4). On p2, most of the oocytes were in direct contact with the presumptive granulosa cells and there were fewer naked oocytes that were surrounded only by other oocytes. By p3, most oocytes had formed oocyte-granulosa cell associations; granulosa-mesenchymal cell associations with a distinct follicular BM predominated.

183

formed follicular BM which appears as a dark blue line in aniline blue stained sections and as a clear region in toluidine stained tissues. Using both toluidine and aniline blue stains, we were able to visualize BM on all 3 days. There were oocytes surrounded by presumptive granulosa cells that were not opposing any stromal cells; these presumptive granulosa cells did not have any BM around them. On p l , some of these “folliclelike structures” were observed toward the surface epithelium and toward the core of the ovary (Fig. 10 A,B). However, these structures had neither the squamous appearance of the first assembled granulosa cells nor a layer of stromal cells opposing the presumptive granulosa cells, and they lacked a follicular BM. On p2, numerous follicle-like structures were observed in different regions of the ovary, but true primordial follicles with complete encircling BMs were observed only near the core of the ovary (Fig. 8). By p3, the follicles predominated in the medullary and cortical regions (Fig. 9>, except toward the extreme periphery where some clusters of oocytes were found.

Collagen in Stroma Sections were stained with aniline blue to observe the collagen distribution in the ovary. This stained the interstitial collagens (type I, 11, 111) as well as the BM collagen (type IV), and thus showed the organization of collagen bundles in the stroma, as well as in the BM. BM was observed beneath the surface epithelium, along the walls of the pockets, around newly formed primordial follicles and capillaries. Interstitial collagen was a prominent component in the stroma of the ovary on all 3 days; the quantity increased from p l through p3. Collagen bundles were found a s discontinuous strands beneath the surface epithelium, but were continuous strands along the outside of the wall of the pockets and in the interstitium between them. A small bundle of collagen was found at the hilum. This bundle of collagen extended from the hilum of the ovary as a stalk (Fig. 11A). All these collagen bundles were more prominent on p2. The subepithelial collagen bundles were more distinct than on p l ; the collagen bundles along the walls of the pockets had increased (Fig. 10B). By p3, the volume of the collagen bundles had inPrimordial Follicles creased further at the hilum along with the increased We chose to define a true primordial follicle as a unit size of the ovary. The collagen bundles beneath the that is enveloped by its own continuous and complete surface epithelium were distinct and continuous. Fine BM (Guraya, 1985). The primordial follicles were thus bundles were found along the newly formed BM, identified by verifying the existence of a completely around the follicles and also in the interstitium between the follicles (Fig. 1lC).

Fig. 7. Histology of pockets at the periphery and the core of the ovary on p2. (A) A few presumptive granulosa cells (arrowheads) within an oocyte cluster toward the periphery. Subepithelial mesenchymal cells (arrows) found between the surface epithelium and oocyte clusters show the pockets that are closed towards the periphery. Original mag. x 400. (B) Numerous presumptive granulosa cells (arrowheads) in the middle of an oocyte cluster in the core. Mesenchymalcells (arrows) form the wall of the pockets. Original mag. x 400.

Histology of Rete Ovarii In 0.5-pm-thick cross sections, the walls of the ER showed pseudostratified large columnar cells that stained dark with toluidine blue. These cells were organized on a distinct basement membrane (BM). Annulations of spindle shaped cells were found around the tubules (Fig. 12). The CR had similar histology as that of the ER, but the tubules were more convoluted and

184

RAJAH ET AL.

Fig. 8. True primordial follicle in the core of the ovary on p2. A single layer of presumptive granulosa cells (arrowheads) are opposed by mesenchymal cells (arrow). Original mag. x 640.

Fig. 9. True primordial follicles on p3. Most of the oocytes are assembled into primordialfollicles (arrow) and are bound by a BM. Original mag. x 787.5.

HISTOGENESIS OF THE NEONATAL RAT OVARY

185

Fig. 10. Follicle-like structure (arrow) on p l . (A) Toward the periphery of the ovary; no mesenchymal cells found opposing the epithelioid cells (arrowheads) that surround the oocyte. Original mag. x400. (6)At the core of the ovary; some of the cells that surround the oocytes are opposed by mesenchymal cells (arrowhead). Original mag. x 400.

compactly arranged beneath the surface epithelium than that of ER in the connective tissue sac (Fig. 13). Histologically, the IR could be distinguished from the CR by its larger lumen and the loosely arranged short stumpy cells that formed its wall. The widened lumen of the IR seemed to engulf the nearby oocytes on p l (Fig. 14A). The oocytes within the IR were in contact with the rete cells (Fig. 14B) that were either forming the wall of the rete or were found loosely organized within the lumen. At these “communication points” where the oocyte-rete cell association is formed, some follicle like structures were observed. The BM that was around the IR at these terminal ends seemed to have disappeared. A few of the oocytes within these rete tubules were degenerating, and these oocytes were not completely surrounded by rete cells and were exposed to stromal cells (Fig. 14C). On p2, numerous such communication points were observed in the core as well as on the two tips of the ovary. By p3, the IR was still observed, but the number of communication points appeared to have diminished. The total rete volume had increased dramatically on p2, but was reduced on p3. On pl, the rete volume was 2.1 x lo5 km3 and on p2, it was 1.4 x lo6 pm3. By p3, it was 8.8 x lo5 km3. In contrast, the ovary and its collagen bundles increased steadily in size over the 3-day period.

Three-Dimensional Image The three dimensional reconstruction clarified the relationship between all three parts of the rete tubules and the other tissues in the ovary (Fig. 15). On pl, the ER was extensive and stood at a right angle to the surface of the cranial end of the ovary, while the IR and CR combined together, occupied less area than the ER and were confined to interior and the core of the ovary respectively (Fig. 15A). On p2, the ER was still at the cranial end, while the C-shaped organization of the CR and the IR was found within the dorsal flank more toward the core rather than toward the cortex of the ovary (Fig. 15B). The increased area of the CR and its subepithelial position within the ovary was also observable in this reconstruction. By p3, though ER was still prominent, it was more extensively engulfed by the growing extra ovarian tissue and was positioned parallel to the surface of the ovary. The CR was still subepithelial, extending between the ER and the IR (Fig. 15C). On p l , the mesenchymal pockets were all open toward the surface (Fig. 15D), but were closed on p2 (Fig. 15E). The shape of the ovary became more of a tight C on p2, from an open C on p l , which reflected on the organization of the pockets. There was an increase in the number of stromal cells as the pocket like orga-

186

RAJAH ET AL.

nization was completed on p2. The ability to reconstruct selected tissue units from the ovary and rotate them at desired angles revealed the close proximity of the newly formed primordial follicles during this same time period (Fig. 15GH). When the outlines of the ovary were added to this reconstructed structure, the space that was occupied by the oocytes that had yet not been assembled into primordial follicles became obvious (Fig. 151). On p3, the number of true primordial follicles increased tremendously (Fig. 15F).

DISCUSSION The majority of authors who have examined the early neonatal ovarian histology of the rat confined their observations to the structure, development, and number of oocytes (Arai, 1920; Beaumont, 1962; Franchi and Mandl, 1963). Most of the detailed histological observations were done during early embryonic (Satoh, 1985) and late neonatal days (Merchant, 1975; Jackson, 1913).During these time periods either the process of follicular assembly was yet to begin, or had already been completed. Our observation of rat ovarian histology during the first 72 neonatal hours focuses on: (1)the time period of first primordial follicle assembly, (2) the heterogeneity in the organization of specific tissue types within the ovary, (3) various cellular associations, and (4)the changing architecture of rete tubules during the assembly of primordial follicles. Heterogeneous Histology The neonatal rat ovary is an “anisometric” (Elias and Hyde, 1983) structure and has heterogeneous organization of different cell types. Therefore, a random section from the ovary during this time period cannot be considered “representative” of the entire ovary; even a random “pocket” within a single section does not represent the distribution of oocytes and presumptive granulosa cells in all the pockets in that section. Within individual pockets on p l , some are packed with numerous oocytes in contact with each other and few epithelial cells around the oocytes (cortex), whereas others have numerous presumptive granulosa cells intervening between the oocytes (medulla). In a longitudinal section running parallel to the flanks these pockets look like large cylindrical structures. But when cut across the C from right to left or top to bottom the pockets have much smaller diameters. This suggests a

Fig. 11. Aniline blue stained collagen bundles showing the organization of the stroma. (A) p l ovary: stromal pockets (small arrows) open toward the periphery. The unstained regions are the oocyte clusters. These clusters are in contact with the surface epithelium. The thick collagen bundle stained with aniline blue at the core (large arrow) of the ovary is the stalk of the ovary. Original mag. x 200. (6)p2 ovary: the stromal pockets (arrow) are closed towards the surface. Original mag. x 400. (C) p3 ovary: each black circle (arrowheads) indicates the Dosi-

HISTOGENESIS OF THE NEONATAL RAT OVARY

Fig. 12. p l ovary: ER tubules in cross section stained with toluidine blue. These tubules are in a connective tissue pouch. The basal layer of the rete cells are opposed by the stromal cells (arrow) where there is a BM (the unstained clear line). Original mag. x 630.

187

Fig. 13. CR beneath the surface epithelium on p2; CR tubules at the point where it enters the ovary beneath the surface epithelium (arrow). Original mag. x 630.

188

RAJAH ET AL.

Fig. 14A-B

HISTOGENESIS OF THE NEONATAL RAT OVARY

189

These heterogeneously organized cell types show a dynamic reorientation during the first 72 hr postpartum which results in the assembly of primordial follicles. Within these 72 hr from the time of birth: (1)the rete tubules shift in position relative to the ovary and also form new associations with the ovarian tissues; (2) the oocytes that were in close contact with each other become separated by intervening presumptive granulosa cells that form a single layer around each oocyte; and (3) the mesenchymal cells “invade” along with the presumptive granulosa cells and an encircling BM is formed around this single layer of presumptive granulosa cells. The three-dimensional reconstruction revealed features that could not be seen from the single sections. For example, the ability to reconstruct selected components of the ovary in three dimensions clarified the proximity of the rete tubules and the newly formed primordial follicles. This observation provides empirical support for the concept that the rete ovarii may be involved in the formation of follicles.

Fig. 14. “Communication point.” (A) IR in the core of the ovary on p l . Section from the middle region of the ovary showing the terminal end of the IR (arrows) with numerous oocytes within the lumen at the core of the ovary. Original mag. x 400. (B)The region within the black square in A enlarged to show the IR (arrows) in the core of the ovary. One of the oocytes is surrounded by some cells (arrowhead) giving a follicle-like appearance. Original mag. x400. (C) IR tubule with the enclosed oocytes in a row that have follicle-like appearance on p2. The degenerating oocytes (arrow) are not completely surrounded by rete cells. Original mag. x630.

funnel like structure of the pockets that are ellipsoid in cross sections, with its longest diameter running parallel to the flanks. On p l and p2, the sections that pass through the periphery at any plane will give the pockets a circular appearance. This is due to the large clusters of oocytes toward the periphery, which are the last ones to be enclosed into pockets. Also, the association of the pockets with the rete tubules at the core of the ovary will be observable only a t certain planes of sections and hence could be easily missed. The heterogeneity in cellular distribution and tissue organization must be taken into consideration in morphometric studies of the neonatal ovaries. For example, studies seeking to quantify the number of oocytes in neonatal ovaries could incur serious errors unless a detailed sampling procedure is used (Elias and Hyde, 1983).

Cell-Cell Recognition and Interaction Our observations indicate that the formation of primordial follicles occurs within a narrow window of time; the second 24 hr postpartum. Future studies should focus on the morphogenic processes that must be occurring during this time period (cell proliferation, cell migration, cell-cell recognition and interaction, and cell differentiation). A chief consideration for future studies should be the dramatic changes in organization of basement membranes during this period. The basement membrane determines the margin of epithelial and stromal cell compartments (Vracko, 1982). We would like t o suggest that, for the basement membrane to be laid down beneath the basal side of the epithelial cells, the presence of both the epithelial (surface or presumptive granulosa) and the mesenchymal (stromall cells are essential. We would like to speculate that the mesenchymal cells direct the formation of primordial follicles, because their presence is essential for the organization of BM; the follicular BM is found only when the stromal cells from the wall of the pocket start invading the presumptive granulosa cell-oocyte clusters.

A Hypothesis to Explain Follicular Assembly Hisaw (1947) suggested that “the arrangement of the granulosa of a primordial follicle is under the control of the ovum.” Peters (1978) emphasized the dynamic role of presumptive granulosa cells by saying that, “the oocyte clusters are to be separated by the presumptive granulosa cells to form primordial follicles.” MerchantLarios and Chimal-Monroy (1989) vaguely referred to the same process as “a problem of somatic4ocyte interaction leading to the splitting up (fragmentation) of sex cords,” but did not define the specific cell interactions involved in the assembly of primordial follicles. We would like to extend these speculations to include

190

RAJAH ET AL.

Fig. 15.

HISTOGENESIS OF THE NEONATAL RAT OVARY

191

the specific role of stromal cells in the process of follicular assembly. Although an oocyte-presumptive granulosa cell interaction is necessary to begin follicle formation, we suggest that presumptive granulosastromal (mesenchymal) cell interaction also is essential for the completion of the assembly of primordial follicles. We base our hypothesis on the fact that a primordial follicle is bound by a complete and continuous BM (Guraya, 1985). We propose that the process of BM organization is the essential step in the completion of the follicular assembly. It has been shown that in all types of tissues BM is organized only when an epithelial/endothelial-mesenchymal interaction takes place. We propose that in ovary the presumptive granulosa cells and stromal cells interact to form a BM. Therefore, both the cellular interaction between the oocytepresumptive granulosa cells as well as presumptive granulosa-stromal cells are essential for follicular assembly.

Most significantly, those oocytes that are first assembled into the primordial follicles are those that are closest to the rete tubules. This suggests that in addition to its possible role in providing the granulosa cells as postulated by Byskov (19751, the rete may have an inductive influence during the assembly of follicles.

sections added to show the space that is occupied by the oocytes that have not yet been assembled into primordial follicles. All the newly formed primordial follicles are close to the rete tubules.

Amsterdam, A,, and Rotmensch, S. (1987) Structure-function relationships during granulosa cell differentiation. Endocr. Rev. 8(3): 309-337.

EXPERIMENTAL PROCEDURES Sprague-Dawley-derived female rat pups, 1through 3 days old, were used in this investigation. Timed pregnant rats were housed under controlled conditions (23“C,12 hr light and 12 hr darkness). On the expected day of parturition the pregnant mothers were observed every hour on the hour. The time of birth was considered as time zero and the following 24 hr as day one. Thus our description of “days” does not always correspond to the solar day. At least 5 animals were studied for each day. The ovaries were collected a t 24-hr intervals. The animals were anesthetized with ice and perfused with cold saline (until the liver cleared, about 10 min) followed by 2.5% glutaraldehyde in 0.15 M caRete Ovarii in Rat Ovary codylate buffer (10 min), then immediately transferred The rete ovarii had been hypothesized to be one of to the same fixative and fixed overnight. The specithe sources for granulosa cells, and was therefore sug- mens were washed extensively with 0.2 M cacodylate gested to play a central role in the assembly of primor- buffer (pH 7.21, postfixed in Dalton’s fixative (Dalton, dial follicles (Byskov, 1975, 1977; Wenzel and Odend- 1955) for 1 hr at 4”C, washed with 0.15 M phosphate hal, 1985). Our observations support this hypothesis. buffer, dehydrated in a series of alcohols, infiltrated, We found that there were major changes in the rete and embedded in Spur (Electron Microscopy Sciences, during the discrete window of time when most of the Ft. Washington, PA 19034). Thin serial sections (500 follicles are assembled. During this time period the nm) were prepared and stained with toluidine blue. communication points between IR and oocytes in- These sections were used for light microscopic observacreased in number. The volume of the rete also in- tions and three-dimensional reconstruction of the creased during this time, then regressed thereafter. ovary using a PC-based Image Combining Microscope system. This system is essentially a computerized camera lucida (“Neurolucida,” Microbrightfield, Inc., ColChester, VT). The user traces regional contours and Fig. 15. Computerized three-dimensional reconstruction of ovaries. marks and identifies cell bodies according to location Color codes: blue, outline of the section of the ovary; red, outline of the and type directly upon the tissue image. A computer rete tubules; green, outline of the pockets; yellow markers, primordial controlled stage makes it possible to view and digitize follicles. (A) p l ovary showing the shape of the ovary and the rete tuan unlimited expanse of tissue a t high magnification; bules. The portion of the rete extending from the surface is ER. The successive tissue sections have their depth separations curved portion of the rete that is in the core of the ovary consists of CR and IR. (B)p2 ovary showing the shape of the ovary and the rete tubules. preserved as part of the image acquisition procedure. The ER still stands at right angle to the surface of the cranial tip that is the The computer graphic representations of individual inferior end in this picture. The extensive rete tubules within the ovary are sections are aligned with respect to one another during mostly CR on the back of the C and IR in the core of the ovary; (C) p3 tracing. Data analysis includes lengths, surface areas, ovary showing the shape of the ovary and the rete tubules. The portion of the rete outside the ovary lying parallel to the surface of the Cis ER. The volumes, cell counts, etc. for the entire reconstruction portion that seems to be within the ovary is also mostly on the surface of (Glaser and Glaser, 1990). Another set of ovaries colthe ovary forming the ER. Some of these tubules constitute the CR and lected after saline perfusion was fixed overnight in IR. (D)Walls of the pockets from p l reconstructed. Cylindrical and/or for 5-pm-thick paraf’fin funnel-like pockets originate from the core of the ovary and extend toward Kahle’s fixative and processed sections. These sections were stained with aniline blue the periphery that remain open. (E)Walls of the pockets from p2 reconstructed to show the closed ends of the pockets toward the periphery. (F) to observe collagen bundles. p3 ovary showing the shape of the ovary, primordial follicles, and the rete ACKNOWLEDGMENTS tubules. Primordial follicles are found in almost all the regions of the ovary. (G) Rete tubules and the primordial follicles reconstructed from p2 This research was supported by NIH Grant HD vary in the same plane parallel to the plane of C as the sections were placed one over the other. (H)G rotated at 90” along the Xaxis showing 27194. a different view that reveals the proximity of the newly formed primordial REFERENCES follicles to that of the rete tubules. (I) Same as H with the outline of the

192

RAJAH ET AL

Anderson, E., Little, B., and Lee, G. S. (1987) Androgen-induced changes in rat granulosa cells i n uitro. Tissue Cell. 19(2):217-234. Arai, H. (1920) On the postnatal development of the ovary (albino rat), with especial reference to the number of ova. Am. J . Anat. 27(4):404-462. Beaumont, H. M., and Mandl, A. M. (1962) A quantitative and cytological study of oogonia and oocytes in the foetal and neonatal rat. Proc. Roy. SOC. B. 155:557-579. Bjorkman, N. (1962) A study of the ultrastructure of the granulosa cells of the rat ovary. Acta. Anat. 51:125-147. Byskov, A. G. (1975) The role of rete ovarii in meiosis and follicle formation in the cat, mink and ferret. J. Reprod. Fertil. 45201-209. Byskov, A. G., and Lintern-Moore, S. (1973) Follicle formation in the immature mouse ovary: The role of the rete ovarii. J. Anat. 116: 207-2 17. Byskov, A. G., Skakkebaek, N. E., Stafanger, G., and Peters, H. (1977) Influence of ovarian surface epithelium and rete ovarii on follicle formation. J. Anat. 123:77-86. Carnegie, J. A,, Byard, R., Dardick, I., and Tsang, B. K. (1988) Culture of granulosa cells in collagen gels: The influence of cell shape on steroidogenesis. Biol. Reprod. 38:881-890. Carson, R., and Smith, J. (1986) Development and steroidogenic activity of preantral follicles in the neonatal rat ovary. J. Endocr. 110:87-92. Costa Gudes, M. L., and Miraglia, T. (1977) The rete ovarii and follicle formation in marmosets. Acta Histochem. Bd. 60.s:247-252. Dalton, A. J. (1955) A chrome-osmium fixative for electron microscopy. Anat. Rec. 121:281. Dawson, A. B., and McCabe, M. (1951) The interstitial tissue of the ovary in the infantile and juvenile rats. J . Morphol. 88543-571. Elias, H., and Hyde, D. M. (1983) Some elementary geometric concepts. In: “A Guide to Practical Steriology.” New York: Krager, 8. Franchi, L. L., and Mandl, A. M. (1963) The ultrastructure of oogonia and oocyte in the foetal and neonatal rat. Proc. Roy. SOC. B. 157: 99-114. Glaser, J., and Glaser, E. M. (1990) Neuron imaging with neurolucida. A PC-based system for image combining microscopy. Comput. Med. Imaging Graph. 14(5):307-317. Guraya S.S. (1985) Primordial follicle. In: “Bioiogy of Ovarian Follicles in Mammals.” Berlin: Springer-Verlag, pp 3, 11. Hisaw, F. L. (1947) Development of the Graffian follicle and ovulation. Physiol. Rev. 27:95-119. Jackson, C. M. (1913) Postnatal growth and variability of the body and of various organs in the albino rat. Am. J . Anat. 15(1):40. Lawrence, T. S., Ginzber, R. D., Gilula, N. B., and Beers, W. H. (1979) Hormonally induced cell shape changes in cultured rat ovarian granulosa cells. J. Cell Biol. 80:21-36. Marx, L., and Bradbury, J. T. (1940) Correlation of ovarian histology and intersexuality of the genital apparatus with special reference to APL treated infantile rats. Anat. Rec. 78:79-103. Merchant, H. (1975)Rat gonadal and ovarian organogenesis with and without germ cells: An ultrastructural study. Dev. Biol. 44:l-21. Merchant-Larios, H. (1978) Ovarian differentiation. In: “The verte-

brate Ovary,” R. Jones, R. E. (edj. New York: Plenum Press, pp 47-81. Merchant-Larios, H. (1979) Origin of somatic cells in the rat gonad: An autoradiographic approach. Ann. Biol. Anim. Biochem. Biophys. 19(4B):1219-1229. Merchant-Larios, H., and Chimal-Monroy, J . (1989) The ontogeny of primordial follicles in the mouse ovary. In: “Developments in U1trastructure of Reproduction.” Motta, P. M. (ed). New York: Liss, pp 55-63. Ohno, S., and Smith, J. B. (1964) Role of fetal follicular cells in meiosis of mammalian oocytes. Cytogenetics 3:324-333. Ojeda, S. R., and Urbanski, H. F. (1988) Puberty in the rat. In: “The Physiology of Reproduction.” Knobil, E., and Neil, J., et al. (eds). New York, Raven Press. p. 1699. Paranko, J. (1987) Expression of type I and type 111 collagen during morphogenesis of fetal rat testis and ovary. Anat. Rec. 219:91-101. Peters, H. (1978) Folliculogenesis in mammals. In: “The Vertebrate Ovary. Comparative Biology and Evolution.” Jones, R. E. (ed).New York: Plenum Press, p 126. Peters, H. (1979) Some aspects of follicular development. In: “Ovarian Follicular Development and Function.” Rees Midgley, A,, and William, A. (eds). New York: Raven Press, pp 1-13. Prepin, J., Gibello-Kervran, C., Charpentier, G., and Jost, A. (1985) Number of germ cells and meiotic prophase stages in fetal rat ovaries cultured in vitro. J. Reprod. Fertil. 73579-583. Rennels, E. G. (1951) Influence of hormones on the histochemistry of ovarian interstitial tissue in the immature rat. Am. J. Anat. 88: 63-107. Satoh, M. (1985) The histogenesis of the gonad in rat embryos. J. Anat. 143:17-37. Stein, L. E., and Anderson, C. H. (1979) A qualitative and quantitative study of rete ovarii development in the fetal rat: Correlation with the onset of meiosis and follicle cell appearance. Anat. Rec. 193:197-212. Torrey, T. W. (1945) The development of the urinogenital system of the albino rat. Am. J. Anat. 76375-397. Ueno, S., Takahashi, M., Manganaro, T. F., Ragin, R. C., and Donahoe, P. K. (1989) Cellular localization of mullarian inhibiting substance in the developing rat ovary. Endocrinology 124(3):10001006. Vracko, R. (1982) The role of basal lamina in maintenance of orderly tissue structure. In: “New Trends in Basement Membrane Research.’’Kuehn, K., Schoene, H. H., and Timpl, R. (eds).New York: Raven Press, pp 1-7. Wenzel, J . G. W., and Odend’hal, S. (1985) The mammalian rete ovarii: A literature review. Cornell Vet. 75:411-425. Wordinger, R., Sutton, J., and Brun-Zinkernagel, A. (1990) Ultrastructure of oocyte migration through the mouse ovarian surface epithelium during neonatal development. Anat. Rec. 227:187-198. Zamboni, L. (1979) The role of mesonephros in the development of the sheep ovary. Ann. Biol. Anim. Biochim. Biophys. 19(4B):11531178.