Morphometric analysis of the uterine endometrium of swine on days 12 and 16 postestrus.код для вставкиСкачать
THE ANATOMICAL RECORD PART A 270A:59 – 66 (2003) Morphometric Analysis of the Uterine Endometrium of Swine on Days 12 and 16 Postestrus DIANNA M. BLACKWELL,1,3 ROBERT C. SPETH,2,3 AND MARK A. MIRANDO1,3* 1 Department of Animal Sciences, Washington State University, Pullman, Washington 2 Department of Veterinary Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, Washington 3 Center for Reproductive Biology, Washington State University, Pullman, Washington ABSTRACT The uterine endometrium of swine is comprised of luminal epithelial, glandular epithelial, and stromal cells that secrete the luteolysin, prostaglandin F2␣ (PGF2␣), during late diestrus. However, which of these cells contribute the most to luteolytic PGF2␣ secretion is unknown because the cellular composition of the endometrium has not been quantiﬁed. Therefore, this study quantiﬁed the cellular composition of the endometrium on days 12 and 16 postestrus by histologic and morphometric analyses. On day 12, the endometrium consisted predominantly of stromal cells (47% of total cell quantity) and glandular epithelial cells (37%), whereas luminal epithelial cells represented only 16% of the total of the three cell types. The number of glandular epithelial cells tended to increase (P ⬍ 0.10) between days 12 and 16, such that they comprised 45% of the endometrium by day 16, while the number of stromal and luminal cells did not change and accounted for 45% and 10% of the cells, respectively. Luminal epithelial cells had a 58% greater cross-sectional area (P ⬍ 0.001) than glandular epithelial cells, whereas glandular epithelial cells had a 22% greater area (P ⬍ 0.001) than stromal cells. Glandular epithelial cells decreased (P ⬍ 0.001) in cross-sectional area between days 12 and 16, whereas the area of luminal epithelial and stromal cells remained unchanged. These results indicate that the porcine endometrium is comprised predominantly of stromal and glandular epithelial cells that are likely to contribute substantially to endometrial PGF2␣ secretion during luteolysis. The contribution of glandular epithelium to luteolytic PGF2␣ secretion probably increases during diestrus as the number of these cells increases. Anat Rec Part A 270A:59 – 66, 2003. © 2003 Wiley-Liss, Inc. Key words: endometrium; uterus; epithelial cell; stromal cell; morphometry; pig Prostaglandin F2␣ (PGF2␣) is released from the uterine endometrium in a highly pulsatile manner during the luteolytic period of late diestrus in domestic ungulates to terminate corpus luteum (CL) function, and oxytocin is the major known stimulus for pulsatile secretion of PGF2␣ (Silvia et al., 1991; Mirando et al., 1996; McCracken et al., 1999). In ruminants, epithelial cells are entirely responsible for the oxytocin-stimulated PGF2␣ secretion, and stromal cells are completely unresponsive to oxytocin (Asselin et al., 1996). In contrast, stromal cells are the porcine endometrial cells that are the most responsive to oxytocin, whereas luminal epithelial cells are completely unresponsive and glandular epithelial cells display an intermediate response (Uzumcu et al., 1998; Braileanu et al., 2001; Hu © 2003 WILEY-LISS, INC. et al., 2001a, b). This pattern of response apparently is opposite to that for expression of oxytocin and oxytocin Grant sponsor: National Institutes of Health; Grant number: HD30268. *Correspondence to: Mark A. Mirando, National Research Initiative Competitive Grants Program, 1400 Independence Ave. SW, Stop 2241, U.S. Department of Agriculture, Washington, DC 20250-2241. Fax: (202) 205-3641. E-mail: email@example.com Received 9 April 2002; Accepted 15 August 2002 DOI 10.1002/ar.a.10182 60 BLACKWELL ET AL. receptors in these cells (Boulton et al., 1995, 1996). Although stromal cells showed a greater PGF2␣ secretory response to oxytocin than did luminal or glandular epithelial cells (Uzumcu et al., 1998; Braileanu et al., 2001; Hu et al., 2001a, b), stromal cells apparently secrete less total PGF2␣ per cell than do epithelial cells, especially compared to those of the luminal surface (Zhang et al., 1991; Uzumcu et al., 1998; Hu et al., 2001a, b), because of the greater basal secretion of PGF2␣ from epithelial cells in the absence of oxytocin stimulation. This would seem to indicate that stromal cells make a relatively minor contribution to luteolytic PGF2␣ secretion in swine despite being more responsive to oxytocin than the two types of epithelial cells. However, this suggestion completely ignores the relative contribution of each cell type to the total morphological composition of the endometrium. If stromal cells comprise a substantial portion of the endometrium in pigs, then they likely also would contribute signiﬁcantly to luteolytic secretion of PGF2␣ during late diestrus. However, to our knowledge morphometric analysis of the cell types in the endometrium has not been reported for pigs or any other species. It was hypothesized that stromal cells are a major constituent of the endometrium in cyclic gilts. Therefore, the primary objective of this study was to determine quantitatively the tissue and cellular composition of the endometrium during the luteolytic period of late diestrus in pigs (i.e., days 12–16) using standard accepted stereological morphometric procedures (Chalkley, 1943; Abercrombie, 1946; Weibel, 1979; Mirando et al., 1989). A secondary objective was to determine if the tissue and cellular composition of the endometrium changed as the luteolytic period progressed from just before the onset of endometrial responsiveness to oxytocin and pulsatile PGF2␣ secretion (i.e., on day 12) to the middle of luteolysis on day 16 (Carnahan et al., 1996; Edgerton et al., 1996). MATERIALS AND METHODS Animals Fourteen crossbred peripubertal gilts (Landrace, Yorkshire, Large White, Hampshire, and Duroc) were observed daily in the presence of an intact boar for standing estrous behavior. Gilts that showed at least one normal interestrous interval of 18 –24 days were allotted, in a completely randomized design, to undergo hysterectomy 12 (n ⫽ 8) or 16 (n ⫽ 6) days after the second or third estrus. At hysterectomy, the uterus was excised, immediately and carefully trimmed of the mesometrium, and weighed. Three samples of tissue, approximately 4 – 6 cm in length, were obtained from one randomly selected uterine horn approximately 5 cm from the uterine body, 5 cm from the utero-tubal junction, and from the center of the uterine horn. Each sample was placed immediately in 100 ml of phosphate-buffered (pH 7.4) 4% paraformaldehyde ﬁxative. Additional tissue from each of the three regions of the uterus was collected, placed in plastic bags, and kept on ice. The tissue samples on ice were transferred to the laboratory where measurements of speciﬁc gravity were obtained by weighing 4- to 6-g portions and then submerging them in a known volume of deionized water to determine volume displacement. Duplicate determinations were made on each region for a total of six measurements per uterus. Histological Preparation of Tissue Paraformaldehyde ﬁxative was replaced with fresh ﬁxative 1 and 3 hr after initial tissue immersion, at which times approximately 0.5 cm was trimmed from each end of the tissue to increase the ratio of ﬁxative volume to tissue mass and to permit increased access of ﬁxative to the center of the tissue while maintaining integrity of the gross tubular anatomy of the tissue. After 48 hr, both ends of the tissue were trimmed to obtain a cross section of tissue approximately 0.5–1 cm in length from near the center of the tissue sample. Tissues were stored in 70% ethanol until cleared, inﬁltrated, and embedded in parafﬁn. Tissues were sectioned at 3 m, and two complete cross sections, 30 m apart, were obtained. After being mounted on glass slides, tissue sections were deparafﬁnized, cleared, stained with Mayer’s hematoxylin, counterstained with eosin/phloxine, and then dehydrated (Humason, 1979), after which coverslips were applied using a drop of Richard Allen mounting medium (Richard Allen Scientiﬁc, Kalamazoo, MI). Histological and Morphometrical Analyses The cross-sectional areas of stromal, glandular epithelial, and luminal epithelial cells (Fig. 1) were measured using a Nikon Labophot microscope (Nikon USA, Melville, NY) attached to an MCID image analysis system (Imaging Research, Inc., St. Catherine, Ontario, Canada). After magniﬁcation at 80⫻, images were displayed on a 33-cm video monitor. A minimum of 35 cells (range ⫽ 35–105) of each type was quantiﬁed on each tissue section obtained from the three regions of the uterus. The cellular crosssectional area and total scanned area were measured in pixels, and data were then converted to area by determining the number of pixels per unit area using a micrometer slide for calibration. The proportion (percentage area) of uterus consisting of endometrium, myometrium, and perimetrium (Fig. 1) was determined by light microscopy at 100⫻ magniﬁcation using a 100-point ocular graticule by the method of Chalkley (1943), as described previously for testicular tissue (Mirando et al., 1989). Four determinations, spanning the cross section from perimetrium to uterine lumen, were made on each section. The proportions (percentage area) of luminal epithelial, glandular epithelial, and stromal tissue within the endometrium were determined similarly, with four determinations made on each section. Calculations All calculations were made such that measurements of variables for each of the three regions of the uterus contributed equally to the mean values calculated. Uterine volumes (cm3) were obtained by dividing the weight of each uterus by its mean speciﬁc gravity measurement. Volumes of endometrium, myometrium, and perimetrium were then obtained by multiplying the mean proportion of each tissue within each uterus by the total volume of each respective uterus (Mirando et al., 1989). Total volumes of luminal epithelium, glandular epithelium, and stroma within the endometrium of each uterus were determined similarly by multiplying the mean proportion of each tissue component within each uterus by the total volume of endometrium within each uterus. The true cell count within a given scanned volume of endometrial tissue was determined from the crude cell count, section thickness, MORPHOMETRY OF PORCINE UTERINE ENDOMETRIUM 61 Fig. 1. Representative hematoxylin and eosin/phloxine-stained photomicrographs showing uterine cross sections from day 12 (left column) and day 16 (right column) pigs at 40⫻ (A) and 400⫻ (B) magniﬁcation. The uterine lumen, endometrium (ENDO), myometrium (MYO), and pe- rimetrium (PERI) are indicated in the top panel. Luminal epithelial cells (LEC), glandular epithelial cells (GEC), stromal cells (SC), and blood vessels (BV) of the endometrium are indicated by arrows in the bottom panel. and cellular diameters (Abercrombie, 1946; Weibel, 1979; Mirando et al., 1989), whereby cellular diameters were calculated from the cellular cross-sectional area of individual cells using the formula for a circle, as described previously (Mirando et al., 1989). The numbers of luminal epithelial, glandular epithelial, and stromal cells per uterus were then obtained from the product of the true cell count per scanned volume of endometrial tissue and volume of endometrial tissue component per uterus (i.e., number of stromal cells per uterus ⫽ stromal cells per unit volume stroma ⫻ stromal volume per uterus, etc.). Although the stromal, glandular epithelial, and luminal epithelial cells examined in the present study were not spherical, a previous study demonstrated that morphometric analyses derived from diameters calculated from cellular components that were highly irregular in shape yielded similar results to those derived from diameter measurements made on spherical cellular components (Mirando et al., 1989). Statistical Analyses Data were subjected to least-squares analysis of variance (ANOVA) for a completely randomized design using the general linear models (GLM) procedure of the Statistical Analysis System (SAS, 1996). Data for uterine weight, speciﬁc gravity, uterine volume, volumes of uterine tissue components (i.e., endometrium, myometrium, and perimetrium), volumes of endometrial tissue components (i.e., stroma, glandular epithelium, and luminal epithelium), and numbers of cells were analyzed for a oneway treatment structure with day as the main effect. Data for cell size (i.e., cross-sectional area of stromal, glandular epithelial, and luminal epithelial cells) were analyzed for a split-plot ANOVA with a three-way treatment structure. Day was the whole-plot effect, gilt was nested within day, and the subplot effects were region of the uterus (i.e., anterior, middle, and posterior) and cell type. Data for proportion (percentage area) of uterine tissue components 62 BLACKWELL ET AL. Fig. 3. Mean (⫹ SEM) volumes of endometrium, myometrium, and perimetrium on days 12 and 16 postestrus in cyclic gilts. The volumes of endometrium, myometrium, and perimetrium did not differ between days 12 and 16. Fig. 2. Mean (⫹ standard error of mean (SEM)) proportion of uterine endometrium (top) and myometrium (bottom) in the anterior, middle, and posterior regions of the uterus on days 12 and 16 postestrus in cyclic gilts. There was a lower (P ⫽ 0.06) proportion of endometrium and more (P ⬍ 0.07) myometrium on day 16 than on day 12. The anterior region of the uterus also contained a smaller (P ⬍ 0.01) proportion of myometrium and a greater (P ⬍ 0.01) proportion of endometrium than did the middle or posterior regions. and endometrial tissue components were analyzed for a split-plot ANOVA with a two-way treatment structure. Day was the whole-plot effect, gilt was nested within day, and the subplot effect was region of the uterus. Because values for proportion of perimetrium and luminal epithelium approached zero, angular (i.e., arc sin) transformations were initially performed for these two characteristics. However, these transformations had only a miniscule effect on the outcome of the ANOVAs, and therefore, only results for the untransformed data were utilized and reported. All tests of hypotheses were performed using the appropriate error terms according to the expectation of the mean squares (Snedecor and Cochran, 1980). RESULTS Mean weight and volume of the uterus did not differ signiﬁcantly between days 12 and 16 (610 ⫾ 61 g vs. 695 ⫾ 70 g and 577 ⫾ 59 cm3 vs. 662 ⫾ 68 cm3, respectively). Mean speciﬁc gravity of the uterus also did not vary signiﬁcantly between day 12 (1.06 ⫾ 0.01 g/cm3) and day 16 (1.05 ⫾ 0.01 g/cm3). The proportions of endometrium and myometrium (Fig. 2) varied by day (P ⬍ 0.07) and region (P ⬍ 0.01). There was a lower (P ⫽ 0.06) proportion of endometrium and more (P ⬍ 0.07) myometrium on day 16 than on day 12. The anterior region of the uterus also contained a greater (P ⬍ 0.01) proportion of endometrium and a lower (P ⬍ 0.01) proportion of myometrium than did the middle or posterior regions. The volumes of endometrium, myometrium, and perimetrium did not differ between days 12 and 16 (Fig. 3). The proportion of luminal epithelium did not vary by day or region (Fig. 4). The proportion of glandular epithelium varied by region (P ⬍ 0.05), with the posterior region possessing a smaller proportion of glandular epithelium than the middle or anterior regions of the uterus (Fig. 4); however, the interaction of region by day (P ⫽ 0.05) indicated that variation among regions of the uterus was primarily due to a decrease in glandular epithelium in the posterior region on day 16. The proportion of endometrial stroma also varied (P ⬍ 0.01) by region of the uterus, with the posterior region possessing the greatest proportion of stroma, the middle of the uterus containing the least, and the anterior region having an intermediate amount (Fig. 4). Within the endometrium, the volumes of stroma, glandular epithelium, and luminal epithelium did not vary signiﬁcantly by day (Fig. 5). The size of individual luminal epithelial, glandular epithelial, and stromal cells varied by cell type (P ⬍ 0.001) with an interaction of day by cell type (P ⬍ 0.01). Across both days, luminal epithelial cells were 58% larger (P ⬍ 0.001) than glandular epithelial cells, and glandular epithelial cells were 22% larger (P ⬍ 0.001) than stromal cells (Fig. 6). Glandular epithelial cells also decreased (P ⬍ 0.01) in size between days 12 and 16, whereas the size of luminal epithelial and stromal cells remained unchanged (Fig. 6). Cell size also varied signiﬁcantly by region of the uterus (cell ⫻ region interaction, P ⬍ 0.001); glandular epithelial cells were of similar size throughout the uterus, whereas luminal epithelial cells were largest near the uterine body, smallest near the anterior region of the uterus, and intermediate in size near the middle of the uterine horn (Fig. 6). In contrast, stromal cells were larger in the posterior and anterior regions of the uterine horn than toward the middle region. Of the three endometrial cell types examined, the endometrium was comprised predominantly of stromal cells (47%) on day 12 and, to a lesser degree, of glandular epithelial cells (37%), whereas luminal epithelial cells represented only 16% of the three endometrial cell types (Fig. 7). The number of glandular MORPHOMETRY OF PORCINE UTERINE ENDOMETRIUM 63 Fig. 5. Mean (⫹ SEM) volumes of luminal epithelium, glandular epithelium, and stroma within the endometrium on days 12 and 16 postestrus in cyclic gilts. Volumes of endometrial tissue components did not vary signiﬁcantly by day. Fig. 4. Mean (⫹ SEM) proportions of luminal epithelium (top), glandular epithelium (middle), and stroma (bottom) within the endometrium in the anterior, middle, and posterior regions of the uterus on days 12 and 16 postestrus in cyclic gilts. Toward the posterior region of the uterus, the endometrium contained a greater (P ⬍ 0.01) proportion of stroma than did the middle or anterior regions of the uterine horns. In contrast, the proportion of glandular epithelium did not vary by region on day 12, but on day 16, the endometrium nearest the uterine body possessed the smallest (P ⬍ 0.05) proportion of glandular epithelium. The interaction of region by day (P ⫽ 0.05) indicated that variation among regions of the uterus was primarily due to a decrease in glandular epithelium in the anterior region on day 16. epithelial cells tended to increase (P ⬍ 0.10) 54% between days 12 and 16. As a result, they comprised 45% of the three endometrial cell types, while stromal cells also accounted for 45% and luminal cells accounted for the remaining 10% (Fig. 7). DISCUSSION This study demonstrated that the uterine endometrium of cyclic pigs is comprised predominantly of stromal and glandular epithelial cells, whereas luminal surface epithelial cells constituted only 10%–16% of the three endometrial cell types during the luteolytic period of late diestrus. Previous studies with pigs demonstrated that oxytocin induced maximal response from endometrial stromal cells, did not stimulate PGF2␣ secretion from luminal epithelial cells, and promoted an intermediate response from glandular epithelial cells (Uzumcu et al., 1998; Braileanu et al., 2001; Hu et al., 2001a, b). However, the contribution of oxytocin-induced PGF2␣ release from stromal and glandular cells was questioned because these cells secrete less PGF2␣ than do luminal epithelial cells (Uzumcu et al., 1998; Braileanu et al., 2001). Based on the prevalence of Fig. 6. Mean (⫹ SEM) cross-sectional area of luminal epithelial, glandular epithelial, and stromal cells on days 12 and 16 postestrus (top) within the endometrium in the anterior, middle, and posterior regions (bottom) in cyclic gilts. Luminal epithelial cells were larger (P ⬍ 0.001) than glandular epithelial cells, and glandular epithelial cells were larger (P ⬍ 0.001) than stromal cells. Glandular epithelial cells also decreased (P ⬍ 0.001) in size between days 12 and 16, whereas the size of luminal epithelial and stromal cells remained unchanged. Glandular epithelial cells were of similar size throughout the uterus, whereas the size of luminal epithelial and stromal cells varied (P ⬍ 0.001) by region of the uterus. the stromal and glandular cells determined in the present study and their closer proximity to the uterine vasculature, it is suggested now that these two cell types are likely to contribute the most to endometrial secretion of PGF2␣ during luteolysis in pigs. The contribution of glandular epithelium to luteolytic secretion of PGF2␣ probably increases during diestrus as 64 BLACKWELL ET AL. Fig. 7. Mean (⫹ SEM) number of luminal epithelial, glandular epithelial, and stromal cells per uterus on days 12 and 16 postestrus in cyclic gilts. The number of glandular epithelial cells tended to increase (P ⬍ 0.10) between days 12 and 16, but the numbers of luminal epithelial and stromal cells did not change signiﬁcantly. the glands proliferate and the number of these cells increases, as indicated by the results of the current study. The uterine endometrium of the pig responds to cues of ovarian steroids with orderly cycles of tissue proliferation, glandular secretion, and prostaglandin production (Corner, 1921; Geisert et al., 1990a, b, 1994; Gray et al., 2001a). The results of the present study indicated that the number of glandular cells increased 54% between days 12 and 16 after estrus, results that generally are consistent with glandular proliferation from early to mid-diestrus reported previously (Corner, 1921; Erices and Schnurrbusch, 1979). The ability of glandular epithelial cells to secrete PGF2␣ also increases sharply between days 12 and 16 (Uzumcu et al., 2000). Taken together, these results indicate that the total amount of PGF2␣ released by the glandular epithelium increases substantially as diestrus progresses. Moreover, the increase in PGF2␣ release from glandular epithelial cells between days 12 and 16 (Uzumcu et al., 2000) occurred in spite of a decrease in size of these cells. In the endometrium of several species, both epithelial and stromal cells are capable of producing prostaglandins (Schatz et al., 1987; Smith and Kelly, 1987, 1988; Fortier et al., 1988; Zhang et al., 1991; Uzumcu et al., 1998, 2000; Braileanu et al., 2001; Hu et al., 2001a, b), although there is discrepancy as to which cell type is the major source of prostaglandin synthesis. Some reports indicate that stromal cells are almost completely responsible for the PGF2␣ release from the human endometrium (Gal et al., 1982) or that glandular epithelial and stromal cells from the human endometrium secrete similar amounts of PGF2␣ (Schatz et al., 1987), whereas others reported that glandular epithelial cells secrete more PGF2␣ than do stromal cells (Lumsden et al., 1984; Smith and Kelly, 1987, 1988). Similarly, glandular epithelial cells of sheep and cattle endometrium secrete substantially more PGF2␣ than do stromal cells, whereas stromal cells are the major source of PGE2 (Fortier et al., 1988; Cherny and Findlay, 1990). Working with porcine endometrium, Zhang et al. (1991) reported that glandular cells secrete more PGF2␣ than PGE, whereas stromal cells secrete more PGE than PGF2␣; however, the two cell types secrete similar amounts of PGF2␣ when obtained from the endometrium of cyclic pigs. Some studies from our laboratory (Uzumcu et al., 1998) have obtained results similar to those of Zhang et al. (1991), although the results of most experiments indicate that luminal epithelial cells secrete the greatest quantity of PGF2␣, stromal cells secrete the least, and glandular epithelial cells release an intermediate amount (Uzumcu et al., 1998; Braileanu et al., 2001). Collectively, those results suggest that stromal cells make a relatively minor contribution to luteolytic PGF2␣ secretion in swine despite being more responsive to oxytocin than the two types of epithelial cells. However, results of the present study indicate that stromal cells probably are a major source of luteolytic PGF2␣ pulses during late diestrus in pigs because they comprise such a substantial portion of the endometrium. Moreover, stromal cells are in closest proximity to the uterine vasculature, and the PGF2␣ released from them may enter the circulation more readily than that secreted from the luminal surface epithelium or the superﬁcial glands. Although stromal cells are most responsive to oxytocin in vitro and epithelial cells of the luminal surface are least responsive (Uzumcu et al., 1998; Braileanu et al., 2001; Hu et al., 2001a, b), those results seemingly are in contrast to results of studies that localized oxytocin receptor mRNA primarily within the luminal epithelium of the porcine endometrium and minimally within stromal cells (Boulton et al., 1995). However, luminal epithelial cells also synthesize substantially more oxytocin than do stromal cells (Boulton et al., 1996). Thus, oxytocin released by luminal epithelial cells in vitro may act in an autocrine manner to either desensitize cultured luminal epithelial cells to exogenous oxytocin or elevate basal PGF2␣ secretion. Either of these effects would reduce the apparent responsiveness of cultured luminal epithelial cells to oxytocin, and both effects were recently reported to occur in luminal epithelial cells from the pig endometrium (Hu et al., 2001b). In contrast, stromal cells did not exhibit noticeable desensitization of the response to oxytocin (Hu et al., 2001b), which also may contribute to these cells displaying a greater response to oxytocin in vitro than do luminal epithelial cells. In addition to an increase in the number of glandular epithelial cells between days 12 and 16, the proportion of endometrium occupied by glandular epithelium varied among regions of the uterus in the present study. Keys and King (1989) reported that the height of the luminal epithelial cell layer was low on days 10 and 13 of the estrous cycle but increased by days 16 –19, observations that agreed with those of Crombie (1972) and King et al. (1982). Taken together, these results indicate that substantial histological remodeling of the uterine endometrium occurred in the absence of changes in uterine weight or volume. A similar conclusion was also reached by Harney and Bazer (1990). Although endometrial glands are known to proliferate primarily during the follicular phase (Corner, 1921; Gray et al., 2001a), they also undergo hyperplasia and hypertrophy during pregnancy in sheep, cattle, and swine (Gray et al., 2001a), which suggests induction of glandular proliferation by progesterone. Consistent with this suggestion, the results of the present study also indicate that extended exposure to progesterone, as diestrus progresses, may promote a substantial increase in number of glandular cells between days 12 and 16 during the diestrous period. How this could be achieved is not clear because progesterone receptors were not de- MORPHOMETRY OF PORCINE UTERINE ENDOMETRIUM tectable in the superﬁcial glandular epithelium of the porcine endometrium after day 12 postestrus (Geisert et al., 1994) or in the ovine endometrium on days 6 –13 postestrus (Spencer and Bazer, 1995), although it could be mediated by underlying stromal cells, as occurs for the response of epithelium to estrogen in mice (Cooke et al., 1997; Kurita et al., 2000). Alternatively, proliferation of glandular epithelium during diestrus may occur through the concerted actions of prolactin and estradiol, as proposed to occur during pregnancy in pigs (Gray et al., 2001a). Several studies have shown that the growth of porcine conceptuses depends on their position within the uterus. For example, Hagen et al. (1980) reported that runts of the litter occupied the middle portion of the uterine horns during fetal development. Uterine effects on fetal size did not result from fetal crowding because Dziuk (1968, 1985) reported that the capacity of the domestic pig averages about 14 live fetuses, a level of productivity infrequently attained by the pig. Perry and Rowell (1969) found that as the number of fetuses increased, fetuses at the anterior end of the uterus tended to have an advantage in growth and were 5%–10% heavier than those in the middle of the uterus. Fetuses toward the anterior end also tended to have an increased advantage over those at the cervical end (Perry and Rowell, 1969). Similarly, placental weights were greater at day 30 of gestation for fetuses toward the oviductal end of the uterus and were less toward the cervical end of the uterus (Wise et al., 1997). The distribution of epithelial and stromal cells within the endometrium and their corresponding secretions may be critical to conceptus growth. Because uterine glands recently were shown to be essential for embryonic growth and survival in sheep (Gray et al., 2001b, c), growth of pig conceptuses could be directly related to the proportion of glandular epithelium within various regions of the uterus. In the present study, the proportion of glandular epithelium was lowest and the proportion of endometrial stroma was greatest near the uterine body. However, the suggestion that conceptus growth could be related to distribution of glandular epithelium and stroma within the endometrium must be tempered by the fact that results of the present study were obtained from diestrous phase endometrium of cyclic pigs. Results for endometrium of pregnancy could differ as the endometrium undergoes remodeling (Harney and Bazer, 1990). In summary, these results indicate that the porcine endometrium is comprised predominantly of glandular epithelial and stromal cells. Therefore, these two cell types are likely to contribute substantially to endometrial secretion of prostaglandin F2␣ during luteolysis in pigs. The contribution of glandular epithelium to luteolytic release of prostaglandin F2␣ probably increases during diestrus as the glands proliferate and the number of these cells increases between days 12 and 16 postestrus. Alterations in uterine morphology occurred during late diestrus that were entirely attributable to remodeling of the uterus because neither weight nor volume of the uterus changed signiﬁcantly during the luteolytic period. The latter results have potential implications in embryonic and fetal growth if changes in uterine morphology occur similarly during pregnancy. 65 ACKNOWLEDGMENTS The authors are grateful to Drs. Ann A. Wiley and Frank F. Bartol, Auburn University, Auburn, Alabama, for advice on histological preparation of the uterine tissues, and to Dr. Troy L. Ott, University of Idaho, Moscow, Idaho, for photomicrography. The authors also thank Dr. Kwan-Hee Kim and the Histology Core Laboratory at Washington State University for assistance with histological preparation of the tissue. The authors are indebted to the members of M.A. Mirando’s laboratory for assistance with surgery and to the staffs of the Washington State University Swine Center and the Experimental Animal Laboratory Building for assistance in care and handling of animals. LITERATURE CITED Abercrombie M. 1946. Estimation of nuclear population from microtome sections. Anat Rec 94:239 –247. Asselin E, Goff AK, Bergeron H, Fortier MA. 1996. Inﬂuence of sex steroids on the production of prostaglandin F2␣ and E2 and response to oxytocin in cultured epithelial and stromal cells of the bovine endometrium. Biol Reprod 54:371–379. Boulton MI, McGrath TJ, Brown DA, Broad KD, Gilbert CL. 1995. Oxytocin receptor mRNA expression in the porcine uterus. J Reprod Fertil Abstr Ser 15:36 –37. Boulton MI, McGrath TJ, Goode JA, Broad KD, Gilbert CL. 1996. Changes in content of mRNA encoding oxytocin in the pig uterus during the oestrous cycle, pregnancy, at parturition and in lactational anoestrous. J Reprod Fertil 108:219 –227. Braileanu GT, Simasko SM, Hu J, Assiri A, Mirando MA. 2001. Effects of arginine- and lysine-vasopressin on phospholipase C activity, intracellular calcium concentration and prostaglandin F2␣ secretion in pig endometrial cells. Reproduction 121:605– 612. Carnahan KG, Prince BC, Mirando MA. 1996. Oxytocin stimulates uterine secretion of prostaglandin F2␣ in cyclic and early pregnant swine. Biol Reprod 55:838 – 843. Chalkley HW. 1943. Method for quantitative morphologic analysis of tissues. J Natl Cancer Inst 4:47–53. Cherny RA, Findlay JK. 1990. Separation and culture of ovine endometrial epithelial and stromal cells: evidence of morphological and functional polarity. Biol Reprod 43:241–250. Cooke PS, Buchanan DL, Young P, Setiawan T, Brody J, Korach KS, Taylor J, Lubahn DB, Cunha GR. 1997. Stromal estrogen receptors mediate mitogenic effects of estradiol on uterine epithelium. Proc Natl Acad Sci USA 94:6535– 6540. Corner GW. 1921. Cyclic changes in the ovaries and uterus of the sow, and their relation to the mechanism of implantation. Contrib Embryol 13:117–146. Crombie PR. 1972. A new specialization of the granular endoplasmic reticulum in the uterine glands of the pig in the later stages of pregnancy. J Cell Sci 10:693–704. Dziuk PJ. 1968. Effect of number of embryos and uterine space on embryo survival in pigs. J Anim Sci 27:673– 676. Dziuk PJ. 1985. Effect migration, distribution and spacing of pig embryos on pregnancy and fetal survival. J Reprod Fertil Suppl 33:57– 63. Edgerton LA, Kaminiski MA, Silvia WJ. 1996. Changes in uterine secretion of prostaglandin F2␣ in response to oxytocin during the estrous cycle, early pregnancy, and estrogen-induced pseudopregnancy in swine. Biol Reprod 55:657– 662. Erices J, Schnurrbusch U. 1979. Uterine development in swine from birth to age of eight months. Arch Exp Vet 33:457– 473. Fortier MA, Guilbault LA, Grasso F. 1988. Speciﬁc properties of epithelial and stromal cells from the endometrium of cows. J Reprod Fertil 83:239 –248. Gal D, Casey ML, Johnston JM, MacDonald PC. 1982. Mesenchymeepithelial interactions in human endometrium. Prostaglandin synthesis in separated cell types. J Clin Invest 70:798 – 805. Geisert RD, Short EC, Morgan GL. 1990a. Establishment of pregnancy in domestic farm species. In: Zavy MT, Geisert RD, editors. 66 BLACKWELL ET AL. Embryonic mortality in domestic species. Baton Rouge, LA: CRC Press. p 23–51. Geisert RD, Zavy MT, Moffatt RJ, Blair RM, Yellin T. 1990b. Embryonic steroids and the establishment of pregnancy in pigs. J Reprod Fertil Suppl 40:293–305. Geisert RD, Pratt TN, Bazer FW, Mayes JS, Watson GH. 1994. Immunocytochemical localization and changes in endometrial progestin receptor protein during the porcine oestrous cycle and early pregnancy. Reprod Fertil Dev 6:749 –760. Gray CA, Bartol FF, Tarleton BJ, Wiley AA, Johnson GA, Bazer FW, Spencer TE. 2001a. Developmental biology of uterine glands. Biol Reprod 65:1311–1323. Gray CA, Bazer FW, Spencer TE. 2001b. Effects of neonatal progestin exposure on female reproductive tract structure and function in the adult ewe. Biol Reprod 64:797– 804. Gray CA, Taylor KM, Ramsey WS, Hill JR, Bazer FW, Bartol FF, Spencer TE. 2001c. Endometrial glands are required for pre-implantation conceptus elongation and survival. Biol Reprod 64:1608 – 1613. Hagen DR, Kephart KB, Wangsness PJ. 1980. Reproduction in domestic and feral swine. II. Interrelationships between fetal size and spacing and litter size. Biol Reprod 23:929 –934. Harney JP, Bazer FW. 1990. Effects of conceptus and conceptus secretory products on uterine development in the pig. Reprod Fertil Dev 2:179 –187. Hu JB, Braileanu GT, Mirando MA. 2001a. Oxytocin stimulates prostaglandin F2␣ secretion from porcine endometrial cells through activation of calcium-dependent protein kinase C. Prostaglandins Lipid Mediat 65:85–101. Hu J, Ludwig TE, Salli U, Stormshak F, Mirando MA. 2001b. Autocrine/paracrine action of oxytocin in pig endometrium. Biol Reprod 64:1682–1688. Humason GL. 1979. Animal tissue techniques, 4th ed. San Francisco: W.H. Freeman and Co. Keys JL, King GJ. 1989. Structural changes in the luminal epithelium of the porcine uterus between days 10 and 19 of the estrous cycle. Am J Anat 185:42–57. King RJ, Townsend PT, Siddle N, Whitehead MI, Taylor RW. 1982. Regulation of estrogen and progesterone receptor levels in epithelium and stroma from pre- and postmenopausal endometria. J Steroid Biochem 16:21–29. Kurita T, Lee KJ, Cooke PS, Taylor JA, Lubahn DB, Cunha GR. 2000. Paracrine regulation of epithelial progesterone receptor by estradiol in the mouse female reproductive tract. Biol Reprod 62:821– 830. Lumsden MA, Brown A, Baird DT. 1984. Prostaglandin production from homogenates of separated glandular epithelium and stroma from human endometrium. Prostaglandins 28:485– 496. McCracken JA, Custer EE, Lamsa JC. 1999. Luteolysis: a neuroendocrine-mediated event. Physiol Rev 79:263–323. Mirando MA, Hoagland TA, Woody CO, Riesen JW. 1989. The inﬂuence of unilateral castration on testicular morphology and function in adult rams. Biol Reprod 41:798 – 806. Mirando MA, Uzumcu M, Carnahan KG, Ludwig TE. 1996. A role for oxytocin during luteolysis and early pregnancy in swine. Reprod Dom Anim 31:455– 461. Perry JS, Rowell JG. 1969. Variations in foetal weight and vascular supply along the uterine horn of the pig. J Reprod Fertil 19:527– 534. SAS, PC-SAS. 1996. Propriety software release 6.12 system for Windows. Cary, NC: Statistical Analysis System Institute, Inc. Schatz F, Markiewicz L, Gurpide E. 1987. Differential effects of estradiol, arachidonic acid and A 23187 on prostaglandin F2␣ output by epithelial and stromal cells of human endometrium. Endocrinology 120:1465–1471. Silvia WJ, Lewis GS, McCracken JA, Thatcher WW, Wilson L. 1991. Hormonal regulation of uterine secretion of prostaglandin F2␣ during luteolysis in ruminants. Biol Reprod 45:655– 663. Smith SK, Kelly RW. 1987. The effect of estradiol-17␤ and actinomycin D on the release of PGF and PGE from separated cells of human endometrium. Prostaglandins 34:553–561. Smith SK, Kelly RW. 1988. Effect of platelet-activating factor on the release of PGF2␣ and PGE2 by separated cells of human endometrium. J Reprod Fertil 82:271–276. Snedecor GW, Cochran WG. 1980. Factorial experiments. In: Statistical methods, 7th edition. Ames, IA: Iowa State University Press. p 298 –333. Spencer TE, Bazer FW. 1995. Temporal and spatial alterations in uterine estrogen receptor and progesterone receptor gene expression during the estrous cycle and early pregnancy in the ewe. Biol Reprod 53:1527–1543. Uzumcu M, Braileanu GT, Carnahan KC, Ludwig TE, Mirando MA. 1998. Oxytocin-stimulated phosphoinositide hydrolysis and prostaglandin F2␣ secretion by luminal epithelial, glandular epithelial, and stromal cells from pig endometrium. I. Response of cyclic pigs on day 16 postestrus. Biol Reprod 59:1259 –1265. Uzumcu M, Carnahan KC, Braileanu GT, Mirando MA. 2000. Oxytocin-stimulated phosphoinositide hydrolysis and prostaglandin F2␣ secretion by luminal epithelial, glandular epithelial and stromal cells from pig endometrium. II. Response of cyclic, pregnant and pseudopregnant pigs on days 12 and 16 post-oestrus. Reprod Fertil Devel 12:157–164. Weibel ER. 1979. Sterological methods. Practical methods for biological morphometry, Vol 1. New York: Academic Press. Wise T, Roberts AJ, Christensen RK. 1997. Relationships of light and heavy fetuses to uterine position, placental weight, gestational age, and fetal cholesterol concentrations. J Anim Sci 75:2197–2207. Zhang Z, Paria BC, Davis DL. 1991. Pig endometrial cells in primary culture: morphology, secretion of prostaglandins and protein, and effect of pregnancy. J Anim Sci 69:3005–3015.