| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Department of Veterinary Biomedical Sciences,3 WCVM, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 5B4
Department of Obstetrics, Gynecology and Reproductive Sciences,4 College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 0W8
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
corpus luteum, corpus luteum function, ovary, ovulation, progesterone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In sheep, as in other species, ovulation has been defined as the disappearance of a large, ovulatory-sized follicle (e.g., 5–7 mm in diameter in sheep) that had been detected and followed by ultrasonography from before estrus [9]. Ovulations have normally been confirmed by ultrasonographic detection of CL 3–5 days later [4], and the mean day of detection of CL in sheep averaged Day 3 after ovulation [4, 10]. However, ovulations followed by short-lived CL have been documented in otherwise normal ovine estrous cycles [4]. In addition, ovulations without CL formation or resulting in short-lived CL have been documented more frequently in different estrous synchronization/ovulation induction regimens in both cyclic and seasonally anestrous ewes [11, 12]. Therefore, detecting CL immediately after ovulation is important to confirm ovulation, to determine ovulation rate, and to allow correlation between the developmental pattern of the ovulated follicle and the luteal outcome.
Considerable histological information is available regarding development of the CL and its cellular composition in domestic species, including the sheep [13–16]. However, correlations among ultrasonographic, morphological, and functional attributes of developing CL in domestic animals are particularly scarce. One study [17] documented such correlations for CL at three discrete stages of the estrous cycle in heifers, namely Days 1–3, 5–9, and 15–19 after ovulation.
Ultrasonographic images are composed of picture elements (i.e., pixels) resulting from the refraction or transmission of high-frequency sound waves [1, 3, 17]. Each pixel represents the ability of a small, discrete unit of tissue to refract or transmit ultrasound waves, resulting in an image displayed in various shades of gray [17]. Collective numerical pixel values (i.e., brightness) and heterogeneity are functions of differences in tissue densities and macromolecular composition [1, 3, 17, 18]. It has been demonstrated that pixel analysis (i.e., echotextural characteristics) of images depicting follicles and CL in cattle and women reflect discrete changes in their morphology and secretory function [3]. To our knowledge, such studies do not exist for sheep, which unlike women and cattle ovulate one or more follicles at the end of each estrous cycle.
We hypothesized that the developing ovine CL could be detected using high-resolution ultrasound equipment as early as 12–24 h after ovulation and that the echotextural characteristics of the developing ovine CL are closely associated with its morphological (i.e., luteal tissue content) and functional (i.e., progesterone production) development. Thus, the objective of the present study was to use newer, high-resolution ultrasonographic equipment to determine the earliest time at which the developing ovine CL could be detected by transrectal ultrasonography postovulation. A second objective was to see if correlations could be found between the echotextural, histological, and functional characteristics of the developing ovine CL. The latter would facilitate rapid, noninvasive studies of terminal follicular development and luteogenesis, in relation to hormone secretion, in this animal model.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transrectal Ultrasonography and Blood Sampling
All ewes underwent transrectal ultrasonography twice a day from the time they were marked by a ram (estrus) until ovariectomy. Transrectal ultrasonography was done with high-resolution, real-time, B-mode ultrasonographic equipment (Aloka SSD-900; Aloka Co. Ltd., Tokyo, Japan) connected to a 7.5-MHz transducer (UST-5821; Aloka). During each scanning session, the settings of the scanner that affect image attributes (e.g., overall time-gain, near-field and far-field gains, compensation, and beam focus) were kept at predetermined levels. Images were displayed at 2x magnification. The number, diameter, and relative position of all follicles 1 mm or greater in diameter and of CL were sketched on ovarian charts, and all ovarian images were recorded on high-grade videotapes (Fuji S-VHS, ST-120 N, Fuji Photo Film Co., Ltd., Tokyo, Japan) using a super VHS-VCR (Panasonic AG-1978; Matsushita Electric, Mississauga, ON, Canada) equipped with digital frame memory. Ovulation was regarded as the disappearance of a large, ovulatory-sized follicle (e.g., 5–7 mm in diameter) [9, 19], which had been detected and followed by ultrasonography from just before estrus. Blood samples (10 ml) were collected by jugular venipuncture using vacutainers (Becton Dickinson, Rutherford, NJ, USA) before each ultrasonographic examination.
Collection of Ovaries, Water-Bath Ultrasonography, and Histology
Ovaries were removed surgically at 12–24 h (four ewes), 36–48 h (three ewes), and 60–72 h (three ewes) after ovulation detected by transrectal ultrasonography. All ovariectomies were performed within 3 h after the last ultrasonographic examination. General anesthesia was induced by i.v. injection of 2.5% thiopental sodium (Pentothal; 25 mg/kg; Abbott Laboratories Ltd., Mississauga, ON, Canada) and maintained by 3–5% halothane (Halothane; Halocarbon Laboratories, River Edge, NJ). Ovaries were exteriorized by midventral laparotomy. Ovaries containing luteal structures were excised, placed immediately in PBS (37°C, 0.1 mol phosphate buffer/L, 0.9% [w/v] sodium chloride, pH 7.2–7.4) and transported to the laboratory within 30 min of dissection. The ovaries were placed in a degassed water bath, and ultrasound images were collected using an Advanced Technology Laboratories ultramark HDI 5000 ultrasound machine (Advanced Technologies Laboratory, Bothell, WA) equipped with a broad-band (5–9 MHz), convex-array transducer designed for transvaginal scanning in humans.
Following water-bath ultrasonography, CL (12–24 h: n = 11; 36–48 h: n = 7; and 60–72 h: n = 6) were dissected and fixed in Haly solution for 24 h. Subsequently, fixed tissues were washed in running water for 6–8 h. The CL were then cut, approximately through the midline (maximum diameter), and placed in PBS until preparation of paraffin blocks. A set of two to three sections (thickness, 5 µm) was obtained for each CL and stained with hematoxylin and eosin for routine histology.
Computer-Assisted Analysis of Ultrasound Images
Ultrasound images of developing CL taken in the water bath, at half-distance from one edge of the CL (maximum diameter), with the HDI 5000 ultrasonographic equipment were stored directly as digitized graphic images. For transrectal images (Aloka SSD-900 equipment) recorded on videotapes, images were selected at half-distance from one edge of the CL. The selected images were digitized at standardized settings with a resolution of 640 x 480 pixels and 256 shades of gray and stored as graphic images. The digitization was done using a digital image-acquisition system (Picvision+; Imaging Technology, Inc., Woburn, MA). Echotextural analysis of images was done using a Sparc Station 10SX (Sun Microsystems, Mountain View, CA) and a custom-developed algorithm optimized for ultrasound images (Synergyne 1, Saskatoon, SK, Canada). Quantitative echotextural analysis was performed based on sequential measurements of numerical pixel values (gray scale values of individual picture elements ranging from 0 [absolute black] to 255 [absolute white]). Each image was divided into four quadrants, and mean numerical pixel values were measured within a computer-generated spot meter encompassing approximately 25% of each quadrant [3, 17, 18]. Special care was taken to prevent anechoic central cavities, which were seen in CL of 60–72 h, from falling under the spot meter. The central cavities were distinguished from follicles by the surrounding border of distinct luteal tissue [20]. The mean numerical pixel value (brightness) for each image was the mean of the numerical pixel values for the four quadrants. Pixel heterogeneity also was calculated; it was defined as the standard deviation of the mean numerical pixel values from the four quadrants of each image.
Luteal structures in digitized ultrasound images were outlined, and the area within the outline was calculated using the algorithm described above (Synergyne 1). In addition, maximal cross-sectional areas of follicles observed at the last transrectal ultrasound scan before ovulation were calculated using the follicle diameters determined with built-in calipers (Aloka SSD-900).
Computer-Assisted Analysis of Histological Images
One stained image per each CL was digitized at 1.25x magnification using imaging software designed for use in light microscopy (Northern Eclipse; Empix Imaging, Inc., Mississauga, ON, Canada). Luteal tissue and islands of blood clot (areas containing clumps of eosin-stained erythrocytes) were outlined on the images at 100–200% zoom-magnification, and the area falling within each outline was calculated. Areas of central cavities within CL containing serous transudate (eosin stained) but not blood clot were also determined. These areas were subtracted from the total area to obtain a corrected luteal area. The total areas of blood clot for each developing CL were expressed as a percentage of the total luteal area.
Hormone Assay
Blood samples were allowed to clot for 18–24 h at room temperature, and serum was harvested and stored at -20°C until assayed. Circulating concentrations of progesterone were determined by a validated radioimmunoassay [21]. The range of the standard curve was from 0.10 to 10 ng/ml. The assay sensitivity, or the minimum concentration of hormone significantly displacing labeled progesterone from the antibody, was 0.03 ng/ml. The intraassay coefficients of variation were 11.3% or 6.8% for reference sera with mean progesterone concentrations of 0.19 or 0.78 ng/ml, respectively.
Statistical Analyses
All statistical analyses were done using SigmaStat statistical software (Version 2.0 for Windows, 1997; SPSS, Inc., Chicago, IL). The effect of time after ovulation on various parameters was assessed by one-way ANOVA. The effect of time after ovulation, the source of ultrasound images, and their interactions were assessed by two-way repeated-measures ANOVA. Multiple comparisons were made by the method of Fisher least significant difference. Correlation coefficients among luteal areas, numerical pixel values, and serum progesterone concentrations were obtained by Pearson product-moment correlations. All results are expressed as the mean ± SEM.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Mean luteal areas obtained using both transrectal ultrasonography and ultrasonography in a water bath increased (P < 0.05) from 12–24 to 60–72 h after ovulation (Table 1). No difference was found (P > 0.05) between luteal areas calculated in images generated using either type of ultrasonographic equipment at 36–48 and 60–72 h postovulation; however, the mean luteal area was smaller in transrectal images as compared to images from the water bath at 12–24 h after ovulation (P < 0.05) (Table 1). A strong and positive correlation was found between luteal areas from the two sources of images (r = 0.89, P < 0.0001). Serum progesterone concentrations and luteal areas were also positively correlated (r = 0.88, P < 0.0001).
Descriptive Morphology of Developing CL
Inspection of ultrasonographic images of CL generated by water-bath scanning using the HDI 5000 equipment revealed that less echoic luteal tissue and more echoic blood clot could be differentiated even in the images at 12–24 h after ovulation (blood clot is toward the ovarian border within the marked area and is encapsulated by luteal tissue in Fig. 1A). In images at 12–24 h generated by transrectal ultrasonography in vivo, such differentiation was not evident; however, all CL could be seen as hyperechoic areas. By 60–72 h after ovulation, CL in the images from both the HDI 5000 and Aloka SSD-900 equipment exhibited a uniform echotextural pattern. Central cavities, if present, were anechoic, similar to follicular antra. Histologically, folding of the follicular wall after ovulation was observed in all the images of CL at 12–24 h after ovulation. This morphological feature was minimal in CL at 36–48 h and absent in those at 60–72 h following ovulation (Fig. 1). A central cavity filled with serous transudate was observed in 33% of the CL (2 of 6) collected at 60–72 h after ovulation.
|
Quantitative Echotextural Analysis of Ultrasound Images
The mean numerical pixel values of images from the water bath (Fig. 2) declined from 12–24 to 60–72 h (136.5 ± 4.2 vs. 91.5 ± 3.7, respectively; P < 0.05), but in transrectal images (Fig. 2), it only declined from 36–48 to 60–72 h (95.0 ± 7.1 vs. 70.1 ± 3.7, respectively; P < 0.05). Mean pixel heterogeneity (Fig. 2) did not differ between 12–24 and 36–48 h (P > 0.05) but declined by 60–72 h for images generated from both scanning in the water bath and transrectal ultrasound scanning (P < 0.05).
|
Quantitative Histomorphometric Analysis
The percentage of the CL occupied by blood clot declined from 12–24 to 60–72 h after ovulation (P < 0.05) (Fig. 3). The mean number of blood clot islets increased from 12–24 to 60–72 h postovulation (P < 0.05) (Fig. 3).
|
Serum Progesterone Concentrations
Mean serum progesterone concentrations increased (P < 0.05) from 12–24 to 60–72 h after ovulation detection (0.05 ± 0.02 and 0.5 ± 0.06 ng/ml, respectively) (Table 1).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The characteristic folding of the follicular wall after ovulation could be observed in all of the images of CL at 12–24 h postovulation, which is in agreement with the results of previous histological studies [22]. However, this morphological feature was reduced in CL at 36–48 h and absent in those at 60–72 h after ovulation (Fig. 1), suggesting the morphological remodeling of luteal tissue by approximately 72 h postovulation. The morphological remodeling of the CL resulting from migration of fibroblasts, endothelial cells, and theca interna cells was presumably facilitated by the folding of the follicular wall [23].
Two quantitative echotextural variables of ultrasound images were utilized in the present study to characterize the formation of CL: the numerical pixel value (brightness) and the pixel heterogeneity [1–3, 17, 18]. The numerical pixel value and heterogeneity are both dependent on differences in tissue densities and macromolecular composition [1, 3, 17, 18]. Numerical pixel values of CL in the present study reached a minimum at 60–72 h after ovulation (Fig. 2), in contrast to approximately 2 days after ovulation in heifers [1]. In mares, high numerical pixel values of CL on Day 1 were attributed to an increase in the number of reflective surfaces created by the apposing folds of the collapsed follicular wall [24]. In a study of mares, only 50% of ovulations resulted in transient corpora hemorrhagica [25], and in heifers, the proportion was even lower (2 of 23) [20]. Therefore, hyperechoic, fibrin-like strands were not a common finding in luteal cavities in these species. In an ultrasonographic study of atherosclerotic carotid plaques in humans [26], fibrin and fibrous tissue had the second-highest gray scale (pixel) intensity among blood, lipid, fibrous tissue, and calcium. In the ewes of the present study, all the ovulations resulted in the formation of a blood clot in the former antrum of the collapsed follicle (Fig. 1). Thus, it is logical to assume that in addition to the increased reflective surfaces of the collapsed follicular wall, the presence of a blood clot influenced the echotexture of the developing CL in sheep.
The greater resolution of the HDI 5000 ultrasound equipment probably accounted for subtle differences in the pattern of change in numerical pixel values of CL images from 12–24 to 60–72 h after ovulation as monitored by the two types of ultrasound scanners used in the present study (Fig. 2). The borders of luteal tissue could be precisely demarcated from the ovarian stroma even at 12–24 h after ovulation using the higher-resolution ultrasound equipment and water-bath scanning. This precision in outlining the luteal tissue may have contributed to the higher mean luteal areas recorded from images generated in the water bath compared to transrectal images (Table 1). However, the slightly reduced precision of outlining the borders of CL in images generated by transrectal ultrasonography was not great enough to undermine our ability to determine the luteal areas, because they were still comparable to the images generated by water-bath scanning and showed a similar trend with time (Table 1). In addition, the luteal areas in images from both sources correlated positively with serum progesterone concentrations. Finally, luteal areas doubled from 12–24 h to 60–72 h after ovulation (Table 1), which is in agreement with the results of earlier histological studies in cyclic ewes [27, 28].
Pixel heterogeneity of CL did not change significantly until 36–48 h after ovulation (Fig. 2). Pixel heterogeneity probably reflected the existence of interspersed hyper- and hypoechoic areas in a heterogeneous luteal tissue composition because of the presence of echoic luteal cells of various types and blood clot as well as extruded, nonechoic serum in the clot [29]. The significant decline in pixel heterogeneity of CL by 60–72 h postovulation could be due to the fact that luteal tissue was predominant by approximately 72 h and islets of blood clots were fewer in number and more diffuse (Figs. 1 and 3), resulting in more homogeneous echogenic patterns. In addition, luteal cell migration and expansion also may have reduced the number of hyperechoic folds of follicular wall [24] that were predominant at 12–24 h after ovulation (Fig. 1) [30].
A central luteal cavity containing serous transudate but not blood clot was observed in 33% of the CL at 60–72 h during the present study. Such central cavities have been observed in several domestic species, including sheep [4], cattle [17, 31], and horse [32]. All CL of Day 1 had a blood clot (Fig. 1) in the former follicular antrum. A few such clots resulted in central cavities filled with anechoic serous transudate. Such fluid-filled cavities in a few CL probably failed to be breached and filled in by proliferating luteal cells.
The importance of our ability to identify CL during luteogenesis lies in the fact that it provides a noninvasive technique to look at relationships between various characteristics of ovulatory follicles (growth rate, time from emergence to ovulation, and endocrine milieu at the time of emergence and during the growth phase) and the formation and ensuing function of CL. Luteal inadequacy can lead to death of the embryo or the fetus [28], and it has been identified as a major cause of recurrent miscarriages in women [33]. Various luteal inadequacies in the ewe, such as short-lived CL and ovulations that do not result in the formation of functional CL, have been documented in our laboratory [4, 10–12]. In addition, it is unclear whether reduced luteal vascularity is secondary to luteal insufficiency or whether inadequate luteal vascularization is a primary cause of luteal dysfunction [34]. Differences in echotextural characteristics between the two ovulatory-sized follicles of the same follicle wave but with different fates (i.e., ovulatory vs. anovulatory follicle) have been shown in women [3]; echotextural characteristics of growing ovarian antral follicles have been documented in heifers [35, 36] but not in sheep. Such studies, along with the present ultrasonographic characterization of CL development, hold promise to increase our understanding of the normal luteogenesis and etiology of CL dysfunction.
In ultrasonographic imaging-based studies of the ovaries in domestic animals, ovulation traditionally has been defined as the disappearance of ovulatory-sized follicles [9, 19]. However, with the present demonstration of early CL detection in the ewe, ovulation can now be defined in this species as the disappearance of ovulatory-sized follicles followed by the detection of forming CL. In addition, recent studies have documented the incidences of several types of abnormal follicular development, failure of ovulation, and inadequate luteal function in women [5]. Apart from laboratory mammals, the sheep has evolved as the principal animal model [37] for studying the physiology of the human reproductive cycle. Thus, the present demonstration of CL visualization in sheep creates a model for studying luteal function in women.
In conclusion, the results of the present study show that it is possible to detect growing CL in the ewe by transrectal ultrasonography as early as 12–24 h after ovulation, thus permitting future studies of the follicular growth-CL formation interphase in this animal model species. Echotextural attributes closely correlated with morphological and functional changes in the growing CL. Ultrasonography and computer-assisted image analyses may greatly aid in studies of luteogenesis and the etiology of CL dysfunction in the ewe.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: N.C. Rawlings, Department of Veterinary Biomedical Sciences, WCVM, University of Saskatchewan, 52 Campus Drive, Saskatoon, SK, Canada S7N 5B4. FAX: 306 966 7376; norman.rawlings{at}usask.ca
Received: 25 February 2003.
First decision: 1 April 2003.
Accepted: 17 April 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
given at different stages of the luteal phase in ewes. Theriogenology 2002 58:1409-1424[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  R. Duggavathi, P.M. Bartlewski, D.M.W. Barrett, E. Bagu, and N.C. Rawlings Short- and Long-Term Effects of Unilateral Ovariectomy in Sheep: Causative Mechanisms Biol Reprod, March 1, 2008; 78(3): 490 - 496. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M.W. Barrett, R. Duggavathi, K. L. Davies, P. M. Bartlewski, E. T. Bagu, and N. C. Rawlings Differential Effects of Various Estradiol-17beta Treatments on Follicle-Stimulating Hormone Peaks, Luteinizing Hormone Pulses, Basal Gonadotropin Concentrations, and Antral Follicle and Luteal Development in Cyclic Ewes Biol Reprod, August 1, 2007; 77(2): 252 - 262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M.W. Barrett, P. M. Bartlewski, R. Duggavathi, K. L. Davies, and N. C. Rawlings Suppression of Follicle Wave Emergence in Cyclic Ewes by Supraphysiologic Concentrations of Estradiol-17beta and Induction with a Physiologic Dose of Exogenous Ovine Follicle-Stimulating Hormone Biol Reprod, October 1, 2006; 75(4): 633 - 641. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
