| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GROWTH, PHYSIOLOGY, AND REPRODUCTION |
Department of Animal Sciences, University of Illinois, Urbana 61801
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: cervix • estradiol • genistein • phytoestrogen • swine • uterus
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Phytoestrogens generally appear to mimic the action of estradiol (Kaldas and Hughes, 1989; Adams, 1995).
The reproductive tract is particularly responsive to estrogen. There are cyclic periods of uterine and cervical tissue growth that occur throughout the estrous cycle. For example, in ovariectomized ewes given silastic implants with estradiol, uterine wet weight increases by more than 2-fold in the first 24 h due to increases in hyperplasia and hypertrophy (Reynolds et al., 1998). Although considered a weak estrogen, genistein administration to ovariectomized rats results in significant increases in height of the lumen epithelial cells of uterine and vaginal tissues (Diel et al., 2001). Other studies also indicate significant changes of reproductive tract tissues in response to phytoestrogen treatment in ovariectomized rats (Perel and Lindner, 1970; Santell et al., 1997).
Modern swine diets contain soybeans and therefore contain phytoestrogens. Drane et al. (1981) observed that prepubertal gilts fed a diet of 20% soybean meal experience a greater increase in vulva size than gilts fed a nonsoybean meal diet. However, there are limited data available on the in vivo effects of soybean phytoestrogens on reproductive tissues of pigs. This study was designed to address the fundamental hypothesis that genistein has estrogen-like effects on estrogen sensitive tissues in ovariectomized postpubertal gilts. The objective was to evaluate the biological responses of uterine and cervical tissues to the administration of genistein.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). All other nutrients also met the requirements suggested by the NRC (1998).
Daily intake of the diet resulted in consumption of approximately 230 g of soybeans. Using the values reported by Wang and Murphy (1994), we estimated that consumption of this diet would result in gilts consuming an estimated 725 mg of total dietary phytoestrogens daily, of which approximately 7 mg would be genistein and 137 mg would be genistin.
Puberty was induced in all gilts used in this trial by administration of a subcutaneous injection of PG600 (400 IU of pregnant mare serum gonadotrophin and 200 IU of human chorionic gonadotrophin; Intervet, Millsboro, DE) in the neck region. Seventy-two hours later, the gilts received an injection of 1,000 IU of human chorionic gonadotrophin. Twenty-four to 36 h after estrus was induced, the gilts were ovariectomized to remove the principal source of endogenous estrogen. Ovariectomy was performed following the approved protocol by trained surgeons.
Gilts were assigned randomly to 1 of 6 treatment groups 15 d postovariectomy. Treatment groups received either vehicle (2 mL of corn oil) with no hormone (negative control, NC; n = 8), 2 mg of estradiol benzoate/d (positive control, PC; n = 5), or genistein at 50 mg/d (G50; n = 4), 100 mg/d (G100; n = 5), 200 mg/d (G200; n = 5), or 400 mg/d (G400; n = 7).
The hormone replacement therapy used for these postpubertal ovariectomized gilts was patterned after an established model of induction of reproductive tract development in ovariectomized gilts (Winn et al., 1994). Dosages of genistein to be administered were estimated to be near the dietary intake levels. In addition, a preliminary study with limited numbers of gilts was conducted to estimate the range of response to genistein administration. Genistein administered at 12.5 and 25 mg/d did not produce a measurable effect on the reproductive tract in ovariectomized gilts. The final genistein dosages chosen for this study were based on the results of that preliminary study. Genistein was purchased from LC Laboratories (Boston, MA) and estradiol benzoate from Sigma (St. Louis, MO); costs of genistein precluded administration of dosages greater than the 400 mg/d.
The gilts received treatment injections containing one-half of the daily dosage at 12-h intervals for 10 consecutive days beginning on d 15 after ovariectomy. On the day of the last injection, the animals were transported to the University of Illinois Meat Science Laboratory abattoir and provided access to water. The following morning (d 25), the gilts were electrically stunned and euthanized by exsanguination.
Tissue Collection and Chemical Analysis
Uteri and cervices were obtained from the gilts at slaughter. Extraneous fat and connective tissue were trimmed, the uterus and cervix were weighed (wet tissue weight), and tissue blocks (4 to 6 mm3) were collected for histology (described in the next section). The remaining tissue was frozen at –20°C for analysis of tissue hydration, protein content, and DNA content. Tissue hydration was measured by freeze-drying; the results were used to calculate tissue dry weight. Freeze-dried tissue was ground, and subsamples of the tissue were used to determine tissue crude protein content by the Kjeldahl method (AOAC, 1995) and total tissue DNA content by a modification of the Labarca and Paigen method (1980), as described previously (Ford et al., 2003).
Histological and Immunohistochemical Analysis
For histological analyses, uterine samples were collected from the mid region (lengthwise) of the right uterine horn, and cervical samples were collected from both the uterine and vaginal regions of the cervix (Eldridge-White et al., 1989). Samples from 5 gilts each from the PC, NC, and G400 groups were analyzed histologically. Samples were initially fixed in 4% paraformaldehyde for 6 to 8 h at room temperature, after which the paraformaldehyde was replaced with fresh paraformaldehyde, and the samples were left to fix overnight. After overnight fixation, tissues were transferred to 70% ethanol. Tissue samples were dehydrated using a Tissue-Tek Vacuum Infiltration Processor Model E150 (Sakura Finetek USA., Inc., Torrance, CA), paraffin embedded, and 5-µm thick sections were dried overnight on a slide warmer (40°C). Tissues were de-paraffinized in xylene, and sections were rehydrated in a series of descending ethanol baths and water. Cervical and uterine samples were then stained with hematoxylin and eosin for morphometric analysis.
Immunohistochemisty was performed to evaluate the differences in tissue localization of progesterone receptor (PR) and proliferating cell nuclear antigen (PCNA). Both PR (Spencer and Bazer, 2002) and PCNA (Klotz et al., 2000) have been used as markers of estrogenic activity in reproductive tissues. Similar procedures were followed for both PR and PCNA staining. Paraffin was removed, and the sections were rehydrated as detailed previously. Target proteins were unmasked by boiling in a microwave oven for 10 min in a 0.01 M citrate buffer. Incubation of the primary antibody was performed overnight (4°C) for both PR and PCNA. Primary antibody for PR detection was a rabbit polyclonal antiPR antibody (Santa Cruz Biotechnology, Santa Cruz, CA) that was diluted 1:200 in BSA/PBS. Primary antibody for PCNA detection was a mouse monoclonal anti-rat PCNA antibody (DAKO Corporation, Carpinteria, CA) diluted 1:250 in BSA/PBS. Secondary antibody for the PR staining was a biotinylated goat anti-rabbit immunoglobulin (Vector Laboratories, Burlingame, CA) diluted 1:100 in BSA/PBS, and for PCNA staining was a biotinylated goat anti-mouse immunoglobulin (DAKO Corporation) diluted 1:100 in BSA/PBS. The avidin-biotinylated peroxidase complex (Vec-tastain ABC Elite Kit; Vector Laboratories, Burlingame, CA) and diaminobenzidine (DAB; Sigma) were used to visualize the antibody binding. Counterstaining was with Mayer’s hematoxylin (Electron Microscopy Sciences, Fort Washington, PA).
Stained sections were visualized and images captured using an Olympus model BX51 microscope (Olympus, Melville, NY). Electronic images were obtained with a digital camera (ProgRes C14; Jenoptik L.O.S. GmbH, Germany) with the aid of Camera Filmware 1.3 software (version 1.5.0; Jenoptik L.O.S. GmbH). Pictures were compiled using Adobe Photoshop software version 7.0 (Adobe Systems, San Jose, CA). Image J (version 1.29x; National Institutes of Health, Bethesda, MD) was used to measure the epithelial heights and uterine gland widths. For epithelial layer heights or uterine gland widths, 4 distinct fields were analyzed with 5 measurements taken per field for a total of 20 measurements. The average of all 20 measurements was used as the epithelial height or uterine gland width for that sample. Quantification of the immunohistochemistry results was done by counting the number of cells stained positive or negative (up to 100 cells) in each of 4 different fields per sample. The average percentage of cells stained positive was then used for each sample.
Statistical Analysis
Data from the reproductive tissue of the NC group were used to establish baseline tissue characteristics. Mean responses were modeled to investigate the effect of treatment level on the uterine and cervical tissue. Statistical analyses of the data were performed using the Regression and GLM procedures of SAS (SAS Inst., Inc., Cary, NC). Data were analyzed by treatment group (control and treated), and across treatment. The model was
 |
in which yijkl was the response variable (e.g., tissue wet weight); µ was the overall population mean; ai was the effect of the i treatment group i, and ei was the error term. The assumptions were that the treatment effects were independent and normally distributed with equal variance, the residuals were independent and normally distributed with equal variance, and both random effects were independent. The differences among treatment groups were evaluated based on the P-values, least squares means, and SE. Significance for histological measurements was obtained using the t-test procedure of SAS. The groups compared were PC vs. NC, PC vs. G400, and NC vs. G400.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Uterine development after 10 d of genistein treatment was indicated by the increases in the wet weight and total DNA mass of the uterus (Table 1). Uterine wet weight was increased above NC gilts in the G200 gilts but not in the G50 or G100 gilts. Uterine wet weight of the G400-treated gilts were more than 250% greater than that of NC gilts. Genistein treatment caused significant changes in uterine wet weight (genistein dosage regression, P < 0.001; treatment effect, P < 0.001). Total uterine DNA also was increased (genistein dosage regression, P < 0.001; treatment effect, P < 0.001). The total uterine DNA in G400 gilts was almost twice that of the NC gilts after 10 d of genistein treatment. Uterine dry weight and protein weight increased with genistein treatment (genistein dosage regression, P < 0.001; treatment effect P < 0.001). The G400 gilts had a uterine dry weight and protein weight approximately 2 and one-half times that of the NC gilts. The significant changes in uterine growth induced by genistein treatment were less than the response seen in the PC gilts in which the estradiol benzoate treatment resulted in a nearly 7-fold increase in uterine wet weight (P < 0.001; Table 1).
|
|
). Cervical wet weight of gilts treated with genistein (G400) increased almost 2 and one-half times compared with the NC gilts (genistein dosage regression, P < 0.001; treatment effect P < 0.001). Total cervical DNA in G400 gilts was more than twice that of the NC gilts. Additionally, there was a dosage-dependent increase in total DNA content with genistein treatment (genistein dosage regression, P < 0.001; treatment effect, P < 0.001). Cervical dry weight and protein weight increased in a dosage-dependent manner (genistein dosage regression, P < 0.001; treatment effect P < 0.001). Cervical dry weight in the G400 gilts was more than double that of NC gilts. Genistein-treated gilts had lower cervical wet weights (P < 0.001), dry weights (P < 0.001), and protein weights (P < 0.001) than found in PC gilts.
Tissue Composition
In contrast to the changes in the mass of uterine tissue components, there were no effects of genistein dosage or estrogen treatment on uterine percentage of hydration (overall mean ± SE = 83.6 ± 2.4%; P = 0.46), percentage of protein (81.9 ± 3.1%, dry basis; P = 0.87), concentration of DNA (11.8 ± 4.8 µg/mg; P = 0.35), cervical percentage of hydration (80.0 ± 2.7%; P = 0.54), or cervical concentration of DNA (6.6 ± 2.1 µg/mg; P = 0.84). However, cervical protein percentage decreased as genistein concentration increased (genistein dosage regression, P < 0.01); the cervical protein percentages of the G200 (86.0 ± 0.7; P < 0.01) and G400 gilts (85.9 ± 0.6; P < 0.01) were less than the NC gilts (88.4 ± 0.5). Cervical protein percentage for PC gilts was less (85.5 ± 0.7; P < 0.01) than for NC gilts but not different from the G50 (P = 0.08), G100 (P = 0.13), G200 (P = 0.62), or G400 (P = 0.65) gilts.
Histomorphometry
Comparisons for histomorphology were made among the NC, PC, and G400 treatment groups. Epithelial cells lining the uterine lumen doubled in height in G400-treated gilts (P < 0.001) and more than tripled in height in the PC gilts (P < 0.001) compared with the NC gilts (Table 2; Figure 2a,b,c). Uterine epithelial cells in the NC gilts were cuboidal (Figure 2b), whereas in the G400 gilts epithelial cells had increased in height and were more columnar (Figure 2c). The PC gilts had uterine epithelial cells that were columnar and had grown to a greater extent than cells of the NC and G400 gilts (Figure 2a). The increase in epithelial layer height in G400-treated gilts and PC gilts compared with NC gilts also was true of the epithelial lining of uterine glands (P < 0.001; Table 2; Figure 2d,e,f). The PC gilts had glandular epithelial cell height that was greater (P < 0.01) than in G400-treated gilts (Table 2). Uterine glands in NC gilts (Figure 2e) were lined by low columnar epithelium cells and had limited apparent luminal volume. The width of the uterine gland lumen in the G400 gilts was almost 3-fold that of the NC gilts (P < 0.01), whereas in the PC gilts the uterine gland lumen width increased almost 5-fold (P < 0.001; Table 2; Figure 2d,e,f). Additionally, the width of the uterine gland lumen in the PC gilts had expanded compared with the G400 gilts (P < 0.05).
|
|
). However, the height of the epithelial layer lining the lumen in the uterine cervix was greater in both the PC- (P < 0.01) and the G400-treated (P < 0.01) gilts than the NC gilts (Table 2; Figure 2g,h,i). The epithelial layer lining the uterine cervix in NC gilts was small, and cells appeared cuboidal in shape (Figure 2h). The epithelial lining of the vaginal cervix increased in height almost 4-fold for G400-treated (P < 0.001) and almost 8-fold with PC-treated gilts (P < 0.001) compared with the NC gilts (Table 2; Figure 2j,k,l). In addition, the height of the epithelial lining of the vaginal cervix was greater for the PC gilts compared with the G400 gilts (P < 0.001). The epithelial lining of the vaginal cervix in NC gilts appeared as a thin layering (1 or 2 cells thick) of stratified epithelial cells (Figure 2k), whereas treatment with either G400 (Figure 2l) or PC (Figure 2j) resulted in a thicker layer of stratified epithelial cells.
Immunohistochemistry
Immunohistochemical staining was performed on the uterine and cervical tissues from the NC, PC, and G400 treatment groups as an indicator of cell division (presence of PCNA) and estrogen responsiveness (induction of PR). The percentage of cells stained positive for PCNA in the uterine lumen was approximately doubled in response to either G400 (P < 0.001) or PC (P < 0.01) treatment compared with NC (Table 3). No differences among the 3 groups were found in the numbers of PCNA-positive epithelial cells lining the uterine cervix (Table 3; PC vs. NC, P = 0.37; PC vs. G400, P = 0.69; G440 vs. NC, P = 0.65). However, epithelial cells lining the vaginal cervix had greater percentages of cells stained positive for PCNA in the G400 (P < 0.05) and PC (P < 0.01) groups compared with the NC group. Epithelial cells staining for PCNA in the vaginal cervix primarily were in the basal layers of the stratified epithelium, although PCNA staining also was observed in stromal cells.
|
). The percentage of cells stained positive for PR in the uterine lumen was not different among the treatment groups (Table 3). However, the PC and the G400 treatments both had a greater percentage of cells stained positive for PR than the NC treatment in the uterine glands (P < 0.001 and P < 0.001, respectively; Table 3). In the uterine cervix the PC treatment resulted in a greater percentage of cells stained positive for PR than the G400 and NC groups (P < 0.05; Table 3), but there were no differences between the G400 and NC groups (P = 0.50). A greater percentage of epithelial cells lining the vaginal cervix stained for PR in the PC group (P < 0.05) or G400 (P < 0.05) compared with the NC group (Table 3), but there were no differences between the G400 and PC groups (P = 0.70). |
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Santell et al., 1997; Diel et al., 2001).
The uterus of the ovariectomized gilt is highly sensitive to estrogenic stimulation (Hall et al., 1992). In that study, ovariectomized prepubertal gilts receiving twice daily injections of 0.2 mg of estradiol benzoate for 16 d respond with a 5-fold increase in uterine wet weight. Injection of estradiol benzoate at 1 mg twice daily for 10 d in the present experiment results in a 7-fold increase in uterine wet weight over untreated control gilts. The porcine cervix also is sensitive to estrogen stimulation, with a 4-fold increase in wet weight of the uterine cervix in ovariectomized gilts induced by 2 mg of estradiol benzoate/d for 15 d (Winn et al., 1994), similar to the fold increase in the total cervical weight observed in the current study. By comparison, the increase in wet weight of both the uterus and cervix to the largest dosage of genistein (400 mg/d) is about 2 and one-half times greater than that for untreated control gilts..
The uterotrophic effect of genistein treatment observed in this study is consistent with other studies that have shown increased uterine growth in response to treatment with phytoestrogens in ovariectomized and immature rats (Whitten et al., 1992; Santell et al., 1997) and ovariectomized sheep (Nwannenna et al., 1995) and cattle (Kallela, 1968). Uterotrophic responses to orally administered genistein (Santell et al., 1997; Degen et al., 2002; Diel et al., 2004b) or injected genistein (Diel et al., 2004a) have been observed in some ovariectomized rat models but not in some intact rat models (Nakai et al., 2005a,b). In addition, endometrial transplants established at extrauterine sites can be maintained by genistein or estrogen injections after ovariectomy of the recipient rats (Cotroneo and Lamartiniere, 2001). Differences in uterotrophic responses among these reports probably arise from the mode of administration, source of genistein, duration of exposure, and endocrine status of the animals.
Morphological changes of the epithelial lining of reproductive tract tissues provide another indicator of estrogenic action. Genistein increases the epithelial cell height in the uterine lumen and uterine glands, as well as the width of uterine glands. These responses are similar to responses observed in ovariectomized rats receiving genistein orally for 3 d (Diel et al., 2001; 2004a,b). Epithelial height also is increased in response to genistein in both the uterine cervix and vaginal cervix of gilts in the current study. In most cases the response to genistein (400 mg/d) treatment is more limited than that to the estrogen treatment. Interestingly, the degree of change in epithelial layer height appeared to be variable among the tissues. Epithelial layer height response to genistein treatment in the uterine cervix is comparable with that of the estrogen treatment but is more limited relative to the estrogen treatment in the other tissues. In ovariectomized rats, vaginal tissue is more sensitive than the uterine tissue to treatment with genistein or estradiol (Diel et al., 2001).
Proliferating nuclear cell antigen is a marker of cell proliferation (Connolly and Bogdanffy, 1993). Proliferating nuclear cell antigen is an auxiliary protein of DNA polymerase delta, and synthesis of PCNA peaks in the S phase of the cell cycle after beginning late in the G1 phase of the cell cycle (Greenwell et al., 1991). In the current study, uterine and cervical tissue sections from NC gilts have few cells that stain positive for PCNA. In most cases, percentages of epithelial cells staining positive for PCNA are increased in uterine and cervical tissues from genistein- and estrogen-treated gilts. The actions of genistein in promoting cellular growth indicate that genistein is having a mitogenic effect in these tissues. However, staining for PCNA does not indicate that a hormonally induced cell proliferation is occurring in the uterine cervical epithelial cells in these ovariectomized gilts.
Progesterone receptor expression in uterine epithelial cells is responsive to estrogen (Spencer and Bazer, 2002). Dietary genistein inhibits development of induced prostate cancer in part by a downregulation of the progesterone receptor mRNA expression (Lamartiniere et al., 2002). Dietary genistein treatment increases uterine PR expression in some rodent studies (McMicheal-Phillips et al., 1998; Yu et al., 2002) but not others (Hargreaves et al., 1999; Yang et al., 2000; Cotroneo and Lamartiniere, 2001). Genistein also increases PR in estrogen-responsive uterine cells when injected in ovariectomized rats (Cotroneo and Lamartiniere, 2001). In the current study of ovariectomized gilts, typically more than 50% of the epithelial cells in each tissue stain positive for PR. Treatment with genistein or estrogen increases the proportion of cells staining positive for PR in each tissue, except for the uterine luminal epithelial cells.
Genistein binds to the estrogen receptor-
at 4% of the affinity of 17ß-estradiol, and to the estrogen receptor-ß at 87% of the affinity of 17ß-estradiol (Kuiper et al., 1998). Within the reproductive tract, estrogen receptor- is found to be present in much greater quantities than estrogen receptor-ß (Couse et al., 1997). The differential magnitude of response to genistein of uterine and cervical tissues noted previously may result from different ratios of the estrogen receptors within the tissues. In the absence of endogenous estrogen, as in the ovariectomized gilts, the action of genistein may occur via the estrogen receptor-ß. Alternatively, genistein may be acting through other receptors and pathways, such as the tyrosine kinase pathway, the mitogen-activated protein kinase pathway, or the epidermal growth factor receptor pathway (reviewed in Rosselli et al., 2000).
The current study demonstrates that the uterus and cervix of ovariectomized gilts respond to phytoestrogen treatment in a manner qualitatively similar to estrogen treatment. This is the first demonstration of such effects of genistein on uterine and cervical tissues in this species. This study establishes a framework for further research on the effects of dietary genistein on reproductive function in intact pigs. It is important to note that no abnormalities in the reproductive tissues are observed in response to the 10-d injection protocol used in this study. Furthermore, although this study does not address the effects of dietary phytoestrogen, the reproductive tissues from these ovariectomized gilts appeared relatively quiescent, suggesting minimal impact of the dietary phytoestrogens on those tissues. This latter observation corroborates rodent studies where ovariectomized rats fed chow known to contain phytoestrogen have no signs of estrogen response in the reproductive tract and are able to respond normally to estrogenic substances in uterotrophic assays (Degen et al., 2002). No effect of dietary soybean meal has been observed on reproductive tract weights or histological abnormalities in the intact gilts, although effects on the vulva have been noted by Drane et al. (1981). Nevertheless, considering the high consumption of soybean-based diets and, therefore, high phytoestrogen intake in swine, more careful study of the effects of these potential endocrine disrupting agents is needed. Soybean phytoestrogens also may have positive effects on the health and growth of swine. Daidzein, another soybean phytoestrogen fed to piglets enhances the ability of the piglets to respond to a viral challenge (Greiner et al., 2001). The sensitivity of the uterus of the gilt to estrogenic substances may make that animal a potential model to examine the impact of environmental endocrine modulators on reproductive tissues (Magnusson, 2005).
|
Footnotes |
|---|
2 Corresponding author: wlhurley{at}uiuc.edu
Received for publication September 5, 2005. Accepted for publication December 1, 2005.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
(ER) and estrogen receptor-ß (ERß) messenger ribonucleic acid in the wild-type and ER-knockout mouse. Endocrinology 138:4613–4621.. Endocrinology 141:3430–3439.This article has been cited by other articles:
|
|
 |
|
  C. F. Kwok, C.-C. Juan, and L.-T. Ho Endothelin-1 decreases CD36 protein expression in vascular smooth muscle cells Am J Physiol Endocrinol Metab, February 1, 2007; 292(2): E648 - E652. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
