| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Department of Neurobiology and Physiology,3 Northwestern University, Evanston, Illinois 60208,
Department of Medicine,4 Feinberg School of Medicine, Northwestern University, Chicago, Illinois 60611,
Robert H. Lurie Comprehensive Cancer Center of Northwestern University,5 Chicago, Illinois 60611
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunit protein is present in all follicle stages but increases in intensity within the mural granulosa cells in large antral follicles. The inhibin ßA and ßB subunit proteins, in addition to the activin type I (ActRIB) and activin type II receptor (ActRIIB), are produced in primordial and primary follicle granulosa cells. Additionally, inhibin ßA subunit is detected in the theca cells from secondary through large antral follicle size classes. GDF-9 is restricted to the oocyte of preantral and antral follicles, whereas the type II BMP receptor (BMP-RII) protein is predominantly localized to primordial- and primary-stage follicles. TGFß1, 2, and 3 ligand immunoreactivity is observed in both small and large follicles, whereas the TGFß type II receptor (TGFß RII) is detected in the oocyte and granulosa cells of antral follicles. The intracellular signaling proteins Smad2 and Smad4 are present in the granulosa cell cytoplasm of all follicle size classes. Smad3 is detected in the granulosa cell nucleus, the oocyte, and the theca cell nucleus of all follicle size classes. These data suggest that the complete activin signal transduction pathway is present in small follicles and that large follicles primarily produce the inhibins. Our data also suggest that TGFß ligands and receptors are colocalized to large antral follicles. Taken together, the ligands, receptors, and signaling proteins for the TGFß superfamily are present at distinct points throughout feline folliculogenesis, suggesting discrete roles for each of these ligands during follicle maturation.
activin, follicle, granulosa cells, inhibin, ovary
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cat family (Felidae) is composed of 37 species in its entirety. With the exception of the domestic cat, all are endangered or threatened by extinction. As with many other endangered animals, the decline in feline populations is primarily due to an accelerated rate of habitat destruction and poaching. Preservation of endangered species is becoming increasingly dependent on captive breeding programs in zoos and in wildlife parks where strategies are employed to manage captive and natural populations in order to sustain genetic lineage and biodiversity. However, while some captive felids breed easily in zoo collections, others often reproduce poorly. Usually the outcome is complicated by a complete lack of physiological information for the species [5]. Conversely, domestic cat ovarian biology can be easily studied due to tissue accessibility and better knowledge regarding its physiology as opposed to its wild counterparts.
While folliculogenesis in the feline ovary has not been well studied, considerably more is known about the feline estrous cycle. The onset of puberty in the domestic cat is observed by 6 to 9 mo of age [6], but this is dependent on breed and photoperiod. Cats are long-day, seasonally polyestrous breeders, which require 12 h of light to maintain normal cycles. If not bred, cats will cycle into heat on an average of every 2 to 3 wk, with estrus lasting about 1 wk [6]. To better understand the control of the feline reproductive axis and investigate paracrine regulation of folliculogenesis, transforming growth factor ß (TGFß)-ligands, receptors, and signaling molecules were localized in cat ovaries collected at various times during the cycle.
The role of the TGFß superfamily members in folliculogenesis has not been investigated in the ovary of the domestic cat. The TGFß superfamily is a large and diverse group of dimeric proteins involved in cell cycle control and differentiation. TGFß as well as activin and GDF-9 act on the ovary in a paracrine manner while inhibin is thought to regulate FSH in an endocrine feedback loop. These ligands have been shown to control steriodogenesis [7], proliferation [8, 9], and stage-dependent apoptosis [10] as well as regulating each other [11]. Thus, it is critical to evaluate these powerful ligands in cat follicle development. Knowledge of an intact TGFß superfamily signal transduction system in the cat ovary may make an important impact on IVM/IVF success for endangered feline species.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ovaries were obtained following routine animal spaying at local veterinary clinics. Nine cats were considered sexually immature at the time of spaying, while three were considered to have just reached puberty. A total of 24 ovaries were analyzed throughout this study. All animal tissue was collected in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals. Tissue was used for RNA isolation or processed for sectioning. The ovaries used for sectioning were rinsed in 1x PBS (Dulbecco phosphate-buffered saline; Gibco, Invitrogen Corp., Carlsbad, CA) then placed into 4% paraformaldehyde (Sigma, St. Louis, MO) fixative at 4°C overnight. The tissue was then dehydrated and paraffin embedded by the Pathology Core Laboratory (Northwestern University, Chicago, IL). Four micrometer microtome sections were obtained and mounted on Superfrost-Plus slides (Vector Laboratories, Inc., Burlington, CA).
Cloning of Domestic Cat Inhibin , ßA, and ßB subunits: cDNA Reverse Transcription (RT) Polymerase Chain Reaction (PCR)
Ovarian total RNA was extracted from five ovaries and 2–4 µg was reverse transcribed into cDNA using Maloney murine leukemia virus reverse transcriptase in the presence of 20 pmol random hexamer oligonucleotides and 10 mM deoxynucleotide triphosphates (Promega Corp., Madison, WI). PCR was performed on the cDNA for 35 cycles with annealing temperatures of 55–58°C using standard techniques. Inhibin , ßA, and ßB subunit-specific primers were initially generated from a universal inhibin subunit sequence and from human ßA, and ßB subunit sequences. All primers used to amplify the subunits from cat ovarian RNA are listed in Table 1. PCR products were cloned into the TA 2.1 cloning vector (Invitrogen) and sequenced using BigDye Terminator Cycle Sequencing (ABI Prism, Foster City, CA). Sequence results were compared with multiple species subunit cDNA, and homology to other species was analyzed. Sequences were entered as GenBank accession numbers AY258627, AY258628, and AY258629.
|
RNA Isolation and Northern Blot Analysis
Total RNA was extracted from ovarian tissue using TRIZOL reagent (Life Technologies Inc., Rockville, MD). Concentration of the isolated RNA was determined by absorption at 260 nm on a spectrophotometer and the integrity of the preparation examined by agarose gel electrophoresis. For RNA blot analysis, 20 µg total RNA was electrophoresed per lane on a 1% agarose, 1·1 M formaldehyde and 1 x (3-[N-Morpholino] propane sulfonic acid) gel using standard procedures. RNA was transferred overnight to BrightStar-Plus positively charged nylon membranes (Ambion, Austin, TX) by capillary action with 20x saline-sodium citrate. Biotinylated probes were synthesized using BrightStar Psoralen-Biotin Nonisotopic Labeling Kit (Ambion). The membranes were hybridized to , ßA, and ßB subunit probes in ULTRAhyb buffer. The biotinylated DNA probes were detected using the BrightStar BioDetect chemiluminescent kit (Ambion). Blots were exposed to film for 1 h and the band size was compared with that of the rat.
Antibodies
Inhibin-directed antibodies have been previously used to analyze the , ßA and ßB subunits from a variety of species and were therefore used in this study [12, 13]. Rabbit polyclonal antibodies directed against , ßA, and ßB subunits of inhibin (a gift from W. Vale and J. Vaughn, The Salk Institute, La Jolla, CA) were used at concentrations of 3 µg/ml, 2 µg/ml, and 1.5 µg/ml, respectively. Rabbit polyclonal antibodies directed against TGFß1, TGFß2, and TGFß3 (a gift from G. Rodriquez, Northwestern University, Evanston, IL) were used at final concentrations of 4 µg/ml, 1.3 µg/ml, and 1.3 µg/ml, respectively. The following five antibodies were purchased from R&D Systems (Minneapolis, MN): goat polyclonal antibodies directed against ActRIB, ActRIIA, and ActRIIB used in immunohistochemical analysis at final concentrations of 10 µg/ml, 4 µg/ml, and 4 µg/ml, respectively; goat polyclonal antibodies directed against TGFß RII used at a final concentration of 10 µg/ml; and goat polyclonal antibodies directed against bone morphometric protein (BMP) RII used at a final concentration of 4 µg/ml. Rat antibodies directed against growth differentiation factor 9 (GDF-9; a gift from R&D Systems) were used at a final concentration of 3.8 µg/ml. The next two antibodies were purchased from Zymed Laboratories (South San Francisco, CA), rabbit polyclonal anti-Smad2 and anti-Smad3 antibodies used at final concentrations of 2.5 µg/ml and 1.25 µg/ml, respectively. Goat polyclonal antibodies directed against anti-Smad4 were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and used at a final concentration of 2 µg/ml.
Immunohistochemistry and Immunofluorescence
For each individual antibody used in the following procedures, 10–15 ovarian sections from at least three different ovaries, all from different cats, were analyzed. Practice cat ovaries were used to figure out the appropriate antibody dilutions for this tissue type; consistent and repeatable results were obtained thereafter. Immunohistochemistry was performed using a previously described method [12]. For immunofluorescence, the TSA Plus Fluorescence System (NEN Life Sciences Products, Inc., Boston, MA) was used for detection via deposition of a fluorophore-labeled tyramide proximal to the horse radish peroxidase enzyme site. Slides were mounted with Vectashield containing 4',6'-diamidino-2-phenylindole. For all instances, replacing the primary antibody with buffer and adding only secondary antibody was used as a negative control. Additionally, if available, blocking peptides or recombinant proteins were used as a specific negative control to preabsorb the primary antibody. Images were acquired using a SPOT RT monochrome digital camera (Diagnostic Instruments, Sterling Heights, MI) and METAMORPH IMAGING (Version 4.6; Universal Imaging, Downington, PA).
Follicle Classification
Currently, there are no studies that specifically classify cat ovarian follicles throughout folliculogenesis. Therefore, follicles were classified according to criteria that have been commonly applied to rodent and human models. Primordial follicles contained mostly oocyte cytoplasm, with less than a 25-µm diameter oocyte, and were categorized as in Gougeon et al. [14] and Meredith et al. [15]. B-class primordial follicles contained squamous pregranulosa cells, and B/C-class primordial follicles contained a single layer of squamous and cuboidal granulosa cells (B/C follicle). Primary follicles (C follicles) contained oocytes greater than 25 µm in diameter, and a single layer of cuboidal granulosa cells. Secondary follicles contained an enlarged oocyte with one or more layers of granulosa cells, plus a theca cell layer. Early antral follicles contained a small antral space, an intact granulosa cell layer, as well as a theca cell layer. Large antral follicles (500 µm–1 mm in diameter) contained healthy intact mural and cumulus granulosa cell layers, theca cell layers, and were positioned near the periphery of the ovary. Atretic follicles varied in size and contained morphological evidence of cellular pyknosis, disorganization of the granulosa cell layer described as a lacy appearance, and shedding of the granulosa cell layer into the antral cavity. Additionally, the atretic follicles contained an oocyte with evidence of cell death, which included oolemma and zona pellucida expansion and a loss of discrete borders. The corpus lutea were structures characterized morphologically by sprawling luteinized cells with an appearance of lipid deposits most likely due to progesterone synthesis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mature region of the cat inhibin subunit cDNA was cloned by RT-PCR (GenBank accession number AY258627) and ligated into a cloning plasmid. The cat mature subunit cDNA sequence was 90%, 86%, and 82% identical to orthologous sequences in horse, human, and rat, respectively. The amplified cat subunit cDNA was 465 base pairs (bp) in length and corresponds to signal peptide sequence and nucleotides 895–1356 of the rat subunit sequence (GenBank accession number M36453). The cat activin ß subunit cDNAs were sequenced and mature ßA subunit was 94%, 93%, and 91% identical to orthologous sequences in horse, human, and rat, respectively. The cat ßA subunit (GenBank accession number AY258628) was 1275 bp in length and corresponds to signal peptide sequence and nucleotides 86–1366 of the human ßA subunit sequence (GenBank accession number NM_002192). The mature ßB subunit was 96%, 94%, and 91% identical to orthologous sequences in human, pig, and rat, respectively. The ßB subunit (GenBank accession number AY258629) was 674 bp in length and corresponded to signal peptide sequence and nucleotides 559–1224 of the human ßB subunit sequence (GenBank accession number NM_002193).
The relative sizes of cat , ßA, and ßB subunit mRNAs were measured by RNA blot analysis and compared with other known species (data not shown). Cat subunit RNA is 1.5 kilobases (kb). This is comparable with the rat (1.5 kb) and the Siberian hamster (1.6 kb) [13, 16]. The cat ßA subunit RNA is 7.0 kb in length (rat and Siberian hamster are 6.8 kb) [13, 16]. The ßB subunit RNA is 4.4 kb (Siberian hamster is 4.4 kb) [13].
Localization of Inhibin and Activin Subunits in the Cat Follicles
To investigate where activin and inhibin subunits are produced in the feline ovary, immunohistochemical localization of , ßA, ßB subunits was done using specific antibodies (summarized in Table 2). Anti- subunit antibody was preincubated with inhibin A medium from a stably expressing Chinese hamster ovary (CHO) cell line and used as a specific negative control. No immunoreactivity was detected using the immunodepleted antibody, indicating that staining represents authentic inhibin subunit protein (Fig. 1A). Inhibin subunit protein is weakly detected in the oocyte cytoplasm of B-class primordial follicles (Fig. 1B). The single layers of cuboidal granulosa cells surrounding two representative C-class primary follicles have higher staining in the granulosa cell cytoplasm when compared with the oocyte cytoplasm (Fig. 1C). Protein levels are intense in the cytoplasm of the granulosa cells of secondary follicles (Fig. 1D) but the oocyte is no longer immunoreactive. This follicle also contains a healthy theca cell layer in which no subunit protein is detected. Of the 27 small antral follicles analyzed for this protein, the mural and cumulus granulosa cells of approximately 60% are immunoreactive (Fig. 1, E and H). However, the subunit is not equally localized in all small antral follicles. About 40% are found to have weak to no granulosa cell staining (Fig. 1F) or are found to contain moderately positive granulosa cell staining (Fig. 1G). Few large antral follicles were present in the collected ovaries; however, one 
1-mm-diameter follicle was examined. Similar to subunit production in the rodent, the mural granulosa cell cytoplasm stained positive, whereas the cumulus cells surrounding the oocyte are negative (Fig. 1I). The inset depicts the clear distinction between stained mural granulosa cells and the lack of protein in the cumulus layer. Many follicles in each tissue section were in various stages of atresia. Atretic follicles in all size classes have evidence of cytoplasmic granulosa cell subunit protein production (Fig. 1J). Morphological criteria described in the Materials and Methods was used to classify atretic follicles, and additional TUNEL staining was completed to demonstrate that the follicle cells were apoptotic (data not shown). No subunit protein is detected in the corpus luteum (CL) (Fig. 1K).
|
|
|
ßA subunit protein was localized to similar size follicles. Anti-ßA subunit antibody was preincubated with activin A media from a stably expressing CHO cell line and used as a specific negative control. Again, no immunoreactivity was detected using the immunodepleted antibody, indicating that staining represents authentic ßA subunit protein (Fig. 2A). Abundant protein is present in the oocyte cytoplasm in B-class primordial follicles (Fig. 2, B and C). Moderate protein levels are detected in the oocyte cytoplasm, with weak production in granulosa cell cytoplasm of the C-class primary follicles observed (Fig. 2C). Secondary follicles display more granulosa cell cytoplasm subunit protein in addition to localized nuclear and cytoplasmic oocyte protein and a weak occurrence of theca cell immunoreactivity (Fig. 2D). ßA subunit protein is present in the mural and cumulus granulosa cells of small antral follicles (Fig. 2E). However, a notable difference from subunit localization is the reactivity of the theca cell layer in small antral follicles as well as protein presence in the nuclear and cytoplasmic compartments of the oocyte. A small amount of staining is also apparent in the antral space of these follicles. The same 1-mm-diameter large antral follicle depicted in Figure 1 also has abundant ßA subunit protein demonstrating clear cytoplasmic localization. A 60x magnification of the mural granulosa cell and theca cell layers is shown (Fig. 2F). Dense staining is apparent in the cytoplasm of granulosa cells directly at the basal lamina with protein abundance decreasing toward the antrum. The theca cells display a moderate amount of immunoreactivity, and consistent throughout, the oocyte cytoplasm and nucleus are positive. Last, the antral space of the large antral follicle is densely stained, indicating protein presence. Analysis of atretic follicles (TUNEL positive) indicates high production of the ßA subunit largely in the cytoplasm of granulosa cells closest to the antrum, those cells falling into the antral space, and those cells directly surrounding the oocyte (Fig. 2G). The oocyte has clearly lost both oolemma and zona pellucida borders. The theca cell layer of atretic follicles does not have detectable ßA subunit protein. Similar to the subunit, no ßA subunit is detected in the CL (Fig. 2H).
|
ßB subunit protein is not detected in the immunodepleted control slides where the primary antibody was preabsorbed with activin B-containing CHO medium (Fig. 3A). Protein is abundant in the cytoplasm and nucleus of the B-class primordial follicles (Fig. 3B). The C-class primary follicles have intense staining in the cytoplasm of the oocyte and granulosa cells (Fig. 3C). The granulosa cell staining pattern diminishes to a moderate signal in the secondary follicles (Fig. 3D), and the oocyte cytoplasm diminishes in signal with a moderate reaction present in the oolemma. The small antral follicle granulosa cells have low protein levels, with coordinate protein detected in the antral space (Fig. 3E). The 1-mm-diameter large antral follicle contains subunit protein throughout the layers of granulosa cell cytoplasm with no apparent difference between mural and antral cells (Fig. 3F). However, there is no theca cell reactivity and no oocyte present. Protein is detected in the antral space in moderate abundance similar to that seen with the ßA subunit antibody. The representative atretic follicle (TUNEL positive) is intensely positive in many areas, including the granulosa cell cytoplasm, antral space, oocyte cytoplasm, oolemma, and zona pellucida (Fig. 3G). In this study, we did not observe ßB protein immunolocalization in CLs (Fig. 3H).
|
Localization of Activin Receptors in the Cat Follicles
To investigate the activin receptor protein localization patterns in the different size classes of follicles, we again performed immunolocalization studies (results summarized in Table 3). Recombinant proteins were obtained to preabsorb ActRIB, ActRIIA, (purchased from R&D Systems, Minneapolis, MN) and ActRIIB antibodies for immunodepleted negative controls (Fig. 4, A, E, and I). ActRIB is intensely positive in both B and B/C-class primordial follicle oocyte cytoplasm and nucleus (Fig. 4B). ActRIB protein in secondary follicles diminishes to moderate levels and becomes primarily localized in the nucleus of the granulosa cells (Fig. 4C). The discrete punctate staining pattern seen in these granulosa cells is known to occur with receptor antibodies [17–19]. The oocyte contains a moderately positive oolemma, while the theca cells are weakly positive. Analysis of small antral follicles show weak ActRIB immunoreactivity in the cytoplasm and moderate levels in the nucleus of granulosa cells in addition to a moderate staining pattern again in the oocyte oolemma (Fig. 4D). The ActRIIA antibody is not very immunoreactive in the cat ovary. A weak stain is present in the B and B/C-class primordial follicles inspected (Fig. 4F). Negative results were obtained for the secondary follicle (Fig. 4G) and the small antral follicle (Fig. 4H) stages for protein localization of this receptor except that the secondary follicle oocyte oolemma is slightly positive (Fig. 4G). The ActRIIB receptor demonstrates stain patterns more like that of ActRIB. The B-class primordial follicles contain cytoplasmic and nuclear oocyte localization (Fig. 4J). The C-class primary follicles have intensely positive granulosa cell and oocyte cytoplasm staining (Fig. 4J). ActRIIB protein is abundant in the cytoplasm of granulosa cells and in the cytoplasm and nucleus of the oocyte of secondary follicles (Fig. 4K). The staining pattern in the oocyte appears a bit punctate, analogous to the secondary follicle staining by anti-ActRIB. Small antral follicles are immunoreactive in the cytoplasm and nucleus of the granulosa cells (Fig. 4L) as well as weakly positive in the oocyte cytoplasm.
|
|
|
Localization of Cytoplasmic Coactivators of Activin
Smad2 protein is moderately localized to the cytoplasm of B-class primordial follicles (Fig. 5A). A notable feature in Figure 5A is the strong Smad2 immunoreactivity of the surface epithelial cells. More intense staining appears in the B/C-class primordial follicle granulosa cells and oocyte cytoplasm (Fig. 5B). Protein levels diminish in the secondary follicle as Smad2 is found in moderate levels in granulosa cell cytoplasm and in the oolemma and zona pellucida of the oocyte (Fig. 5C). The theca cells are also weakly immunopositive (Fig. 5C). The granulosa cells of the small antral follicle remain the same as the secondary follicle, but immunoreactivity of the zona pellucida is lower and the oolemma slightly higher (Fig. 5D). For a Smad2-negative control, the primary antibody was replaced with buffer and the specimens were only incubated with secondary antibody (data not shown). Smad4 protein is apparent at moderate levels in the cytoplasm of B-class primordial follicles (Fig. 5E). Similar to Smad2 production, there is Smad4 surface epithelial cell reactivity (Fig. 5E). The C-class primary follicle depicted acquired granulosa cells with moderate cytoplasmic staining and contains an oocyte that is moderately immunoreactive in its cytoplasm and nucleus (Fig. 5F). The secondary follicle contains intensely stained granulosa cells, weak theca cell protein levels, and a weak appearance of Smad4 protein in the oocyte cytoplasm (Fig. 5G). Smad4 in the small antral follicles is intensely localized in the granulosa cell cytoplasm, but in contrast with Smad2, weakly apparent in only the oocyte cytoplasm (Fig. 5H). For a Smad4-negative control, blocking peptides were purchased for Smad4 (Santa Cruz Biotechnology, Inc.) and preabsorption of anti-Smad4 was done (data not shown).
|
Localization of Smad3 in Developing Follicles
To best colocalize Smad3 in nuclear vs. cytoplasmic compartments, immunofluorescent studies were performed. Smad3 is detected in the nucleus of all primordial follicle oocytes (Fig. 6, A and B). The two B-class primordial follicles in Figure 6A and one in Figure 6B only show strong Smad3 localization in the oocyte nucleus. The B/C-class primordial follicle contains the same oocyte nuclear protein localization in addition to two lateralized cuboidal granulosa cells that appear moderately fluorescent (Fig. 6A). Figure 6B also shows a B/C-class primordial follicle with Smad3 in the nucleus of the granulosa cells. The C-class primary follicle containing only cuboidal granulosa cells possesses intense nuclear fluorescence of the oocyte (Fig. 6C). Note that the granulosa cells of the C-class follicle in Figure 6C produce Smad3 protein in a somewhat asymmetric pattern, with the lower portion of the follicle more intensely fluorescent than the upper portion. Another feature of Smad3 immunostaining is its localization in the surface epithelial cells similar to that seen with anti-Smad2 and anti-Smad4 antibodies (Fig. 6A). The secondary follicles contain intense localization in both the nuclear and cytoplasmic areas of the granulosa cells (Fig. 6D). Again, the oocyte is positive along with slight Smad3 protein concentrated in the nucleus of the theca cells. The control panels (Fig. 6, E–H) are images taken of control slides (addition of secondary antibody only) at the respective magnification for each follicle size class at the same exposure time as the Smad3 antibody containing slides. Punctate fluorescence is considered to be background due to a moderate amount of autofluorescence that is known to appear in the mammalian ovary.
|
Smad3 was localized by the same methods in larger follicle size classes as well. Control (addition of secondary antibody only) images represent background fluorescence (Fig. 7, A and B). As expected, large antral follicles have high levels of Smad3 protein in the nucleus and cytoplasm of the granulosa cells, especially the cumulus granulosa cells (Fig. 7C). The 60x views of the large antral follicle cumulus and mural granulosa cell protein levels are shown (Fig. 7, D and E). Another 60x view of the area between two large antral follicles, including their theca cell layers, is depicted (Fig. 7F). The strong nuclear and cytoplasmic granulosa cell Smad3 localization can now be distinguished from the moderate nuclear localization in the theca cell layer.
|
Other TGFß Superfamily Ligands and Receptors in Cat Follicles
Other members of the TGFß superfamily may also be important in cat follicle development; therefore, analysis of TGFß and GDF-9 along with their respective receptors was examined (results summarized in Table 4). Only two follicle size classes for each antibody were examined, primary follicles and antral follicles. In all cases for Figure 8, negative controls included the addition of secondary antibody only. TGFß1 protein is strongly localized to the granulosa cell cytoplasm of C-class primary follicles (Fig. 8A) and moderately produced in the granulosa cells of small antral follicles (Fig. 8E). Additionally, a small amount of protein is detected in the oocyte cytoplasm of small antral follicles. TGFß2 protein is intensely localized in the granulosa cell cytoplasm as well as in the oocyte cytoplasm of C-class primary follicles (Fig. 8B). In the large antral follicles, the staining pattern is the same as that for the primary follicles; however, there is protein present in the nucleus of these cells and a small amount in the theca cells (Fig. 8F). TGFß3 protein is only weakly produced in the cytoplasm of the granulosa cells and oocyte of C-class primary follicles (Fig. 8C). Localization of TGFß3 protein in large antral follicles is identical to the staining pattern for TGFß2 (Fig. 8G). Abundant GDF-9 protein is detected exclusively in the oocyte cytoplasm of both C-class primary and small antral follicles (Fig. 8, D and H). This further confirms previous reports found by others.
|
|
Localization of the type II receptor for the TGFß ligands was assessed by immunofluorescence. TGFß RII protein is not present in the primary follicle population (data not shown). However, it is largely localized to the cytoplasm of granulosa cells and the zona pellucida of the oocyte in large antral follicles (Fig. 8, K and L). Control images were taken at the same magnification and time exposure (Fig. 8, I and J). Comparison of the control images with those that were incubated with TGFß RII antibody show definite reactivity over background fluorescence.
GDF-9 shares the BMP type II receptor with other BMP ligands. This receptor could be detected by immunohistochemical stain, and protein is present in large amounts in the oocyte of primordial follicles and in modest amounts in the granulosa cell and oocyte cytoplasm of C-class primary follicles (Fig. 8M). Weak protein detection is present in the cytoplasm of granulosa cells in small antral follicles in addition to the oolemma of the oocyte (Fig. 8N).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cat inhibin/activin subunits were cloned and conservation with other species was analyzed. The protein sequence encoding the mature region of the cat subunit was 89%, 92%, and 92% identical to protein sequences of rat, horse, and human, respectively [16, 21, 22]. The protein sequences of the ß subunits were even more highly conserved. The mature region of the cat ßA subunit shared 98%, 100%, and 100% protein sequence identity with horse, human, and rat, respectively [16, 22, 23]. The mature region of the cat ßB subunit shared 97%, 99%, and 100% protein sequence identity with rat, pig, and human, respectively [24, 25]. This degree of conservation between orthologs suggests that the ligands have been highly conserved during evolution. Therefore, the positions where differences occur may point to nonessential amino acids with respect to function.
Inhibin and activin subunits are present in many follicle size classes in the cat. Specifically, the inhibin subunit is present in the granulosa cell cytoplasm of small, large, and atretic follicles. With only a few exceptions, these findings concur with protein localization in the mare [26], mRNA and protein localization in the human [27, 28], mRNA localization in the primate [29], and mRNA and protein localization in the rat [30]. In the human, and only slightly in the mare, do the theca cells express inhibin subunit [26–28]. This was not found to be true in the cat ovary nor is it the case in the rat [31] and primate ovary [29]. Low levels of inhibin subunit protein were detected in primordial oocytes. This correlates with previous work by Meunier et al. who showed subunit mRNA and protein in rat primordial follicles [30]. Intriguingly, not all small antral follicles produce the same level of subunit protein. A gradient of protein production was apparent, consistent with an emerging notion that these ligands establish morphogenlike gradients in certain tissues [32]. Additionally, subunit expression was localized to only the mural granulosa cells of the large antral follicle and not the cumulus granulosa cell layer. This pattern is analogous to subunit mRNA expression in the rat [31].
The activin ßA and ßB subunits are present in all follicle size classes in the cat. Activin ßA subunit mRNA is primarily localized in antral follicles in the human, primate, and rat [27, 29, 30]. Activin ßB subunit is found in granulosa cells of all size follicles in the cow [33], small antral follicles in the human [27], and preovulatory follicles in the primate [29, 34]. In corroboration with these previous studies, the cat ovary contains both subunit proteins in the oocyte and granulosa cell cytoplasm of primordial- and primary-stage follicles. ßA subunit was more abundant in secondary follicles, while ßB was moderately present in the oocyte oolemma of secondary follicles. Hence, both activin ß subunits are intense in smaller sized follicles, suggesting a role for activin in the early stages of development. Granulosa cells of large antral follicles produced considerable amounts of both subunits, and a gradient of ßA protein was detected in these follicles. Similar to results found in the human, ßA subunit is localized to the theca cells of large antral cat follicles whereas ßB is not [27, 28]. Because the ß subunits coincide with subunit position, inhibin dimers may play a more dominant role in follicles at late stages of development. Analogous to the subunit, ßA and ßB subunits were localized to granulosa cells of atretic cat follicles, ßB subunit abundantly present in the zona pellucida and oolemma of the oocyte. Immunoreactivity of granulosa cells in atretic follicles for the three subunits has also been shown in the human [12]. While human [27] and primate [29, 34] ovaries express subunit mRNA and protein in the CL, luteal cells in the cat CL did not produce any of the subunits. Therefore, the feline CL does not appear to be a source of inhibin or activin. This finding is consistent with that found in the mare [26] and rat [35].
The activin RIB and RIIB receptors are necessary for local ligand action. ActRIIB was present in all follicles examined, while ActRIB was strongly present in the primordial-, primary-, and secondary-stage follicles and only weakly in antral follicles. These data are consistent with the finding that ActRIB mRNA is present in the postnatal rat ovary, but the level of expression declines with increasing follicular development [36]. Feline ActRIIA was the least abundant of the activin receptor subtypes and was only slightly positive in the oocyte. This is in concert with the finding that granulosa cells of preantral and early antral follicles are negative for ActRIIA in the human, whereas secondary oocytes display ActRIIA immunoreactivity [12]. Contrary to our work, other studies have shown ActRIIA more highly expressed than ActRIIB [36–38]. These data suggest that activin signaling takes place in the early stages of follicle development.
Smad protein was localized to specific follicle populations by immunohistochemistry and immunofluorescence. Similar to other species, Smad2 and Smad4 were present in primordial follicles, suggesting that this follicle population is responsive to TGFß-like ligands [36]. Smad3 was located in the nucleus of granulosa cells and oocytes of all follicle stages. Theca cells from large antral follicles were Smad3 positive. Finally, Smads2, 3, and 4 are present in the ovarian surface epithelial cells. Smads have been shown to act as tumor suppressors in many cell types, and presence and abundance of Smads2, 3, and 4 in the surface epithelial cells suggest that future studies could investigate the role these coregulators play in ovarian surface epithelium neoplasia. Smad2 protein localization in the zona pellucida was a puzzling observation; therefore, further studies are required to determine if this discrete staining pattern is specific.
Multioocytic follicles are a natural occurrence in cats, and as depicted in Figure 5, they are positive for the Smads as are unioocytic follicles. Multioocytic follicles are also frequently found in the dog, rabbit, and rhesus monkey [39]. Additionally, they occur in humans in association with in vitro fertilization although at a much lower percentage [40]. The origins and functionality of these structures are unknown.
In previous studies, GDF-9 was expressed in the mammalian oocyte throughout follicle development [41, 42]. Our results phenocopy this pattern because protein levels were found to be restricted to the oocyte of primary and antral follicles. The GDF-9 receptor (BMP RII), however, is mainly abundant in primordial and primary follicles, therefore suggesting signaling action in the early stages of follicle maturation. TGFß ligands 1, 2, and 3 were localized to granulosa cells, oocytes, and theca cells at different intensities for the follicle classes observed. TGFß RII was absent in primary follicles but present in the granulosa cells of antral follicles. These data suggest that TGFß ligands can act on antral follicles and activate the Smad signaling pathway, whereas GDF-9 action may be restricted to small follicles.
To summarize, these results suggest discrete follicle populations and times of action for activin, inhibin, GDF-9, and TGFß in the cat ovary. Specifically, activin and GDF-9 may act in the earliest stages of follicular development, TGFß on intermediate follicle populations, and inhibin may act in the latest stages of follicular development. This comprehensive analysis of ligands, receptors, and signaling proteins provides a new paradigm for dynamic and follicle-restricted roles of the TGFß superfamily. In conjunction with known in vitro maturation requirements, these findings may help to advance reproductive technology aimed at preserving endangered feline species.
|
ACKNOWLEDGMENTS |
|---|
, ßA, ßB subunit antibodies, and Gus Rodriguez (Northwestern University) for providing the TGFß ligand antibodies. Additionally, we thank Jaroslav Jelen for help with cloning the ßB subunit and Stacey Chapman Tobin for critical reading of the manuscript. |
FOOTNOTES |
|---|
2 Correspondence: Teresa K. Woodruff, Northwestern University, Department of Neurobiology, O.T. Hogan 4-150, 2205 Tech Drive, Evanston, IL 60208. FAX: 847 491 2224; tkw{at}northwestern.edu
Received: 1 August 2003.
First decision: 20 August 2003.
Accepted: 25 November 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
- and ß-subunit complementary deoxyribonucleic acids and expression in the ovary. Mol Endocrinol 1987 01:561-568, ßA, ßB subunits in the ovary. Biol Reprod 1998 59:1392-1398, ßA, ßB messenger ribonucleic acids in the normal human ovary and in polycystic ovarian syndrome. J Endocrinol 1994 143:127-137This article has been cited by other articles:
|
|
 |
|
  M.-M. Dolmans, B. Martinez-Madrid, E. Gadisseux, Y. Guiot, W. Y. Yuan, A. Torre, A. Camboni, A. Van Langendonckt, and J. Donnez Short-term transplantation of isolated human ovarian follicles and cortical tissue into nude mice Reproduction, August 1, 2007; 134(2): 253 - 262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. G Knight and C. Glister TGF-{beta} superfamily members and ovarian follicle development. Reproduction, August 1, 2006; 132(2): 191 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Anttonen, H Parviainen, A Kyronlahti, M Bielinska, D B Wilson, O Ritvos, and M Heikinheimo GATA-4 is a granulosa cell factor employed in inhibin-{alpha} activation by the TGF-{beta} pathway. J. Mol. Endocrinol., June 1, 2006; 36(3): 557 - 568. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. I Suszko and T. K Woodruff Cell-specificity of transforming growth factor-{beta} response is dictated by receptor bioavailability. J. Mol. Endocrinol., June 1, 2006; 36(3): 591 - 600. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Bristol-Gould, C. G. Hutten, C. Sturgis, S. M. Kilen, K. E. Mayo, and T. K. Woodruff The Development of a Mouse Model of Ovarian Endosalpingiosis Endocrinology, December 1, 2005; 146(12): 5228 - 5236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Hussein Apoptosis in the ovary: molecular mechanisms Hum. Reprod. Update, March 1, 2005; 11(2): 162 - 178. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.L. Juengel and K.P. McNatty The role of proteins of the transforming growth factor-{beta} superfamily in the intraovarian regulation of follicular development Hum. Reprod. Update, March 1, 2005; 11(2): 144 - 161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J R V Silva, R van den Hurk, H T A van Tol, B A J Roelen, and J R Figueiredo Gene expression and protein localisation for activin-A, follistatin and activin receptors in goat ovaries J. Endocrinol., November 1, 2004; 183(2): 405 - 415. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
