|
|
||||||||
Exodomain from Sheep Follicular Endothelium: Implication in the Ovulatory Process1
a Department of Animal Science, University of Wyoming, Laramie, Wyoming 82071
| ABSTRACT |
|---|
|
|
|---|
from contiguous endothelium, and apoptotic cell death. The objectives of this investigation were to determine whether plasmin elicits TNF
secretion from thecal endothelium of ovine follicles, to characterize the site(s) of enzymatic attack, and to assess the physiological consequence of soluble TNF
action. Endothelial cells of thecal tissues isolated from antral follicles of eCG-primed anestrous ewes shed (histochemical depletion) TNF
into incubation medium (ovarian cell DNA fragmentation bioassay, Western blot detection) upon exposure to plasmin. Immunopurification and N-terminal sequence analysis indicated that TNF
was excised from its transmembrane precursor at the Arg79-Ser80 and Lys88-Pro89 linkages. Microinjection of TNF
into the apical wall of explanted follicles induced cellular apoptosis and stigma development. We suggest that plasmin-mediated cleavage of TNF
exodomain from its membrane anchor along thecal endothelium is a determinant of tissue dissolution within the formative ovulatory rupture site of ewes.
| INTRODUCTION |
|---|
|
|
|---|
from thecal endothelium [4] and consequent programmed cell death (apoptosis) [5, 6]. Plasmin is a pleiotropic serine protease derived from plasminogen by enzymatic activation. Two forms of plasminogen activators have been characterized in vertebratesurokinase (uPA) and tissue (tPA) types [7]. Accumulation of plasmin within the apical wall of preovulatory ovine follicles was attributed to uPA secretion by juxtaposed ovarian surface epithelial cells stimulated by gonadotropin [8].
TNF
is a cytokine expressed as a 26-kDa integral transmembrane polypeptide that upon cleavage yields a 17-kDa extracellular domain subunit; the structure is well conserved across mammalian species (~80% homology at the amino acid level). Mature TNF
is a soluble noncovalent homotrimer. Common cell types that produce TNF
are leukocytes, smooth muscle, fibroblasts, and endothelial cells. Plasma membrane receptors for TNF
are present on virtually all nucleated cells [9, 10], including ovarian cells [11]. It is now apparent that, in addition to its ability to induce cytolytic cell death (hemorrhagic necrosis), TNF
can convey a signal that results in apoptosis [1214].
Early-stage apoptosis is characterized by endonuclease activation, internucleosomal DNA fragmentation, and nuclear condensation. Atrophic cells committed to apoptosis lose contact with their neighbors and supporting basement membranes; those that line cavities are sloughed and may completely disappear within a few hours [1518].
It was hypothesized that plasmin-mediated release of follicular TNF
dictates apoptosis within a limited diffusion radius encompassing the ovulatory site. Indeed, TNF
is a candidate substrate for serine protease attack [19, 20]. Experiments were conducted to establish whether plasmin liberates bioactive (apoptosis inducing) TNF
from thecal endothelium and to elucidate the proteolytic sites of enzymatic action. The capacity of apical follicular TNF
microinjection to elicit stigma development also was assessed.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Regulation by Plasmin of Bioactive Follicular TNF
Secretion
Five mature western-range ewes were treated during the anestrous season with 1000 IU eCG. Follicles
5 mm in diameter were isolated from the ovaries (25 per animal) at the time animals were killed (Beuthanasia-D i.v.; Schering-Plough Animal Health, Kenilworth, NJ) 36 h after eCG.
Granulosa cells were entirely dislodged from the antral surfaces of hemisected follicles with a polytetrafluoroethylene scraper designed to remove adherent cells from culture flasks (Becton Dickinson Co., Lincoln Park, NJ). Tissue forceps were used to detach theca from surrounding interstitium. Thecal shells of each follicle were incubated (1 h, 37°C, 70 µl Medium 199) with 0.15 IU native or heat-inactivated (control) bovine plasmin (n = 18). The functional dose of plasmin cleaved an enzyme-specific chromogenic substrate (H-D-norleucyl-hexahydrotyrosyl-lysine-p-nitroanilide diacetate salt) at a rate similar to that induced by extracts of preovulatory apical follicular wall [8].
Tissues were recovered from incubation media, fixed in Histochoice (Amresco, Solon, OH), embedded in paraffin, serially sectioned at 5-µm thickness, transferred onto microscope slides treated with subbing solution (0.025% chromium potassium sulfate/0.25% gelatin; to enhance tissue adherence), and immunostained for TNF
. Samples were incubated for 10 min in 10% normal goat serum and then for 30 min with a rabbit polyclonal antiserum (1:200 dilution) directed against recombinant ovine TNF
(AB1842; Chemicon International, Temecula, CA). Sections were washed in three changes of PBS. Immune complexes were detected with a secondary goat anti-rabbit IgG-fluorescein isothiocyanate conjugate (F 9887; 1:160, 10 min). Negative control reactions were performed in the absence of primary antiserum and with antibody preadsorbed with purified recombinant human TNF
(210-TA; R & D Systems Inc., Minneapolis, MN). Endothelial cells within four areas per tissue sample were photographed (Olympus [Tokyo, Japan] BH-2 equipped with a reflected light fluorescence attachment; Ektachrome 400 HC [Eastman Kodak, Rochester, NY], 40-sec exposure; x1000) and classified (without knowledge of treatment) as intensely immunoreactive or not.
Tissue-conditioned media of control and plasmin-treated groups were preabsorbed with TNF
antiserum or normal rabbit serum (5 µl, 1 h; n = 9) and combined in suspension with freshly prepared ovarian surface epithelial cells (1 x 103/0.1 ml final volume, 4 h, 37°C) [21]. Aliquots of the ovarian cell pool also were incubated with recombinant TNF
(0, 0.25, 0.5, 1, 2, 4, 8 ng; n = 5). An Oncor (Gaithersburg, MD) ApopTag In Situ S7111 Kit was used to detect internucleosomal DNA fragmentation (onset of apoptosis) in permeabilized cells [2, 4]. Briefly, exposed 3'-OH ends of DNA fragments were labeled with digoxigenin-11-D-uridine triphosphate by terminal deoxynucleotidyl transferase (TdT) catalysis. Incorporated nucleotide heteropolymers were localized with antidigoxigenin Fab-fluorescein isothiocyanate. Images of individual cells (
50 per sample selected at random) were categorized (Optimas, Bothell, WA) as positively labeled (luminance intensity > 2 times control [-TdT/conjugate] cell background) or nonreactive.
Western Detection of TNF
Thecal shells obtained from hemisected follicles (
5 mm) of four gonadotropin (1000 IU eCG)-primed ewes were or were not incubated with bioactive plasmin (n = 16). Conditioned media within treatment were combined and desalted/concentrated by filtrative centrifugation (Centricon-3; Amicon, Beverly, MA; 7500 x g, 2 h). Secreted proteins were separated by PAGE (15%) under reducing conditions (SDS) and transferred (single-strength Towbin's buffer with 10% methanol, 100 V, 1 h) to a nitrocellulose membrane (0.22 µm; Micron Separations Inc., Westborough, MA). Prestained molecular weight markers were used for standard comparisons. Membrane was blocked with single-strength Tris-buffered saline/Tween 20 (TBST) containing 1% BSA (40 min). The blot was subjected to reaction with TNF
antiserum (1:2000, 30 min; 3-strength TBST wash) and secondary anti-rabbit IgG-alkaline phosphatase (1:5000, S 373B; Promega, Madison, WI; TBST/TBS rinse). Color developer (Promega) was added for 30 min followed by a wash in deionized water. Negative and positive controls were carried out in the absence and presence of primary antibody with TNF
(10 ng).
Excision Links of TNF
Halved thecal shells recovered from follicles (
5 mm) of 10 eCG (1000 IU)-treated ewes were incubated with plasmin (n = 96). Media were pooled and ultrafiltered. TNF
was isolated from retentate using an Immunopure Protein G-IgG Orientation Kit according to the instructions of the manufacturer (Pierce, Rockford, IL); the goat polyclonal affinity-purified capture antibody was directed against the epitope corresponding to the 20-amino acid carboxy terminus of TNF
(sc-1347; Santa Cruz Biotechnology, Santa Cruz, CA). Eluted product was subjected to N-terminal sequence analysis (automated Edman degradation/HPLC) [22].
Injection of TNF
into the Ovarian Apex
Antral follicles (
5 mm) of four ewes treated with eCG (1000 IU) were isolated from ovaries in tissue blocks using a single-edged razor blade. Recombinant TNF
(2 ng) or vehicle (2 µl) was injected (5 µl Hamilton [Reno, NV] syringe fitted with a 27-gauge hypodermic needle) into the apical wall of follicles, which were then incubated in Medium 199 (4 ml, 37°C) for 4 or 8 h (n = 4). Ovarian surface and granulosa epithelial cells were removed from the apical hemisphere and granulosa cells from the basal hemisphere of follicles after the 4-h incubation and analyzed for fragmented DNA (
20 cells per sample). Follicles assigned to the 8-h incubation were fixed, paraffin-embedded, sectioned, and stained in hematoxylin and eosin using standard techniques; thickness of apical ovarian walls (theca + tunica albuginea) was estimated with Optimas software.
Statistical Analyses
Morphometric subsample data were averaged. Assignments of tissue/cellular/media samples to treatments were made at random. Treatment mean contrasts were made by Student's t-test or ANOVA and protected least-significant difference. Differences were considered significant at p < 0.05.
| RESULTS |
|---|
|
|
|---|
was diminished by incubation of tissues with plasmin (Fig. 1). There was a coincident increase in TNF
bioactivity (ovarian surface cell DNA fragmentation) in media conditioned by tissues exposed to plasmin; the effect was neutralized by TNF
antibodies. The apoptotic reaction to plasmin was within the linear dose-response range of recombinant TNF
(Fig. 2).
|
|
Western analysis of media conditioned by explants of plasmin-treated theca revealed a distinctive protein band of approximately 17 kDa in accord with soluble TNF
standard. There was no evidence of TNF
secretion from control thecal tissues (Fig. 3).
|
Immunoaffinity isolation yielded a dominant protein with an amino terminus, Ser-Ser-Ser-Gln-Ala-Ser-Asn, that corresponds to deduced ovine TNF
truncated at the carboxy side of Arg79. A secondary N-sequence also was detected, Pro-Val-Ala, denoting a cleavage at Lys88 [23].
Injection of TNF
into the apex of follicles maintained in vitro for 4 h resulted in a localized phenomenon of cellular DNA fragmentation (Fig. 4); the dose (2 ng) was relative to the efficacy of incubation media conditioned by plasmin-treated thecal tissues (Fig. 2). After 8 h of incubation, complete deprivation of the ovarian surface epithelium was apparent in the immediate area of cytokine microinjections; either circumjacent granulosa cells were totally absent (~50% of sections examined), or diminution of the layer was evident. Cells of the basal follicular wall appeared healthy. Thickness of the connective tissue matrices of the theca and tunica albuginea along the ovarian dome were attenuated after exposure to TNF
(Fig. 5). Follicular rupture did not occur within the time frame of the experimental protocol.
|
|
| DISCUSSION |
|---|
|
|
|---|
release, and suppressed by intrafollicular injections of uPA antibodies,
2-antiplasmin, or TNF
antibodies [2, 4, 8], provided circumstantial evidence for an obligate enzyme-cytokine interplay in the ovulatory process [3]. Results of this investigation demonstrate that plasmin cleaves TNF
exodomain from thecal endothelium and that TNF
induces changes in explanted follicular walls consistent with those of stigma formation (i.e., loss of epithelial cells and theca/tunica albuginea dissolution). Thus, the prerequisite phase in the proposed ovulatory cascade is plasmin dependent. Secretion of plasminogen activators by thecal and granulosa cells of gonadotropin-stimulated follicles of rodents has been established; both uPA and tPA apparently contribute to ovarian plasmin biosynthesis and ovulation [2427]. An alternative source of proteolytic enzymes is the ovarian surface epithelium [28]. Receptors for gonadotropins have been detected on ovarian surface cells [29]. Plasmin accretion within the apical ovarian-follicular wall in the sheep was due primarily to uPA secretion (tPA was undetectable and ovulation was not altered by tPA antibodies) by surface epithelium [8] stimulated locally by LH derived from hyperemic thecal blood vessels ([30, 31], unpublished results). Plasminogen activators also were preferentially elevated within the apices of preovulatory porcine [32] and rat [33] follicles. In certain species (e.g., horse and armadillo) ovulation is restricted to a discrete ovarian depression (fossa) covered by prototypical (coelomic mesothelial derived) surface epithelium [34]. Stigma formation and follicular rupture were inhibited in ewes by surgical removal of ovarian surface epithelium [8].
Cleavages of TNF
at arginine or lysine linkages of the parent molecule are consistent with reported (trypsin-like) hydrolytic specificities of plasmin [7]. There is some variability in the predicted amino acid sequences of N-termini TNF
exodomains among species that could dictate substrate susceptibility to enzymatic processing (Arg2 and Lys11, which correspond to Arg79 and Lys88 of the precursor molecule, are conserved) [23]. Human TNF
is released from the cell surface at the Ala76-Val1 bond by a metalloproteinase disintegrin (TACE, TNF
-converting enzyme) [35]; however, additional putative sites of enzymatic attack exist between (exposed) residues +1 and +12 [36].
Insight into the signal transduction pathway that mediates the apoptotic effect of TNF
is evolving. Receptors of 55 (TNFR1) and 75 (TNFR2) kDa bind trimeric ligand through a homologous extracellular motif [10, 37, 38]. TNFR1, which contains a cytoplasmic death domain (DD), initiates apoptosis (TNFR2 activates the transcriptional factor NFkB). Aggregation of receptors orients the death domains in a conformation that recruits the adapter proteins TRADD (TNFR-associated protein with a DD) and FADD (Fas-associated protein with a DD). The death effector region of FADD can evidently recruit the zymogen forms of certain cysteine proteases (e.g., FLICE/Mach
1), which upon activation evoke a proteolytic cascade leading to DNA fragmentation and cytotoxicity [3942].
At high tissue concentrations, TNF
also initiates microvascular coagulation associated with necrotic cell death [12] and inflammatory tissue damage symptomatic of the ovulatory process [43]. Vascular lesions typical of hemorrhagic necrosis are observed within the immediate area surrounding the stigma of the preovulatory ovine follicle [6, 31, 44, 45]. A lack of blood flow (ischemia) into the ovulation papilla, leading to oxygen deprivation and toxic metabolite accumulation, would predictably potentiate cell death.
The prospective role of plasmin in the mechanics of follicular rupture is not limited to TNF
activation and cell death. Plasmin activates latent collagenases [7] that degrade the connective tissue matrices of the follicular theca and tunica albugineathereby weakening the ovarian wall [2, 24, 46] (the general consensus of functional studies indicates that uPA regulates tissue degradation, whereas tPA, which has a strong affinity for fibrin, is involved in thrombolysis [7]). Endogenous collagenolytic activity was greater within the follicular apex than the base of preovulatory ovine follicles [47]. Moreover, TNF
deposition within the apical ovarian hemisphere of follicles (present study) appeared to provoke connective tissue breakdown indicative of stigma formation. In fact, collagenase gene expression in dermal fibroblasts and chondrocytes was induced by TNF
[4850]. TNF
might therefore augment ovulatory collagenolysis by stimulation of interstitial de novo metalloproteinase biosynthesis. That ovulation did not occur after TNF
microinjection of explanted follicles may reflect a lack of obligate in vivo inputs, such as smooth muscle contractility [51], that impel rupture of a devitalized tissue.
Research findings with species other than the sheep are supportive of the concept that TNF
is a determinant of ovulation. Perfused rat ovaries released TNF
into venous effluent during ovulation [52], and the addition of TNF
to perfusates increased LH-induced ovulation rates [53]. Preovulatory human and bovine follicles also secrete bioactive TNF
[11, 5456].
A model is proposed of temporal interactions of gonadotropin, ovarian cell types, plasminogen activator/plasmin, collagenases, and TNF
in the breakdown of the apical follicular wall during ovulation: vascular transudate containing LH is delivered to receptor-bearing cells (i.e., granulosa, theca interna, surface epithelium), stimulating secretion of plasminogen activator; interstitial plasminogen is converted to plasmin, which activates collagenases and cleaves TNF
from thecal endothelium; collagenases disrupt the fibril network of the theca and tunica albuginea and absolve the basement membranes supporting the ovarian and granulosa epithelia; TNF
induces apoptosis and enhances collagenolysis; collagen degradation and cellular effacement instigate stigma development and follicular rupture.
| FOOTNOTES |
|---|
2 Correspondence: W.J. Murdoch, Department of Animal Science, P.O. Box 3684, 16th and Gibbon St., University of Wyoming, Laramie, WY 82071. FAX: 307 766 2355; wmurdoch{at}uwyo.edu ![]()
Accepted: December 11, 1998.
Received: October 12, 1998.
| REFERENCES |
|---|
|
|
|---|
in the ovulatory mechanism of ewes. J Anim Sci 1997; 75:16011605.
in follicular development, ovulation, and the life span of the corpus luteum. Domest Anim Endocrinol 1997; 14:115.[CrossRef][Medline]
. FASEB J 1990; 4:32153223.[Abstract]
: problems with cloning by inverse PCR. Gene 1991; 109:203210.[CrossRef][Medline]
from cells. Nature 1997; 385:729733.[CrossRef][Medline]
stimulated collagenase and stromelysin message levels in chondrocytes. Biochem Biophys Res Commun 1993; 196:11331142.[CrossRef][Medline]
induces collagenase gene expression by human skin fibroblasts. J Invest Dermatol 1995; 105:197202.[CrossRef][Medline]
(TNF
) on ovulation in the rat ovary. Reprod Fertil Dev 1995; 7:6773.[CrossRef][Medline]
. FEBS Lett 1990; 261:253255.[CrossRef][Medline]
from the human preovulatory follicle. Am J Reprod Immunol 1998; 39:387390.This article has been cited by other articles:
![]() |
J. Prange-Kiel, C. Kreutzkamm, U. Wehrenberg, and G. M. Rune Role of Tumor Necrosis Factor in Preovulatory Follicles of Swine Biol Reprod, September 1, 2001; 65(3): 928 - 935. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. L. Johnson, J. Murdoch, E. A. Van Kirk, J. E. Kaltenbach, and W. J. Murdoch Tumor Necrosis Factor {alpha} Regulates Collagenolytic Activity in Preovulatory Ovine Follicles: Relationship to Cytokine Secretion by the Oocyte-Cumulus Cell Complex Biol Reprod, December 1, 1999; 61(6): 1581 - 1585. [Abstract] [Full Text] |
||||
![]() |
W. J. Murdoch and S. A. Lund Prostaglandin-Independent Anovulatory Mechanism of Indomethacin Action: Inhibition of Tumor Necrosis Factor {alpha}-Induced Sheep Ovarian Cell Apoptosis Biol Reprod, December 1, 1999; 61(6): 1655 - 1659. [Abstract] [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |