| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Center for Research on Reproduction and Women's Health and Departments of Obstetrics and Gynecology and
b Pathology and Laboratory Medicine, University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104
c Department of Medicine, Beth Israel Hospital and Harvard Medical School, Boston, Massachusetts 02215
d Bristol Heart Institute, Bristol Royal Infirmary, Bristol BS2 8HW, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1(IV) and 2(IV) chains. Amnion basement membrane collagens were more sensitive to degradation by collagenases than were adult rat kidney basement membrane collagens, which are enriched in 3(IV), 4(IV), and 6(IV) chains. Amnion type IV collagen content per unit of protein was markedly reduced by Day 21 of pregnancy, the day of delivery. Increased amnion levels of matrix metalloproteinase (MMP)-2 and MMP-9, gelatinases that degrade type IV collagen, were found by Day 21, suggesting that collagen breakdown was responsible, in part, for the decline in amnion type IV collagen. Infection of organ cultures of Day 18 rat amnions with a recombinant adenovirus expressing MMP-9 (AdMMP-9) caused release of collagen fragments detected as hydroxyproline in the culture fluid, amnion cell detachment, and apoptosis. The AdMMP-9-induced apoptosis was prevented by the MMP inhibitor batimastat. These findings suggest that MMPs are implicated in anoikis and apoptotic death of amnion cells, and may be part of a complex program of fetal membrane remodeling that occurs before delivery.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Type IV collagen is primarily localized to basement membranes. In the fetal membranes, type IV collagen is found in the basal lamina beneath the amniotic epithelial cells [10, 11]. It is also interspersed among trophoblast cells in the human chorion. Type IV collagen protomers are composed of three alpha chains joined in a triple helix. Six different alpha chains, each derived from a distinct gene, have been identified [12]. There is marked tissue specificity of the type IV collagen chain repertoire, which appears to reflect functional roles of specific basement membranes. Beyond serving a structural role, type IV collagen is believed to influence cellular function. Notably, degradation of type IV collagen in the basement membrane causes loss of cell attachment (anoikis) and apoptosis of epithelial cells [13–17].
We carried out studies to characterize the type IV collagen composition and content of the rat amnion near the end of pregnancy and correlated changes in type IV collagen levels with expression of enzymes that degrade it. The hypothesis that induction of enzymes that degrade type IV collagen promotes amnion cell detachment, and apoptosis was tested using an in vitro model.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Timed pregnant Sprague-Dawley rats (Zivic-Miller, Portersville, PA) were killed on Days 18–21 of pregnancy. The animals usually deliver on the afternoon of Day 21. Amnions were collected as previously described [6]. Tissue extracts prepared in a buffer consisting of 50 mM Tris, 1% Nonidet NP-40, 0.1% deoxycholate, 0.1% SDS, 150 mM NaCl, 10 µg/ml aprotinin, and 1 mM EDTA were used for Western blot analysis as previously described [5, 8]. Extracts made in 50 mM Tris buffer containing 100 mM CaCl2 and 150 mM NaCl, pH 7.4, were used for zymography [9]. All experiments were performed on at least two separate occasions.
Antibodies
Polyclonal antibodies raised against the C-terminal noncollagenous globular domains (NC1 domains) of specific alpha chains have been described previously [18]. Polyclonal antibodies against MMP-2 (Ab17) [19] and MMP-9 [20] were generously provided by William Stetler-Stevenson (National Institutes of Health) and Rafi Fridman (Wayne State University, Detroit, MI), respectively.
Western Blotting
Western blot analyses were carried out as previously reported [5, 8]. Briefly, tissue extracts (50–80 µg protein/lane) were separated by SDS PAGE, and the separated proteins were then transferred to membranes and probed with rabbit polyclonal antibodies. Antigen-antibody complexes bound to the membranes were detected using the enhanced chemiluminescence (ECL) reagent system (Amersham, Piscataway, NJ).
In Vitro Basement Membrane Degradation Assays
Basement membranes were prepared from rat amnions and maternal kidney as previously described [18]. The basement membrane extracts (1 mg dry weight) were incubated with bacterial collagenase (Worthington, Freehold, NJ; 18.5 U/mg) or active MMP-9 enzyme (Calbiochem, San Diego, CA; 1 µg of enzyme/mg of detergent-extracted basement membrane) over a 2-h period or 5-h period, respectively, at 37°C [18]. Hydroxyproline in peptides released during the digestion was quantitated as previously reported [18].
Infection of Rat Amnion Organ Cultures with Recombinant Adenovirus
Amnions were collected on Day 18 of pregnancy, washed free of amniotic fluid in PBS, and placed into organ culture in 1 ml of serum-free Kennett's HY medium (Gibco BRL Life Technologies, Gaithersburg, MD) containing 1% BSA in 12-well plastic dishes (2 or 3 amnions/well; within each experiment the number of amnions per well was equal in all treatment groups). Organ cultures were incubated at 37°C under an atmosphere of 5% CO2:95% humidified air. The amnions were infected with either recombinant adenovirus expressing human proMMP-9 (AdMMP-9) constructed as previously reported [21], or with a control recombinant adenovirus containing the LacZ cDNA, at 5.8 x 108 plaque-forming units (pfu)/well immediately after being placed into culture. This concentration of recombinant adenovirus was selected on the basis of previous experience with transduction of cultured cells [21] and experiments in which rat amnions were infected with lower concentrations of virus. In some experiments, the MMP inhibitor batimastat (BB-94: 4-[N-hydroxylamino]-2R-isobutyl-3S-[thienylthiomethyl]-succinyl-L-phenylalanine-N-methylamide; British Biotech, Oxford, UK) was added to the culture medium at a concentration of 3 µM in dimethylsulfoxide (DMSO). Control cultures received the DMSO vehicle. Thirty hours after infection, the conditioned media were collected and centrifuged before zymographic analysis for MMP activity [9] and hydroxyproline assay as modified from our previously reported method [18].
For the hydroxyproline assay, 50 µl of the conditioned medium was dried in plastic tubes in an oven at 100°C. Fifty microliters of 4 N NaOH was added, and the tubes were then placed into boiling water for 90 min. Fifty microliters of 1.4 N citric acid was then added to bring the pH to 6.0. One milliliter of chloramine-T solution was then added, and the tubes were incubated at room temperature for 20 min. One milliliter of aldehyde/perchloric acid solution was then added, and the tubes were incubated at 65°C. After 15 min, absorbance at 550 nm was determined. The absorbance values of the starting culture medium were subtracted from those of the conditioned media. Samples were assayed in triplicate.
The amnion tissue was processed for detection of ß-galactosidase activity [22] and fixed in formalin for histological examination and analysis of nuclear DNA fragmentation in apoptotic cells.
The recombinant adenoviruses used in these studies were free of wild-type adenovirus on the basis of titration studies carried out in nonpermissive HeLa cells and the absence of E1a in infected cells as assessed by immunofluorescence using an anti-E1a antibody.
Histological Quantitation of Apoptosis
The method of Wijsman et al. [23] was employed to detect nuclear DNA fragmentation in formalin-fixed, paraffin-embedded tissue sections. This immunohistochemical procedure identifies nuclear 3'-end labeled DNA fragments. We have previously reported that apoptosis detected by this method is directly correlated with intranucleosomal DNA cleavage determined by formation of 180- to 200-basepair (bp) DNA ladders [6]. The percentage of apoptotic cells in each specimen was assessed by determining the percentage of dark brown-stained nuclei in 10 random fields viewed at x400, amounting to the scoring of approximately 1000 nuclei/specimen.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1(IV) and 2(IV) Predominated in Rat Amnion
Extracts of amnions collected on Day 19 of pregnancy were treated with bacterial collagenase and then subjected to SDS PAGE to resolve the NC1 domains. The proteins were transferred to membranes, which were subsequently probed with polyclonal antibodies recognizing specific NC1 sequences in the 1(IV)/2(IV), 3(IV), 4(IV), 5(IV), and 6(IV) chains. These Western blot analyses demonstrated that 1(IV)/2(IV) NC1 domain dimers and monomers were most abundant in rat amnion extracts (Fig. 1). Some 3(IV) was detected, whereas 4(IV), 5(IV), and 6(IV) chains were present at negligible levels. The same antibodies detected 1(IV)/2(IV), 3(IV), 4(IV), 5(IV), and 6(IV) chains in rat kidney extracts (Fig. 1).
|
Rat Amnion Basement Membrane Extracts Were More Susceptible to Collagenase Digestion than Rat Kidney Basement Membrane Extracts
In vitro degradation assays indicated that collagens in the rat amnion basement membrane extract were degraded at a higher rate (Fig. 2) than the collagens in rat kidney basement membranes, which are enriched in 3(IV), 4(IV), and 6(IV) chains [18]. Similar observations were also made when purified activated human MMP-9 was used for basement membrane degradation (Fig. 2), with the exception that bacterial collagenase degraded the basement membrane extracts at a faster rate than did MMP-9.
|
Loss of Type IV Collagen from Rat Amnions Before Labor and Increased MMP Expression
Western blot analyses of extracts of rat amnions collected on Days 18–21 of pregnancy using an antibody recognizing 1(IV)/2(IV) chains of type IV collagen demonstrated a marked reduction in the content of type IV collagens per unit protein by Day 21 (Fig. 3). Immunoreactive bands of 180 kDa and 120 kDa predominated on Days 18 and 19, with some NC1 dimer evident at 45 kDa. By Day 20, these immunoreactive bands were reduced in intensity, and they were markedly depleted by Day 21 when approximately 75% of the immunoreactive type IV collagen had been depleted as assessed by densitometric analysis of Western blots. Brilliant Blue G staining of the polyacrylamide gels demonstrated that equal amounts of amnion extract protein had been loaded in each lane but that the relative abundance of some of the protein components changed between Day 18 and Day 21 (Fig. 3B).
|
Western blot analysis of extracts for the gelatinases MMP-2 and MMP-9 demonstrated increased expression of these enzymes by Day 20, with substantial increases on Day 21 (Fig. 3C). In agreement with our previous zymographic studies [9], MMP-2 (66-kDa lysis band) and MMP-9 (92- and 180-kDa lysis bands) activities were increased on Day 21 compared to Day 19 (Fig. 3D). A lysis band at 68 kDa, possibly reflecting the activated form of MMP-9 [24], was also detected on Day 21. A faint lysis band at 31 kDa was found consistently in amnion samples collected on Days 18 and 19 but was absent on Day 21.
Infection of Rat Amnion Organ Cultures with a Recombinant Adenovirus Expressing proMMP-9 Promoted Collagenolysis and Amnion Cell Apoptosis
Altering expression of MMP-9 results in profound changes in extracellular matrix catabolism and cellular characteristics, including cell death [16, 17]. Transfection of rat embryo cells with an MMP-9 expression vector produced cells with a metastatic phenotype, a reflection of increased capacity to degrade extracellular matrix components [25]. Conversely, inhibition of MMP-9 expression in transformed cells that constitutively produce MMP-9 by using a ribozyme suppressed metastasis without affecting tumorigenesis [26].
To test the hypothesis that induction of a type IV collagen-degrading activity causes collagen degradation, cell detachment, and subsequent cell death, we infected organ cultures of rat amnion collected on Day 18 with recombinant adenovirus expressing proMMP-9 (AdMMP-9) or with a control virus (AdLacZ) expressing LacZ (5.2 x 108 pfu/well). Thirty hours after infection, the medium surrounding amnions infected with AdMMP-9 contained a prominent proMMP-9 signal detected as a 92-kDa lysis band by zymography and an 83-kDa band beneath it, reflecting processed proenzyme (Fig. 4), results that are consistent with transduction of cultured cells with this vector [21]. MMP-2 activity was present in the media from cultures of both AdLacZ- and AdMMP-9-infected amnions at approximately equivalent levels. Amnions infected with AdLacZ stained uniformly for ß-galactosidase activity, reflecting efficient gene transfer, whereas amnions infected with AdMMP-9 did not stain for the presence of ß-galactosidase.
|
The AdMMP-9-infected amnions were characterized by a ragged appearance and detachment of cells from the membranes, histological features that are found in amnions collected on Days 20 and 21 of pregnancy [2, 6] (Fig. 5). AdLacZ-infected amnions were relatively intact, with normal cellularity and without evidence of cytopathic effect. Evidence for increased collagen breakdown in AdMMP-9-infected amnions was obtained from measurement of hydroxyproline in the conditioned medium. Medium from cultures of amnions infected with AdMMP-9 contained greater than 4-fold more hydroxyproline than did medium from cultures of AdLacZ-infected amnions (p < 0.01 by Student's t-test) (Fig. 4).
|
AdMMP-9 infection resulted in a marked increase in apoptosis, assessed by immunohistochemical staining for nuclear DNA fragmentation (Fig. 4). In the AdMMP-9-infected tissues, 6.3 ± 1.1% of the cells were undergoing apoptosis (mean ± SD, n = 3 cultures per group) in the experiment shown in Figure 5, a finding that is similar to what we have observed in amnions collected on Day 21 of pregnancy [6]. In contrast, in the AdLacZ-infected amnions, 0.4 ± 0.2% of the cells were apoptotic (significantly different from cells in AdMMP-9-infected amnions; p < 0.01 by Student's t-test).
The effects of AdMMP-9 infection on amnion structure and amnion cell apoptosis were blocked by the MMP inhibitor, batimastat. Batimastat treatment retained amnion membrane integrity, reduced the release of hydroxyproline-containing peptides into the culture fluid, and completely prevented amnion cell apoptosis (Fig. 5).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
On the basis of the dramatic changes in MMP-9 expression, we elected to focus our attention on this enzyme. The emphasis of this study on MMP-9 should not be construed as a dismissal of a role for MMP-2 in amnion extracellular matrix catabolism before parturition. The increase in MMP-2 activity and protein in the amnion on Day 21 suggests that this enzyme could also participate in type IV collagen breakdown. Vu et al. [27] recently targeted the murine MMP-9 gene. The nullizygous mutants demonstrated an apparent defect in apoptosis in the skeletal growth plate, but no apparent reproductive abnormalities, although the fetal membranes were evidently not examined. Moreover, deficiency of MMP-2 does not appear to cause developmental abnormalities [28]. However, these two enzymes may be able to compensate for each other in processes involved in amnion matrix degradation.
Our studies demonstrating differential sensitivity of kidney and amnion basement membranes to collagen breakdown by active MMP-9 and bacterial collagenase suggest that the rat amnion type IV collagen chain composition potentially favors collagenase degradation, as compared to the type IV collagen composition of adult renal tubular basement membranes. The relative abundance of cysteine-enriched type IV collagen chains in renal basement membranes may impart greater strength to the basement membrane as a result of cross-linking through disulfide bridges [12, 18]. Hence, it is conceivable that type IV collagen composition determines the longevity and integrity of the basement membrane, with basement membranes that are relatively poor in the more cross-linked chains, like amnion, being destined for turnover. However, it must be recognized that our experiments were carried out on extracted basement membrane material and that our findings may not reflect the relative susceptibilities of the collagens to degradation in situ.
Degradation of the type IV collagen in amnion basement membranes may initiate the process of apoptotic cell death that we have previously described [2, 6]. Indeed, ultrastructural studies of rat amnions documented loss of basement membrane components and delamination of amnion epithelial cells in association with features of cell death [2]. The present studies show loss of type IV collagen by Day 20. Consistent with this view, our in vitro experiments demonstrated that increasing MMP-9 expression promotes collagen breakdown and apoptosis in rat amnions in organ culture. It should be noted that cells remaining anchored to the amnion, in addition to those that had separated from their substrates, were undergoing apoptosis as evidenced by nuclear DNA fragmentation. Hence, apoptosis appears to be initiated before complete separation of the amnion cells from their substrate, perhaps as a result of partial detachment or changes in cell shape due to matrix degradation.
The MMP inhibitor batimastat prevented the amnion cell apoptosis induced by AdMMP-9. This inhibitor is not specific for MMP-9, so we cannot conclude that it's effects were solely on MMP-9 [29, 30]. It is possible that other MMPs are activated by AdMMP-9 infection and that these enzymes participate in the induction of apoptosis.
We corroborated the rat amnion organ culture studies by using cultures of human amnion cells (WISH cells) grown in serum-free medium on a combined matrix of type I and type IV collagen. Transfection of the WISH cells with an MMP-9 expression plasmid resulted in cell detachment and apoptosis, whereas transfection of the cells with an empty vector had no effect on cell attachment and cell death (unpublished observations).
The simplest interpretation of our in vitro studies is that increased expression of the MMP-9 proenzyme leads to the formation of active enzyme, which, in turn, degrades type IV collagen, leading to amnion cell apoptosis. Alternative interpretations include the possibility that the increased proMMP-9 sequestered endogenous tissue inhibitors of metalloproteinases, leading to increased activities of endogenous MMPs. This alternative explanation still invokes a role for MMPs in the induction of apoptosis and could be part of the physiological process of fetal membrane breakdown in that proMMP-9 levels increase markedly before active labor in the rat amnion.
Apoptosis is now recognized to play an important role in fetal and maternal tissues near the end of pregnancy [6, 31–34]. However, the factors that provoke programmed cell death in preparation for labor and after delivery remain to be clarified. Our observations suggest that increased expression of MMPs initiates apoptosis by catalyzing degradation of the basement membranes or pericellular matrix, leading to cell detachment or change in cell shape and subsequently cell demise. It should be noted that MMP activation is not the only mechanism by which apoptosis can be induced in involuting tissues [17]. In addition to catabolic processes, altered production of matrix proteins by amnion cells [35–37] and changes in levels of cytokines or growth factors may also promote cell death [38]. Furthermore, detachment of cells from the amnion is not an immediate death sentence since viable amniocytes can be collected from amniotic fluid, and limited digestion of amnion tissue with bacterial collagenase disperses cells that can be maintained in culture. Thus, the state of the amnion cells, the process leading to their separation from the matrix, or the duration of the separation may be key determinants of cell fate.
Although the biochemical and structural changes that we have described in the rat amnion parallel to some degree changes that have been found in the human amnion [3, 8], there are important differences between the rat and human membranes. For example, the rat amnion undergoes striking changes before labor that essentially encompass the whole amnion. In contrast, structural and biochemical changes in human fetal membranes appear to be restricted to specific regions near the rupture site. Unlike the human fetal membranes, the rat amnion is not adherent to a chorion laeve, so the nature of cellular communications in the rat and human fetal membranes could be substantially different. These and other differences must be considered by investigators attempting to extrapolate from studies conducted in rodents to humans.
In summary, the findings presented in this report suggest that the sequence of events involved in the terminal changes in amnion structure before onset of labor are initiated by increased MMP activity, leading to collagenolysis and, subsequently, to apoptotic death of amnion cells. This sequence predicts that detection of increased MMP activity could presage fetal membrane rupture.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Jerome F. Strauss, III, 778 Clinical Research Building, 415 Curie Boulevard, Philadelphia, PA 19104. FAX: 215 573 5408; jfs3{at}mail.med.upenn.edu
Accepted: August 27, 1998.
Received: June 16, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. S. LeBleu, B. MacDonald, and R. Kalluri Structure and Function of Basement Membranes Experimental Biology and Medicine, October 1, 2007; 232(9): 1121 - 1129. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Khwad, V. Pandey, B. Stetzer, B. M. Mercer, D. Kumar, R. M. Moore, J. Fox, R. W. Redline, J. M. Mansour, and J. J. Moore Fetal Membranes From Term VAginal Deliveries Have a Zone of Weakness Exhibiting Characteristics of Apoptosis and Remodeling Reproductive Sciences, April 1, 2006; 13(3): 191 - 195. [Abstract] [PDF]
|
|
|
|
 |
|
 M. El Khwad, B. Stetzer, R.M. Moore, D. Kumar, B. Mercer, S. Arikat, R.W. Redline, J.M. Mansour, and J.J. Moore Term Human Fetal Membranes Have a Weak Zone Overlying the Lower Uterine Pole and Cervix Before Onset of Labor Biol Reprod, March 1, 2005; 72(3): 720 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Moore, D. W. Lundgren, R. J. Silver, and J. J. Moore Lactosylceramide-Induced Apoptosis in Primary Amnion Cells and Amnion-Derived WISH Cells Reproductive Sciences, September 1, 2002; 9(5): 282 - 289. [Abstract] [PDF]
|
|
|
|
 |
|
 P. E. Ferrand, S. Parry, M. Sammel, G. A. Macones, H. Kuivaniemi, R. Romero, and J. F. Strauss III A polymorphism in the matrix metalloproteinase-9 promoter is associated with increased risk of preterm premature rupture of membranes in African Americans Mol. Hum. Reprod., May 1, 2002; 8(5): 494 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Ulug, S. Goldman, I. Ben-Shlomo, and E. Shalev Matrix metalloproteinase (MMP)-2 and MMP-9 and their inhibitor, TIMP-1, in human term decidua and fetal membranes: the effect of prostaglandin F2{alpha} and indomethacin Mol. Hum. Reprod., December 1, 2001; 7(12): 1187 - 1193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Alhasan, O. Aranha, and F. H. Sarkar Genistein Elicits Pleiotropic Molecular Effects on Head and Neck Cancer Cells Clin. Cancer Res., December 1, 2001; 7(12): 4174 - 4181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kumagai, Y. Otsuki, Y. Ito, M-A. Shibata, H. Abe, and M. Ueki Apoptosis in the normal human amnion at term, independent of Bcl-2 regulation and onset of labour Mol. Hum. Reprod., July 1, 2001; 7(7): 681 - 689. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Allard, P Adin, L Gouedard, N di Clemente, N Josso, M. Orgebin-Crist, J. Picard, and F Xavier Molecular mechanisms of hormone-mediated Mullerian duct regression: involvement of beta-catenin Development, January 8, 2000; 127(15): 3349 - 3360. [Abstract] [PDF]
|
|
|
|
 |
|
 J. McLaren, D.J. Taylor, and S.C. Bell Increased incidence of apoptosis in non-labour-affected cytotrophoblast cells in term fetal membranes overlying the cervix Hum. Reprod., November 1, 1999; 14(11): 2895 - 2900. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Lei, E. E. Furth, R. Kalluri, P. Wakenell, C. B. Kallen, J. J. Jeffrey, P. S. Leboy, and J. F. Strauss III Induction of Matrix Metalloproteinases and Collagenolysis in Chick Embryonic Membranes before Hatching Biol Reprod, January 1, 1999; 60(1): 183 - 189. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
