Preeclamptic Decidual Microvascular Endothelial Cells Express Lower Levels of Matrix Metalloproteinase-1 Than Normals

Microvascular Research 57, 340 –346 (1999) Article ID mvre.1998.2142, available online at http://www.idealibrary.com on

Preeclamptic Decidual Microvascular Endothelial Cells Express Lower Levels of Matrix Metalloproteinase-1 Than Normals Eileen D. M. Gallery,* ,† Suzanne Campbell,* Jacky Arkell,‡ Minh Nguyen,‡ and Christopher J. Jackson‡ *Department of Renal Medicine, †Department of Obstetrics and Gynaecology, and ‡Sutton Arthritis Research Laboratory, University of Sydney at Royal North Shore Hospital, St. Leonards, NSW 2065, Australia Received December 15, 1998

In preeclampsia, invasion of intrauterine decidual blood vessels by placental cytotrophoblasts is significantly reduced. This study examined the secretion of matrix metalloproteinases (MMP) by cultured human decidual endothelial cells from normal (NDEC) and preeclamptic (PEDEC) pregnancies. MMPs secreted into the culture medium were measured using zymography, Western blotting, and ELISA. Results were confirmed by Northern analysis. Phorbol myristate acetate, known to induce protease activity in other endothelial cell populations, stimulated MMP1, MMP9, and TIMP1 secretion in both NDEC and PEDEC. Neither tumor necrosis factor-a nor transforming growth factor-b, both thought to have significant roles in the control of placentation, affected MMP secretion. MMP9 and TIMP1 levels were similar between the two cell types; however, MMP1 secretion was markedly different between the cell types. NDEC expressed higher levels of MMP1 under both basal (160 6 32 ng/10 6 cells) and stimulated (275 6 50) conditions compared to PEDEC (32 6 24 and 70 6 53, respectively). The lower MMP1 expression of decidual endothelial cells from preeclamptic women may inhibit endovascular invasion by cytotrophoblasts. These findings may, at least partly, explain the relative failure of trophoblasts to invade maternal decidual blood vessels in preeclamptic pregnancy. © 1999 Academic Press

340

Key Words: preeclampsia; decidual endothelial cells; matrix metalloproteinase; tissue inhibitor of matrix metalloproteinase.

INTRODUCTION Human placentation is unique, with more extensive decidual endovascular trophoblast invasion than in any other animal species, including nonhuman primates (Ramsey et al., 1976). Preeclampsia is a relatively common disorder peculiar to human pregnancy. It is characterized by impaired placental perfusion and fetal growth and a maternal syndrome of hypertension, vasoconstriction, and multiorgan dysfunction which results from generalized endothelial cell dysfunction and activation of the clotting system. The occurrence of endothelial injury throughout the decidua in preeclampsia, and of widespread maternal endothelial dysfunction (Shanklin and Sibai, 1989), implicates the endothelial cell in development and progression of preeclampsia. The earliest described lesion in women destined to develop preeclampsia is a relative failure of trophoblasts to invade maternal decidual vessels (Shanklin and Sibai, 1989). The ability of trophoblasts to adhere 0026-2862/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

341

MMP-1 Expression by Decidual Endothelium

to and proteolytically digest an extracellular matrix in vitro has been extensively studied (Fisher et al., 1989; Kliman and Feinberg, 1990; Lim et al., 1997). It is known that matrix metalloproteinase (MMP) secretion by trophoblasts is reduced in preeclampsia (Lim et al., 1997), but little is known of the contribution of the host decidual endothelium to these aspects of placentation in the human. Evaluation of the secretion of MMPs and TIMP by human decidual endothelium is of potential value in elucidating its role in controlling the extent of trophoblast invasion. The MMPs are a group of matrix-degrading enzymes which together can degrade all the components of the extracellular matrix (Woessner, 1991; Matrisian, 1992). They are secreted by many cell types including endothelial cells and trophoblasts. Type I collagenase (MMP1) can cleave the a chains of fibrillar collagens, which is present in endometrial interstitium. The gelatinases MMP2 (72 kDa) and MMP9 (92 kDa) degrade denatured collagen (gelatin) and the collagen present in basement membrane. The regulation of MMP secretion is complex and tightly controlled at several levels of expression. In addition, they can be inhibited by the specific tissue inhibitors of metalloproteinases (TIMPs). The secretion of TIMP1 by endothelial cells provides tight control of MMP activity surrounding blood vessels (Shingu et al., 1993). TIMPs are detectable in human placenta and also in decidua (Damsky et al., 1993; Huppertz et al., 1998), but it is unknown whether they are of endothelial cell origin. The vasculature is composed of heterogenous populations of endothelial cells, and there are functional differences between those from large and small vessels and among different microvascular beds. We have shown that there are differences between the MMP and TIMP1 secretory profiles of microvascular (neonatal foreskin) and macrovascular (umbilical vein) endothelial cells (Jackson and Nguyen, 1997). In addition, we have recently demonstrated that microvascular endothelial cells derived from the joints of patients with rheumatoid arthritis secrete decreased levels of TIMP1 compared to normals (Jackson et al., 1998). We have now developed methodology which allows us to examine for the first time secretion of MMPs by the microvascular endothelial cells of human decidua. This study was undertaken to investigate the hypothesis that there is relatively lower

expression of MMPs and TIMP1 in decidual endothelial cells from patients with preeclampsia than in those from normal pregnancies.

MATERIALS AND METHODS Subjects Pregnant women undergoing delivery by primary cesarean section at Royal North Shore Hospital participated in the study. Informed consent was obtained from those involved. Patients were excluded if they were in labor at the time of delivery or if they had ruptured membranes prior to cesarean section. Preeclampsia was defined as hypertension (sitting BP . 140/90 mm Hg), proteinuria (24-h urinary excretion .500 mg), and hyperuricaemia (s. uric acid . 0.35 mmol/L) occurring de novo after 20 weeks amenorrhea (Brown et al., 1993) with a return to normal postpartum. A woman was classified as normal if her BP never rose above 130/80 mm Hg during the index pregnancy and she had no proteinuria. The median duration of pregnancy in women with preeclampsia was 37 weeks, and that of normal women 38 weeks.

Sample Collection Decidual tissue. Biopsies (0.5–2.0 g) were collected from the uterus at the time of surgery by the attending obstetrician from both normotensive and preeclamptic women, and transported to the laboratory in sterile Hanks’ balanced salt solution (Trace Scientific Pty Ltd, Castle Hill, NSW, Australia). Foreskin tissue. Control populations of microvascular endothelial cells were prepared from foreskin tissue collected following routine infant circumcision.

Endothelial Cell Isolation and Culture Decidual endothelial cells (DEC) and foreskin endothelial cells (FSEC) were isolated by enzyme digestion followed by selective binding to Ulex Europaeus Icoated beads as previously published (Gallery et al., 1991; Jackson et al., 1990). These cells demonstrated

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

342

positive staining for von Willebrand factor which is typical of endothelium (data not shown). DEC were subsequently cultured on gelatin-coated plastic, in Medium 199 (ICN Biochemicals Ltd, Thame, Oxfordshire, UK) with added endothelial cell growth supplement (50 mg/ml, prepared as described by Maciag et al. (1979)), heparin (100 mg/ml, Sigma Chemical Co., St. Louis, MO), and 40% pooled maternal serum. No antibiotics were added to the medium. FSEC were grown in Biorich medium (ICN Biomedicals, Aurora, OH) containing 30% normal pooled human serum (derived from healthy volunteers) plus 100 mg/ml ECGS and 50 mg/ml heparin. Cells were maintained in a humidified 95% air:5% CO 2 mixture.

Experimental Protocol Cells (9 3 10 4/well, in 400 mL of medium) were grown to confluence, maintained in growth medium for 48 h and then underwent a 4-h quiescing period in basal medium. Incubations were then performed for 24 h under basal conditions (Medium 199 plus 4% human pregnancy serum stripped of MMPs), and the responses to phorbol myristate acetate (PMA, 100 ng/ ml), tumor necrosis factor-a (TNF, 50 ng/ml), and transforming growth factor-b (TGF, 10 ng/ml) were assessed. PMA was used as it is known to upregulate gene expression of MMPs. At the end of incubations, medium was collected and stored at 270°C until assay for MMPs and TIMP1 using zymography, Western blotting, or ELISA. To standardize among wells the number of cells was approximated by the method of Oliver et al. (1989).

Zymography MMP2 and MMP9 were measured using zymography as previously described (Herron et al., 1986). The gels were scanned into an IBM PC and the intensity of the bands was semiquantitated using Molecular Analyst software (Bio-Rad Laboratories, Hercules, CA).

Immunoblotting MMP1 and TIMP1 were detected by Western immunoblot analysis. Antibodies were purchased from Oncogene Science (Uniondale, NY) and used at 1 mg/ml.

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Gallery et al.

ELISA ELISA for MMP1 and TIMP1 were performed using commercially available Biotrak kits (Amersham). Results for group data are shown as means 6 SEM. Differences between groups were analyzed by ANOVA, while within group changes were assessed by means of the two-tailed paired Student’s t test.

Northern Blotting Endothelial cells were grown to confluence in 75cm 2 flasks and treated in basal medium with PMA or no test agent for 24 h. RNA extraction and Northern analysis were performed as previously described (Jackson and Nguyen, 1997). Probe DNA for MMP1 and TIMP1 was generously provided by Professor Lyn Matrisian (Vanderbilt University, Nashville, TN) and Kate Gibbons (Sutton Laboratory, Royal North Shore Hospital, Sydney, Australia), respectively.

RESULTS MMP Secretion of FSEC versus NDEC MMP2 and MMP9. MMP secretion was initially compared between FSEC and DEC from normal pregnancies (NDEC). Under basal conditions, both NDEC and FSEC secreted substantial amounts of the constitutively expressed MMP2, as detected by zymography (Fig. 1). PMA did not increase the total amount of MMP2 secretion but did activate the enzyme in both cell populations. There was minimal secretion of MMP9 under basal conditions by either cell type. In response to PMA, there was marked induction of MMP9, with NDEC secreting more abundant amounts than FSEC. This was consistent across three cell lines. Scanning densitometry of zymograms showed a mean 3.6-fold increase in MMP9 secretion by NDEC compared to FSEC. In the presence of EDTA, the bands representing both MMP2 and MMP9 disappeared indicating that these enzymes are MMPs (data not shown).

MMP-1 Expression by Decidual Endothelium

FIG. 1. Gelatin zymography of NDEC- and FSEC-conditioned medium. Confluent endothelial monolayers of NDEC or FSEC were preincubated in basal medium for 4 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA or no test agent (basal). The conditioned medium was collected and zymography was performed as described under Materials and Methods. These experiments were performed on three different FSEC and NDEC cell lines, each with similar results.

MMP1. Zymography also showed that both FSEC and NDEC secreted MMP activity with a M r of approximately 44 kDa (Fig. 1). This band, which is likely to represent MMP1 (Jackson and Nguyen, 1997), was more prominent in medium harvested from NDEC than from FSEC. Western analysis confirmed that NDEC secreted elevated levels of MMP1 compared to FSEC both under basal conditions and in response to PMA (Fig. 2). Similar results were obtained for three cell lines each of FSEC and NDEC.

MMP Secretion of NDEC versus PEDEC MMP2 and MMP9. We next examined the secretion of MMPs between NDEC and PEDEC. Zymography results for four cell lines of each were semiquantitated using scanning densitometry and results shown in Fig. 3. Under basal conditions, both cell types secreted MMP2. Although two of the four NDEC cell lines tested secreted substantially more MMP2 than PEDEC (data not shown), when the results for all four cell lines were pooled there was no significant difference between NDEC and PEDEC (Fig. 3, top). No stimulation of MMP2 was observed in response to PMA, TNF, or TGF. Results for MMP9 are shown in Fig. 3, bottom. Low but detectable levels of MMP9 were secreted under basal conditions. PMA stimulated MMP9 secretion in both NDEC and PEDEC, with no difference between the two cell types.

343

Neither TNF or TGF had any effect of the secretion of MMP9. MMP1. The band of MMP activity at 44 kDa, detected by zymography in the culture medium of NDEC, corresponding to MMP1 (Fig. 1), was not seen in media from PEDEC (data not shown). Western immunoblotting confirmed that PEDEC secreted low levels of MMP1 compared to NDEC (Fig. 4). The amount of MMP1 secreted was quantitated by ELISA in four cell lines from each clinical group (Fig. 5). Under basal conditions, PEDEC secreted significantly lower levels of MMP1 (28 6 23 ng/10 6 cells, mean 6 SEM) than did NDEC (157 6 29 ng/10 6 cells; P , 0.01). PMA significantly stimulated MMP1 synthesis in both cell types, compared to basal conditions (P , 0.05), the final values for PEDEC (68 6 54) being substantially lower than those for NDEC (263 6 54; P , 0.01). Northern analysis also revealed a corresponding increase in MMP1 mRNA by NDEC compared to PEDEC (Fig. 6). Neither TNF nor TGF had any effect on MMP1 secretion. TIMP1. Results for TIMP1, quantitated by ELISA in four cell lines of both NDEC and PEDEC, are shown in Fig. 5b. Under basal conditions both decidual cell types secreted measurable and equivalent levels of TIMP1. PMA increased TIMP1 secretion by both cell types to the same extent while neither TNF nor TGF had any demonstrable effect. Northern blotting showed that mRNA expression correlated with TIMP1 protein secretion (Fig. 6). Since the overall activity of MMPs results from a balance between the amount of MMPs and TIMPs present

FIG. 2. Western analysis of NDEC- and FSEC-conditioned medium for MMP1. Confluent endothelial monolayers of FSEC or NDEC were preincubated in basal medium for 4 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA or no test agent (basal). MMP1 antigen was measured in the conditioned medium using an antibody to MMP1.

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

344

FIG. 3. Scanning densitometry of zymography using conditioned medium from NDEC and PEDEC. Confluent endothelial monolayers of NDEC or PEDEC were preincubated in basal medium for 4 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA, 50 ng/ml TNF, 10 ng/ml TGF, or no test agent (basal). The conditioned medium was collected and zymography performed as described under Materials and Methods. These zymograms were scanned and quantitated as described under Materials and Methods. The results are expressed as the mean 6 standard deviation of three experiments using three different cell lines each of NDEC (black bars) and PEDEC (hatched bars).

we calculated the ratio of MMP1:TIMP1. The ratio, in both stimulated and unstimulated cells, was more than five times higher in NDEC (.10) than PEDEC (,2), suggesting that NDEC have a markedly higher capacity to promote matrix degradation than PEDEC.

Gallery et al.

PMA is a potent angiogenic factor capable of inducing protease activity in endothelial cells. The results of the present study have shown that PMA upregulates the secretion of MMP1, MMP9, and TIMP1 in decidual endothelial cells. Stimulation of MMP secretion was accompanied by upregulation of its mRNA. The cytokines TNF and TGF are both present in deciduum (Saksela and Jaatela, 1989; Lysiak et al., 1995) and play a significant role during trophoblast invasion and placentation. In the current study, TNF did not alter the secretion of MMPs in decidual endothelial cells. In agreement, previous reports (Hanemaaijer et al., 1993; Jackson and Nguyen, 1997) have also shown that TNF has minimal effect on the expression of MMPs and/or TIMP1 by human neonatal foreskin or umbilical vein endothelial cells. TGF has been shown to have variable effects on MMP expression depending on the cell type studied (Agarwal et al., 1994; Uria et al., 1998). We found no demonstrable effect of TGF on secretion of MMPs or TIMP1 by decidual endothelial cells. Our findings suggest that neither TNF nor TGF plays a role in matrix degradation by decidual endothelial cells. Trophoblast invasion of the intrauterine decidual vessels is an essential step in human placentation. Endogenous production and activation of MMP9 is a prerequisite for the inherent invasiveness of trophoblasts (Librach et al., 1991). Although studies of in vivo expression of MMPs by trophoblasts have given conflicting results. Huppertz et al. (1998) have reported that much of the variability may be due to differences

DISCUSSION We have shown for the first time that microvascular endothelial cells derived from normal human deciduum secrete MMPs. Their secretion of MMP1 and MMP9 was higher than another microvascular endothelial cell population, derived from neonatal foreskin. This supports and extends previous findings of diversity among endothelium derived from different vascular beds (Jackson and Nguyen, 1997). It also emphasises the importance of examining endothelial cells from the decidual vascular bed when studying related disorders.

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

FIG. 4. Western analysis of MMP1 and TIMP1 for NDEC and PEDEC. Confluent endothelial monolayers of NDEC or PEDEC were preincubated in basal medium for 4 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA, 50 ng/ml TNF, 10 ng/ml TGF, or no test agent (basal). MMP1 and TIMP1 antigens were measured in the conditioned medium using specific antibodies.

MMP-1 Expression by Decidual Endothelium

345

FIG. 5. ELISA of MMP1 and TIMP1 for NDEC and PEDEC. Confluent endothelial monolayers of NDEC (black bars) or PEDEC (hatched bars) were preincubated in basal medium for 6 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA, 50 ng/ml TNF, 10 ng/ml TGF, or no test agent (basal). Conditioned media were collected and assayed for (a) MMP1 and (b) TIMP1 using specific ELISA kits. Results are expressed as means 6 SEM of four different experiments using four different cell lines of each. Significance levels are shown for differences between NDEC and PEDEC (*P , 0.05, **P , 0.01).

trophoblast invasion of the extracellular matrix in preeclampsia. Whether there is an inherent abnormality in the ability of PEDEC to produce MMP1 is unclear. We have previously shown that there are differences in MMP secretion between macrovascular and microvascular endothelial cells (Jackson and Nguyen, 1997). It is unlikely that the vessel size accounts for the differences observed in the current study between NDEC and PEDEC, since both cell types are derived from similar tissue using the same isolation procedure. It is feasible that reduced MMP1 secretion by PEDEC is an acquired functional abnormality of these cells. The underlying mechanism of decreased MMP1 secretion by PEDEC is the subject of current investigation. In summary, we have described a unique MMP secretory profile by human decidual endothelial cells. The increased MMP activity of endothelial cells from normal women may facilitate, while those of preeclamptic women may inhibit, endovascular invasion by cytotrophoblasts. These findings may, at least, partly explain the relative failure of trophoblasts to invade maternal decidual blood vessels in preeclamptic pregnancy.

in methods of tissue handling prior to immunohistochemistry by light microscopy. They reported the presence of a wide repertoire of MMPs and TIMPs in the extracellular matrix surrounding invasive trophoblasts in the term placenta. The high levels of gene expression and secretion of matrix degrading enzymes by the normal decidual host endothelium which we have described in this work may be of central importance in control of trophoblast invasion. Our results clearly indicate that PEDEC secrete markedly lower levels of MMP1 than NDEC. Furthermore, although MMP1 secretion by PEDEC was elevated in response to PMA, levels remained consistently lower than those of NDEC. There was also a trend toward lower MMP2 secretion by PEDEC, although values were more variable. The particularly low ratio of MMP1:TIMP1 secreted by PEDEC supports the concept that NDEC have a greater capacity to promote matrix degradation and thus facilitate neighboring trophoblast invasion. The described abnormalities are likely to result in the inhibition of

FIG. 6. Northern blot analysis of MMP1 and TIMP1 mRNA expression by NDEC and PEDEC. Confluent endothelial monolayers of NDEC or PEDEC were preincubated in basal medium for 4 h, followed by incubation for 24 h in fresh basal medium in the presence of 100 ng/ml PMA or no test agent (basal). mRNA was extracted and subjected to Northern analysis. Ribosomal RNA verified equal sample loading of RNA (data not shown).

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

346

ACKNOWLEDGMENTS Our particular thanks are due to Sr Mei Ling Chan and the Obstetricians at Royal North Shore Hospital for collection of decidual samples. This work was supported by grants from the Australian Kidney Foundation, Wenkart Foundation, Northern Sydney Area Health Service, Baxter Extramural Grant Program, and the National Health and Medical Research Council of Australia.

REFERENCES Agarwal, C., Hembree, J. R., Rorke, E. A., and Eckert, R. L. (1994). Transforming growth factor beta-1 regulation of metalloproteinase production in cultured human cervical epithelial cells. Cancer Res. 54, 943–949. Brown, M. A., Gallery, E. D. M., Gatt, S. P., Leslie, G., and Robinson, J. (1993). Management of hypertension in pregnancy: Summary of consensus statement of Australasian Society for the Study of Hypertension in Pregnancy. Med. J. Australia 158, 700 –702. Damsky, C., Sutherland, A., and Fisher, S. (1993). Extracellular matrix 5: Adhesive interactins in early mammalian embryogenesis, implantation, and placentation. FASEB J. 7, 1320 –1329. Fisher, S. J., Cui, T., Zhang, L., Hartman, L., Grahl, K., Zhang, G. Y., Tarpey, J., and Damsky, C. H. (1989). Adhesive and degradative properties of human placental cytotrophoblast cells in vitro. J. Cell. Biol. 109, 891–902. Gallery, E. D. M., Rowe, J., Schrieber, L., and Jackson, C. J. (1991). Isolation and purification of microvascular endothelium from human decidual tissue in the late phase of pregnancy. Am. J. Obstet. Gynecol. 165, 191–196. Hanemaaijer, R., Koolwijk, P., Leclercq, L., Devree, W. J. A., and Vanhinsbergh, V. W. M. (1993). Regulation of matrix metalloproteinase expression in human vein and microvascular endothelial cells—Effects of tumour necrosis factor-alpha, interleukin-1 and phorbol ester. Biochem. J. 296, 803– 809. Herron, G. S., Werb, Z., Dwyer, K., and Banda, M. J. (1986). Secretion of metalloproteinases by stimulated capillary endothelial cells. I. Production of procollagenase and prostromelysin exceeds expression of proteolytic activity. J. Biol. Chem. 261, 2810 –2813. Huppertz, B., Kertschanska, S., Demir, A. Y., Frank, H. G., and Kaufmann, P. (1998). Immunohistochemistry of matrix metalloproteinases (MMP), their substrates, and their inhibitors (TIMP) during trophoblast invasion in the human placenta. Cell Tiss. Res. 291, 133–148. Jackson, C. J., Arkell, J., and Nguyen, M. (1998). Rheumatoid synovial endothelial cells secrete decreased levels of tissue inhibitor of MMP (TIMP1). Ann. Rheum. Dis. 57, 158 –161. Jackson, C. J., Garbett, P. K., Nissen, B., and Schrieber, L. (1990). Binding of human endothelium to Ulex Europaeus I-coated dyna-

Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Gallery et al.

beads: Application to the isolation of microvascular endothelium. J. Cell Sci. 96, 257–262. Jackson, C. J., and Nguyen, M. (1997). Human microvascular endothelial cells differ from macrovascular endothelial cells in their expression of matrix metalloproteinases. Int. J. Biochem. Cell Biol. 29, 1167–1177. Kliman, H. J., and Feinberg, R. F. (1990). Human trophoblast– extracellular matrix (ECM) interactions in vitro: ECM thickness modulates morphology and proteolytic activity. Proc. Natl. Acad. Sci. USA 87, 3057–3061. Librach, C. L., Werb, Z., Fitzgerald, M. L., Chiu, K., Corwin, N. M., Esteves, R. A., Grobelny, D., Galardy, R., Damsky, C. H., and Fisher, S. J. (1991). 92kD type IV collagenase mediates invasion of human cytotrophoblasts. J. Cell. Biol. 113, 437– 449. Lim, K. H., Zhou, Y., Janatpour, M., Mcmaster, M., Bass, K., Chun, S. H., and Fisher, S. J. (1997). Human cytotrophoblast differentiation/invasion is abnormal in pre-eclampsia. Am. J. Pathol. 151, 1809 –1818. Lysiak, J. J., Hunt, J., Pringle, G. A., and Lala, P. K. (1995). Localization of transforming growth factor beta and its natural inhibitor decorin in the human placenta and decidua throughout gestation. Placenta 16, 221–231. Maciag, T., Cerundolo, J., Ilsley, S., Kelley, P. R., and Forand, R. (1979). An endothelial cell growth factor from bovine hypothalamus: Identification and partial characterization. Proc. Natl. Acad. Sci. USA 76, 5674 –5678. Matrisian, L. M. (1992). The matrix-degrading metalloproteinases. BioEssays 14, 455– 463. Oliver, M. H., Harrison, N. K., Bishop, J. E., Cole, P. J., and Laurent, G. J. (1989). A rapid and convenient assay for counting cells cultured in microwell plates: Application for assessment of growth factors. J. Cell Sci. 92, 513–518. Ramsey, E. M., Houston, M. L., and Harris, J. N. S. (1976). Interaction of the trophoblast and maternal tissue in three closely related primate species. Am. J. Obstet. Gynecol. 124, 647– 652. Saksela, E., and Jaatela, M. (1989). Tumor necrosis factor in the human fetoplacentary unit. Int. J. Dev. Biol. 33, 13–16. Shanklin, D. R., and Sibai, B. M. (1989). Ultrastructural aspects of pre-eclampsia. 1. Placental bed and uterine boundary vessels. Am. J. Obstet. Gynecol. 161, 735–741. Shingu, M., Nagai, Y., Isayama, T., Naono, T., and Nobunaga, M. (1993). The effects of cytokines on metalloproteinase inhibitors (TIMP) and collagenase production by human chondrocytes and TIMP production by synovial cells and endothelial cells. Clin. Exp. Immunol. 94, 145–149. Uria, J. A., Jimenez, M. G., Balbin, M., Freije, J. M. P., and Lopezotin, C. (1998). Differential effects of transforming growth factor-beta on the expression of collagenase-1 and collagenase-3 in human fibroblasts. J. Biol. Chem. 273, 9769 –9777. Woessner, J. F. J. (1991). Matrix metalloproteinases and their inhibitors in connective tissue remodeling. FASEB J. 5, 2145–2154.