Fibrin Enhances Differentiation, but not Apoptosis, and Limits Hypoxic Injury of Cultured Term Human Trophoblasts

Placenta (2005), 26, 491e497 doi:10.1016/j.placenta.2004.08.011

Fibrin Enhances Differentiation, but not Apoptosis, and Limits Hypoxic Injury of Cultured Term Human Trophoblasts R. G. Humphreya, S. D. Smitha, L. Panga, Y. Sadovskya,b and D. M. Nelsona,* a

Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO 63110, USA; Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, MO 63110, USA Paper accepted 30 August 2004 b

We hypothesized that fibrin enhances apoptosis and modulates differentiation of trophoblast in vitro. Cytotrophoblasts isolated from normal term human placentas were cultured %72 h in DMEMe10%-FBS on a fibrin matrix in standard or hypoxic conditions. Trophoblasts were cultured on plastic (control), type I collagen (matrix control), or dishes with fibrinogen, fibrin degradation products (FDP), thrombin, plasma fibronectin or cellular fibronectin. Apoptosis was determined by western analysis of the cleavage products of poly-ADP-ribose polymerase and cytokeratin 18 and caspase 3 activity. Cell cycle regulation was quantified by expression of proliferating cell nuclear antigen (PCNA) and p27 protein. Differentiation was determined by media level of hCG and hPL. Compared to the two controls, fibrin matrix had no effect on trophoblast apoptosis or total cell death in standard conditions. Neither fibrin nor collagen altered expression of PCNA or p27. In contrast, fibrin significantly increased the secretion of both hCG and hPL. Fibrin, but not FDP, thrombin or fibronectins, promoted hormonal differentiation. Fibrin limited the impact of a %8 h of hypoxia on trophoblast hormone release but did not avert the effects of 24 h of low oxygen and did not alter apoptosis in hypoxic trophoblast. We conclude that fibrin provides an environment conducive for trophoblast reepithelialization of the surface of villi, where injury is marked by fibrin deposition. Placenta (2005), 26, 491e497 Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Trophoblast; Fibrin; Apoptosis; Differentiation

INTRODUCTION The trophoblast on human placental villi is a unique epithelium that regulates maternalefetal exchange of gases, nutrients, and waste products. Villous trophoblast near term consists of a discontinuous layer of mitotically active cytotrophoblasts that differentiate into syncytiotrophoblast by membrane fusion. This process is accompanied by hormonal differentiation, with increased synthesis of the hormones human chorionic gonadotropin (hCG) and human placental lactogen (hPL). Denudations develop in the trophoblast epithelium lining the intervillous space. Consequently, exposure of the trophoblast basement membrane may stimulate thrombolytic cleavage of fibrinogen, with deposition of a matrix at the discontinuities [1]. The matrix is rich in fibrin, and therefore termed fibrincontaining fibrinoid [2], and is characteristic of injury and * Corresponding author. Department of Obstetrics and Gynecology, Washington University School of Medicine, Campus Box 8064, 4566 Scott Avenue, St. Louis, MO 63110, USA. Tel.: C1 314 747 0739; fax: C1 314 362 8580. E-mail address: [email protected] (D.M. Nelson). 0143e4004/$esee front matter

repair in the villous trophoblast [1,3]. This histopathological feature is distinguished from matrix-type fibrinoid in the basal plate, which contains abundant collagens, laminins and other extracellular components but little fibrin [2]. Importantly, fibrin-containing fibrinoid is prevalent in placentas from pregnancies complicated by fetal growth restriction or preeclampsia [4], where trophoblast apoptosis is also common [5,6]. The association of apoptotic nuclei within fibrincontaining fibrinoid deposits [7] and the re-epithelialization of the matrix by trophoblast [1] suggest that fibrin may influence trophoblast biology. Here we tested the hypothesis that fibrin enhances apoptosis and modulates differentiation of trophoblast from term human placentas. MATERIALS AND METHODS Cell isolation and culture This study was approved by the Institutional Review Board of Washington University School of Medicine. Cytotrophoblasts were isolated from normal term human placentas as described by Kliman et al. [8] with modifications. Elimination of fragments of syncytia was performed as described by Guilbert et al. [9], resulting in 95e99% of the cells immunoreactive for Ó 2004 Elsevier Ltd. All rights reserved.

Placenta (2005), Vol. 26

492

cytokeratin 7 [10]. Culture medium was DMEM with 10% fetal bovine serum (FBS; HyClone, Logan, UT), 20 mmol/L HEPES (pH 7.4), 0.5 mmol/L L-glutamine (Sigma, St. Louis, MO), penicillin (10 U/mL), streptomycin (10 mg/mL), and Fungizone (0.25 mg/ml). Time zero for all experiments was defined as 4 h after plating, to allow for cell adherence. Cultures were maintained in standard conditions at 37 (C in a 5% CO2-air atmosphere in an incubator (Nuaire, Plymouth, MN). Media were collected daily and stored at ÿ20 (C until assayed. Cells were plated at 3.5 ! 105 cells/cm2 grown on fibrin and compared with those grown on serum coated plastic (control) or type I collagen (matrix control) [11] which is not characteristic of fibrin-containing fibrinoid [2]. Wells were coated with fibrin as described [12], with modifications. Human stock fibrinogen (92% clottable, lot # F-4883; Sigma) was reconstituted to 7.5 mM in HBSS with HEPES. Human stock thrombin (Sigma) was reconstituted at 40 NIH units/ mL in HBSS with HEPES. The stock solutions were passed through 0.2 mm filters and stored at ÿ20 (C. The fibrin matrix was formed by adding 10 mM CaCl2 to 1 mL fibrinogen stock, mixed, incubated 5 min on ice, and 0.4 NIH units/mL thrombin was added immediately prior to pipetting in 1 mL increments to cover the bottom of 3.5 cm diameter, non-tissue culture dishes (Falcon). Collagen-coated wells were made with rat-tail type 1 collagen (BD Biosciences, Bedford, MA) as described by the manufacturer. The collagen solution was stored on ice, and 1 mL aliquots were added to non-tissue culture dishes (Falcon). After overnight incubation at 37 (C for both, the fibrin clots and collagen gels were removed with forceps, the adherent fibrin and collagen were incubated in 1.5 mL DMEM with 10% FBS for 1 h at 37 (C, and the wells were rinsed with DMEM prior to seeding as above for primary culture. Control plastic culture dishes (Falcon) were incubated in 1.5 ml DMEM with 10% FBS for 1 h at 37 (C without cells, rinsed with DMEM, and seeded identically for primary culture. The studies of the individual matrix components were done in 1 cm diameter tissue culture dishes (Falcon) coated with each of the following. The 92% clottable human fibrinogen (Sigma) in DMEM was diluted to 2.9 mM, and for fibrinogen, 0.5 mL of this solution was layered directly onto the dish. For fibrin coating 1 mL of the fibrinogen solution was mixed with 0.05 NIH units of thrombin stock (Sigma; 10 NIH U/mL) diluted in sterile HBSS with HEPES (pH Z 7.4) and layered as above. Fibrin degradation products (FDP) were made by adding 0.01 U/mL plasmin (from a stock solution of 1 U/mL in HBSS with HEPES, pH Z 7.4) to the fibrin solution for 20 min at 37 (C. After a gentle mix the FDP containing solution was pipetted in 0.5 ml increments to the 1 cm wells. Purified fibrin matrix was made from 99.1% clottable human fibrinogen (American Diagnostica, Samford, CT), reconstituted according to the manufacturer’s protocol, diluted in DMEM to 2.9 mM, mixed with thrombin as above, and plated in 0.5 mL increments in 1 cm wells. All matrices were incubated overnight; the clot was removed as above and

rinsed with fresh DMEMe10% FBS before primary culture. For exposure to thrombin and the fibronectins DMEMe10% FBS was supplemented with 100 nM thrombin, 0.4 mg/mL plasma fibronectin, or 15 mg/mL cellular fibronectin. Exposure to hypoxia These experiments were conducted in a Forma 1025 glove-box type anaerobic chamber (Thermo Electron Corp, Marietta, OH). The partial pressure of oxygen was recorded at !1%. An attached interchange chamber ensured that the atmosphere of the main chamber remained hypoxic. Media inside the main chamber was purged of dissolved oxygen by bubbling with anaerobic (nitrogen) gas for 5 min/100 mL, which based on our determination reduced the pO2 of the media to %15 mmHg for culture. Western analysis Cells were washed once with cold PBS, total cell lysates prepared [12], and western analysis performed as described [13], with 30 mg of protein per lane. After transfer the blot was blocked for 1 h with skim milk and incubated in 5% milk and 0.1e0.5% TBS-T with monoclonal antibodies to the cleavage products of cytokeratin 18 (cyt 18; mouse anti-human; 1:1000; Roche), the cleavage products of poly-adenosine diphosphate ribose polymerase (PARP; rabbit anti-human; 1:1000; Promega), proliferation cell nuclear antigen (PCNA; mouse antihuman; 1:1000; Santa Cruz), or p27 (mouse anti-human; 1:100; Santa Cruz). Gels were washed 10 min !2 in Trisbuffered saline solution with Tween 20, and incubated for 1 h with horseradish peroxidase-conjugated secondary antibodies from the appropriate species (1:1000, Santa Cruz). Membranes were also probed for b-actin (goat; 1:400; Santa Cruz). Densitometric analyses, which included normalization to b-actin, were performed using a densitometer and ImageQuant version 3.3 software (Molecular Dynamics, Inc., Sunnyvale, CA) or the BiochemiSystem photo imager and Labworks 4.0 analysis software (UVP, Inc., Upland, CA). Caspase 3 activity assay Cells were lysed by scraping into buffer and caspase 3 activity was assessed in 30 mg of lysate as previously described [14]. Substrate cleavage was measured in a FluoroCount microplate fluorometer (Packard, Meridan, CT) by quantifying fluorescence that resulted from proteolytic cleavage of a synthetic peptide linked to 7-amino-4-methylcoumarin (Ac-DEVDAMC; Anaspec, San Jose, CA). Background fluorescence was subtracted, and data were expressed as relative fluorescent units (cRFU), normalized to protein. Lactate dehydrogenase (LDH) activity LDH activity in the media was quantified in triplicate by an enzyme immunoassay kit, per the manufacturer’s protocol (Roche, Basel, Switzerland). A positive control for the LDH assay was cultures exposed to 1 mM staurosporine, 10 ng/ml

Humphrey et al.: Fibrin Modulates Trophoblast Differentiation

DNA determination Cells were fixed in -20 C methanol for 20 min, washed !2 with PBS and incubated in 100 ng/ml 4#,6-diamidino-2phenyl-indole (DAPI; Sigma, St. Louis) for 5 min. After two washes with PBS fluorescence was quantified on a FluoroCount microplate fluorometer (Packard, Meridan, CT).

200

Collagen

Fibrin

Control

Collagen

Fibrin

100

Control

Media were assayed for hCG and hPL in duplicate using enzyme immunoassays as described by the manufacturer (DRG International, Mountainside, NJ). Inter- and intraassay variabilities for hCG were 6.6% and 3.4%, and for hPL were 5.6% and 6.7%, respectively. Values represent hCG and hPL secreted per 24 h interval, normalized to total cellular protein or DNA.

(% of control)

Hormone analyses

PARP Expression

TNF-a and 25 mg/ml etoposide, designated the death cocktail. The LDH activity of media from each paradigm was thereby expressed as the percentage of the maximal cell death induced by the death cocktail at the comparable time in culture.

493

PARP β−Actin 24 h 72 h Figure 1. The effect of fibrin on PARP expression in primary term trophoblasts. The upper panel is the densitometric analysis (mean G SEM) of PARP expression (n Z 5), analyzed by western immunoblotting and normalized to b-actin, for cells grown on control, fibrin matrix, or the matrix control. A representative western blot is shown in the lower panel. There were no significant differences in PARP expression among the paradigms of growth substrates.

Statistical analysis

200

Collagen

Fibrin

Control

Collagen

Fibrin

100

Control

We first examined the influence of fibrin matrix on apoptosis in trophoblasts cultured for 24 h in standard conditions, when the cytotrophoblast phenotype predominates, and after 72 h in culture, when syncytiotrophoblasts are prominent [15]. We examined the expression of cleavage products of PARP and cyt 18, both reliable and complementary markers of apoptosis in the nucleus and cytoplasm, respectively. There was no difference in the expression of either cleaved PARP (Figure 1) or cyt 18 (Figure 2) in cells grown on fibrin matrix, compared to cells grown on either control substrate. As expected [13] the level of apoptosis was higher in cultures with a dominance of cytotrophoblasts at 24 h, compared to the more differentiated cells at 72 h, independent of growth substrate (Figures 1 and 2). To exclude the possibility that fibrin matrix might increase non-apoptotic cell death, we assayed for release of LDH, a nonspecific marker of cell death. Again, there was no difference in the level of cell death among the three growth substrates when each was compared to the percentage of maximal cell death induced by the death cocktail at each time, as described in Materials and methods. The level of global cell death at 24 h (plastic: 57 G 10%; fibrin: 54 G 9%; collagen: 52 G 9%) was

(% of control)

RESULTS

higher in all culture paradigms compared to the cells at 72 h (plastic: 8 G 1%; fibrin: 6 G 1%; collagen: 9 G 2%). Taken together, these studies demonstrate that under standard culture conditions cell death was not influenced by the growth substrate. Because re-epithelialization of fibrin deposits at villous surface discontinuities in vivo may require cytotrophoblast proliferation, we wondered if fibrin altered the expression of cell cycle regulators in cultured trophoblasts. Comparing fibrin matrix to controls, we found no differences among the

CK-18 Expression

Data are expressed as mean G SD or mean G SEM, when more than two paradigms were compared among experiments. Comparisons were made using t test and ANOVA with post hoc Bonferroni correction used for multiple comparisons. A P ! 0.05 was considered significant. The number of primary cultures for each set of experiments is listed in the figure legends.

CK-18 β−Actin 72h 24 h Figure 2. The effect of fibrin on cytokeratin 18 (cyt 18) expression in primary term trophoblasts. The upper panel is the densitometric analysis (mean G SEM) of cyt 18 expression (n Z 3), analyzed by western immunoblotting and normalized to b-actin, for trophoblasts cultured on control, fibrin matrix, or matrix control. A representative western blot is shown in the lower panel. There were no significant differences in the expression of cyt 18 among the paradigms of growth substrates.

Placenta (2005), Vol. 26

494

200

Collagen

Control

Fibrin

(% of control)

p27 Expression

100

Collagen

Fibrin

100

Control

(% of control)

PCNA Expression

200

p27 β−Actin

PCNA β−Actin

72 h 72 h Figure 3. The effect of fibrin on cell cycle regulators. Primary trophoblasts were cultured for 72 h on control, fibrin or type I collagen growth substrate and proteins were harvested for western analysis. The upper panels are the densitometric analyses (mean G SEM) of expression of PCNA (n Z 5) and p27 (n Z 3). Control is defined as 100% compared to fibrin and collagen. Representative western blots are shown in the lower panels. When normalized to b-actin, there were no significant differences in expression of either cell cycle regulator among the paradigms.

paradigms at either 24 h (data not shown) or 72 h (Figure 3) in the expression of either the S phase protein PCNA or the p27 protein that limits proliferation. We next assessed the effect of matrix on trophoblast differentiation. As we have previously shown [15], we found that under control conditions hCG levels increased by more than 15-fold in trophoblasts cultured in vitro from 24 h to 72 h. The level of hCG in medium from 24 h of culture was

B

300

*

C

24 h

72 h

200

Fi

br in

100

C

100

(% of control)

hPL

200

Fi br ol in la ge n Fi C brin ol la ge n

(% of control)

hCG

300

la ge n

*

ol

400

A

low as expected in all cultures, although cells on fibrin showed the highest hCG in medium compared to control cultures. Importantly, hCG in media of cells grown 72 h on fibrin was 3-fold higher than control cells on plastic (P ! 0.05; Figure 4a). hCG levels for trophoblasts on collagen were insignificantly higher than control and lower than fibrin. Similarly, we found that hPL release into medium from 48 h to 72 h was 2.2fold higher (P ! 0.05) on fibrin compared to control, but not significantly higher than the 2-fold higher level of hPL in medium for cultures on collagen (Figure 4b). Taken together, these data indicate that fibrin facilitates hormonal differentiation of trophoblasts, but fibrin matrix has no effect on constitutive apoptosis or cell cycle regulation, compared to the control substrates. We next questioned whether or not the enhanced hormonal differentiation of trophoblast cultured on fibrin matrix was due to signals from fibrin components or from other matrix constituents. To address this question, we measured hCG levels in medium from primary trophoblasts cultured in the presence of several constituents expected in fibrin matrices in vivo, as described in Materials and Methods and listed in Figure 5. In two of these experiments we also included cultures on a fibrin matrix made from purified fibrinogen (99% clottable fibrinogen, free of fibronectins). Cells grown on fibrin matrix, whether from regular or purified sources, exhibited the highest level of hCG release from 48 to 72 h, averaging 4.7fold and 9.3-fold higher than control (P ! 0.05), respectively. Cells grown on native fibrinogen, or in the presence of FDP, showed a slight increase in medium hCG levels compared to control. Importantly, medium hCG levels for trophoblasts cultured with thrombin, plasma fibronectin, or cellular

72 h

Figure 4. The effect of fibrin on trophoblast hormonal differentiation. Primary trophoblasts (n Z 7) were cultured on control, fibrin, or type I collagen growth substrates. The mean level of hCG (mIU/mg protein/24 h) in control cultures (set at 100%) was 3.1 mIU/mg protein at 24 h and 62.6 mIU/ mg protein at 72 h. Mean level of medium hPL in control cultures (n Z 4) during 48e72 h was 2.1 mIU/ml, and assigned a value of 100% in panel B. Values for fibrin and collagen are mean percentage (G SEM) of control. * Denotes P ! 0.05 compared to control.

Figure 5. The influence of matrix components on trophoblast hCG production. Primary trophoblasts were cultured on control, on regular fibrin matrix (fibrin), fibrin matrix made from purified fibrinogen (pure fibrin), fibrinogen, fibrin degradation products (FDP) created from action of plasmin on fibrin, or in media containing thrombin, plasma (p) fibronectin, or cellular (c) fibronectin, as described in Materials and methods. The media harvested after 48 h to 72 h in culture were assayed for hCG and normalized to DNA (mean G SD). The experiment shown is representative of results from four primary cultures for all matrix components, except pure fibrin (n Z 2). * Denotes P ! 0.01 compared to control.

Humphrey et al.: Fibrin Modulates Trophoblast Differentiation

495

Control Fibrin 150 (% of control)

Caspase 3 activity

fibronectin were not significantly different from control. The same qualitative pattern was observed for hPL, with hPL in medium from cultures on regular fibrin (2.8-fold) and purified fibrin (3.2-fold) higher than control (P ! 0.05). Again, there was no significant difference in hPL levels for cells grown with thrombin or either fibronectin compared to control. These data indicate that fibrin molecules, but not other components within the fibrin matrix, promote hormonal differentiation of cultured trophoblasts. Fibrin-containing fibrinoid deposits are prevalent on villi of placentas in clinical conditions associated with hypo perfusion and villous hypoxia [4]. We wondered if a fibrin matrix would limit the adverse effects of hypoxia on cultured trophoblasts, previously demonstrated by us [15] and others [16,17]. Primary cultures of trophoblasts on control or regular fibrin growth substrates were exposed to progressive intervals of hypoxia between 48 and 72 h of culture as shown in Figure 6. We found that growth on a fibrin matrix diminished the effect of hypoxia on trophoblast differentiation, reflected by the significantly higher media hCG levels (P ! 0.05), in cells exposed to 4 h or 8 h of low oxygen during the 48e72 h culture (Figure 6). However, fibrin had no significant effect on the hypoxia-induced reduction of hCG in cultures after 24 h of low oxygen. Interestingly, the degree of apoptosis, reflected by caspase 3 activity, was not reduced at these time points for cells grown on fibrin compared to control (Figure 7). These

100 50

4

8

24

Hours of Hypoxia Figure 7. The effect of fibrin matrix on hypoxia induced apoptosis in trophoblast. Primary trophoblasts (n Z 4) were cultured under standard conditions for 48 h prior to exposure to hypoxia as illustrated in the lower panel of Figure 6. Caspase 3 activity was determined by fluorometric analysis, and the level of apoptosis (mean G SEM) in cultures grown on fibrin and exposed to hypoxia was expressed as a percentage of the level of caspase 3 activity in control cultures, similarly exposed to hypoxia. None of the differences were significant.

findings suggest that fibrin matrix attenuates the impact of a few hours of hypoxia on trophoblast hCG production, but cannot abrogate the effects of more prolonged hypoxic intervals.

DISCUSSION

*

Control

hCG

(mIUx102/µg DNA)

30

Fibrin

* 20

* 10

0

4

8

24

Hours of Hypoxia (shaded)

Hours of Hypoxia 4 8 24 24

48

72

Time in culture Figure 6. Fibrin matrix modulates the effect of hypoxia on trophoblast hCG release. Primary trophoblasts were cultured under standard conditions for 48 h on control or fibrin growth substrates, and then exposed to hypoxia for the time intervals indicated in the lower panel of the figure. Media were collected at 72 h and assayed for hCG (mean G SD). The figure depicts a representative experiment from four primary cultures. *Denotes P ! 0.05 compared to control at the comparable time point in culture.

The data show that under standard conditions fibrin matrix does not alter basal levels of trophoblast apoptosis or global cell death and that fibrin has no effect on the expression of two cell cycle regulators in trophoblasts. Importantly, fibrin promotes hormonal differentiation of cultured trophoblasts. This effect is attributed to the fibrin components within the matrix and is not due to an epitope unique to fibrinogen or FDP. Although present in our fibrin matrix, thrombin, cellular fibronectin and plasma fibronectin have no influence on differentiation of trophoblasts in the absence of fibrin. Furthermore, growth on a fibrin matrix diminishes the effect of hypoxia on hormonal differentiation of trophoblasts cultured up to 72 h. These data suggest that fibrin promotes repair of villous denudations in vivo. Fibrin is a well-known constituent of placentas from uncomplicated pregnancies, and fibrin deposits are more common in many [4], though not all [18], pregnancies complicated by placental dysfunction. Originally called fibrinoid necrosis [19], villous fibrinoid deposits are enriched in fibrin, and thus were re-named fibrin-containing fibrinoid [2]. Ultrastructural studies of term placental villi suggest that injury to the trophoblast layer exposes the basement membrane at a syncytiotrophoblast discontinuity [1]. Syncytiotrophoblast also undergoes apoptosis [20], possibly as a part of normal epithelial turnover, and enhanced trophoblast apoptosis associates with pregnancy complications [4e6].

496

Apoptosis could thereby create denudations in the trophoblast layer. Extensive morphometric analyses [3,20e24] have clarified that fibrin-type fibrinoid is preferentially located at sites of de-epithelialization and demonstrate that fibrin-type fibrinoid deposition is more prevalent in some complicated pregnancies [23,25]. However, other pregnancies complicated by placental dysfunction exhibit less villous fibrin in the placenta [18]. This finding, attributed to altered fibrinolysis in affected pregnancies [18], underscores the fact that fibrin deposits result from interactions of pro- and anti-coagulant cascades that regulate fibrin formation and re-modeling within the placenta. Although fibrin matrix in vitro enhances morphological differentiation of cultured trophoblasts [26], trophoblasts on placental villi often exhibit ultrastructural features of apoptosis within and adjacent to fibrin-containing fibrinoid deposits [7]. The identification of apoptotic features of trophoblast in villous fibrinoid deposits by electron microscopy was recently complemented by immunohistochemical detection of cleavage products of cyt 18, a reliable marker of apoptosis, adjacent to fibrin-containing fibrinoid [27]. The multi-component nature of fibrin matrix on villi in vivo raised the question of whether fibrin components contributed to the observed apoptosis in trophoblast, the foundation for the hypothesis posed in our study. We established that neither fibrin nor other matrix components in our in vitro system affect the level of apoptosis in trophoblasts cultured under standard conditions.

Placenta (2005), Vol. 26

Fibrin formation from thrombin treated fibrinogen in vitro was designed to mimic conditions to which villous trophoblasts are exposed in vivo. Among all components that were tested, fibrin maximally enhanced the hormonal differentiation of trophoblasts. Key components that are known to have receptors in trophoblast, such as thrombin [28] and fibronectin [11,29] showed no difference from control. We chose type I collagen as our matrix control because this extracellular component is not present in fibrin-containing fibrinoid [2] and would not contribute to villous trophoblast differentiation in these deposits in vivo. We found that unlike fibrin, collagen as a matrix control enhanced but did not maximize hormonal differentiation of trophoblasts. Compared to other villous syncytiotrophoblasts, expression of the anti-apoptotic Bcl-2 protein is higher in trophoblast adjacent to fibrin deposits [30,31]. This protein may confer protection from apoptosis for trophoblasts as they reepithelialize discontinuities. We demonstrate that in the presence of hypoxia, fibrin matrix ameliorates the impact of hypoxia on trophoblast hormone production, at least for a few hours of hypoxic exposure but cannot abrogate the effects of 24 h of low oxygen. Whether fibrin modulates Bcl-2 expression is unknown, but the fibrin in matrix provides some protection from insults like hypoxia in vivo. Our data suggest that fibrin components provide an environment conducive for trophoblast survival and villous repair and thereby, assist to maintain the integrity of the trophoblast layer on villi.

ACKNOWLEDGMENTS We are grateful to the Society for Maternal Fetal Medicine for the generous scholarship award to RGH. We also thank the National Institutes of Health for grant support from HD 29190 (DMN) and ES11597 (YS).

REFERENCES [1] Nelson DM, Crouch EC, Curran EM, Farmer DR. Trophoblast interaction with fibrin matrix. Epithelialization of perivillous fibrin deposits as a mechanism for villous repair in the human placenta. Am J Pathol 1990;136:855e65. [2] Frank HG, Malekzadeh F, Kertschanska S, Crescimanno C, Castellucci M, Lang I, et al. Immunohistochemistry of two different types of placental fibrinoid. Acta Anat 1994;150:55e68. [3] Mayhew TM, Bowles C, Orme G. A stereological method for testing whether or not there is random deposition of perivillous fibrin-type fibrinoid at the villous surface: description and pilot applications to term placentae. Placenta 2000;21:684e92. [4] Kanfer A, Bruch JF, Nguyen G, He C, Delarue F, Flahault A, et al. Increased placental antifibrinolytic potential and fibrin deposits in pregnancy-induced hypertension and preeclampsia. Lab Invest 1996;74: 253e8. [5] Smith SC, Baker PN, Symonds EM. Increased placental apoptosis in intrauterine growth restriction. Am J Obstet Gynecol 1997;177: 1395e401. [6] Allaire AD, Ballenger KA, Wells SR, McMahon MJ, Lessey BA. Placental apoptosis in preeclampsia. Obstet Gynecol 2000 Aug;96:271e6. [7] Nelson DM. Apoptotic changes occur in syncytiotrophoblast of human placental villi where fibrin type fibrinoid is deposited at discontinuities in the villous trophoblast. Placenta 1996;17:387e91. [8] Kliman HJ, Nestler JE, Sermasi E, Sanger JM, Strauss 3rd JF. Purification, characterization, and in vitro differentiation of cytotrophoblasts from human term placentae. Endocrinology 1986;118:1567e82.

[9] Guilbert LJ, Winkler-Lowen B, Sherburne R, Rote NS, Li H, Morrish DW. Preparation and functional characterization of villous cytotrophoblasts free of syncytial fragments. Placenta 2002;23:175e83. [10] Kamudhamas A, Pang L, Smith SD, Sadovsky Y, Nelson DM. Homocysteine thiolactone induces apoptosis in cultured human trophoblasts; a mechanism for homocysteine-mediated placental dysfunction? Am J Obstet Gynecol 2004;191:563e71. [11] Kao LC, Caltabiano S, Wu S, Strauss 3rd JF, Kliman HJ. The human villous cytotrophoblast: interactions with extracellular matrix proteins, endocrine function, and cytoplasmic differentiation in the absence of syncytium formation. Dev Biol 1988;130:693e702. [12] Corbett SA, Wilson CL, Schwarzbauer JE. Changes in cell spreading and cytoskeletal organization are induced by adhesion to a fibronectin-fibrin matrix. Blood 1996;88:158e66. [13] Levy R, Smith SD, Chandler K, Sadovsky Y, Nelson DM. Apoptosis in human cultured trophoblasts is enhanced by hypoxia and diminished by epidermal growth factor. Am J Physiol Cell Physiol 2000;278:C982e8. [14] Yusuf K, Smith SD, Sadovsky Y, Nelson DM. Trophoblast differentiation modulates the activity of caspases in primary cultures of term human trophoblasts. Pediatr Res 2002;52:411e5. [15] Nelson DM, Johnson RD, Smith SD, Anteby EY, Sadovsky Y. Hypoxia limits differentiation and up-regulates expression and activity of prostaglandin H synthase 2 in cultured trophoblast from term human placenta. Am J Obstet Gynecol 1999;180:896e902. [16] Alsat E, Wyplosz P, Malassine A, Guibourdenche J, Porquet D, Nessmann C, et al. Hypoxia impairs cell fusion and differentiation process in human cytotrophoblast, in vitro. J Cell Physiol 1996;168:346e53.

Humphrey et al.: Fibrin Modulates Trophoblast Differentiation [17] Esterman A, Finlay TH, Dancis J. The effect of hypoxia on term trophoblast: hormone synthesis and release. Placenta 1996;17:217e22. [18] Mayhew TM, Bowles C, Yucel F. Hypobaric hypoxia and villous trophoblast: evidence that human pregnancy at high altitude (3600 m) perturbs epithelial turnover and coagulation-fibrinolysis in the intervillous space. Placenta 2002;23:154e62. [19] Fox H. Perivillous fibrin deposition in the human placenta. Am J Obstet Gynecol 1967;98:245e51. [20] Mayhew TM, Leach L, McGee R, Ismail WW, Myklebust R, Lammiman MJ. Proliferation, differentiation and apoptosis in villous trophoblast at 13e41 weeks of gestation (including observations on annulate lamellae and nuclear pore complexes). Placenta 1999;20:407e22. [21] Mayhew TM, Barker BL. Villous trophoblast: morphometric perspectives on growth, differentiation, turnover and deposition of fibrin-type fibrinoid during gestation. Placenta 2001;22:628e38. [22] Mayhew TM, Brotherton L, Holliday E, Orme G, Bush PG. Fibrin-type fibrinoid in placentae from pregnancies associated with maternal smoking: association with villous trophoblast and impact on intervillous porosity. Placenta 2003;24:501e9. [23] Mayhew TM, Sampson C. Maternal diabetes mellitus is associated with altered deposition of fibrin-type fibrinoid at the villous surface in term placenta. Placenta 2003;24:524e31. [24] Mayhew TM. Villous trophoblast of human placenta: a coherent view of its turnover, repair and contributions to villous development and maturation. Histol Histopathol 2001;16:1213e24.

497 [25] Van Horn JT, Craven C, Ward K, Branch DW, Silver RM. Histologic features of placentas and abortion specimens from women with antiphospholipid and antiphospholipid-like syndromes. Placenta 2004; 25:642e8. [26] Farmer DR, Nelson DM. A fibrin matrix modulates the proliferation, hormone secretion and morphologic differentiation of cultured human placental trophoblast. Placenta 1992;13: 163e77. [27] Austgulen R, Chedwick L, Vogt Isaksen C, Vatten L, Craven C. Trophoblast apoptosis in human placenta at term as detected by expression of a cytokeratin 18 degradation product of caspase. Arch Pathol Lab Med 2002;126:1480e6. [28] Lubega J. Thrombin receptor on the placental syncytiotrophoblastic plasma membrane. Ric Clin Lab 1990;20:203e8. [29] Burrows TD, King A, Loke YW. Expression of integrins by human trophoblast and differential adhesion to laminin or fibronectin. Hum Reprod 1993;8:475e84. [30] Marzioni D, Muhlhauser J, Crescimanno C, Banita M, Pierleoni C, Castellucci M. BCL-2 expression in the human placenta and its correlation with fibrin deposits. Hum Reprod 1998;13:1717e22. [31] Toki T, Horiuchi A, Ichikawa N, Mori A, Nikaido T, Fujii S. Inverse relationship between apoptosis and Bcl-2 expression in syncytiotrophoblast and fibrin-type fibrinoid in early gestation. Mol Hum Reprod 1999; 5:246e51.