Proteomic Identification of Caldesmon as a Physiological Substrate of

Sep 18, 2009 - Proteomic Identification of Caldesmon as a Physiological Substrate of Proprotein Convertase 6 in Human Uterine Decidual Cells Essential...
0 downloads 0 Views 5MB Size
Download PDF
Proteomic Identification of Caldesmon as a Physiological Substrate of Proprotein Convertase 6 in Human Uterine Decidual Cells Essential for Pregnancy Establishment Lynette M. Kilpatrick,§ Andrew N. Stephens,§ Belinda M. Hardman,§ Lois A. Salamonsen, Ying Li, Peter G. Stanton, and Guiying Nie* Prince Henry’s Institute of Medical Research, 246 Clayton Road, Clayton, Victoria, 3168, Australia Received April 30, 2009

Proprotein convertase 5/6 (PC6), a member of the proprotein convertase (PC) family, is a critical regulator in the uterus for embryo implantation. In particular, PC6 is essential for the differentiation of uterine stromal fibroblasts into decidual cells (decidualization). Knockdown of PC6 in the mouse uterus leads to complete failure of decidualization and implantation. It has been envisaged that PC6 functions by proteolytically activating a number of important growth factors and other precursor proteins. However, to date, the precise mechanism of PC6 action in the uterus, particularly during decidualization, is unknown. In this study, we aimed to investigate the mechanisms of PC6 action in decidualization by identifying its physiological substrates using a proteomic approach. Primary human endometrial stromal cells were decidualized and treated with or without recombinant human PC6 (rhPC6). The proteins cleaved by rhPC6 were identified by 2-dimensional fluorescent differential gel electrophoresis. The candidate proteins were validated as PC6 substrates by a number of approaches, including determining their coexpression with PC6 in vivo in decidual cells in the human uterus. A total of 18 protein spots were significantly altered by rhPC6 treatment, 8 of which were assigned clear identities by mass spectrometry. One of these was confirmed to be caldesmon, a key protein involved in organizing the actin microfilaments and regulating cytoskeletal structure. This study demonstrates a novel approach to identify PC-regulated proteins of physiological relevance, and provides important insight into the mechanism by which PC6 regulates decidualization. Keywords: Proprotein convertase • PC5/6 • decidualization • pregnancy • 2-dimensional gel electrophoresis

Introduction Implantation of the embryo into the uterus is critical for establishing a pregnancy. To facilitate implantation, in rodents and primates including women, the uterus must undergo considerable morphological and physiological changes, one of which is the differentiation of uterine stromal fibroblasts into phenotypically and functionally distinct decidual cells (decidualization). Defects in decidualization cause implantation failure and pregnancy loss.1-3 We have previously shown that proprotein convertase 5/6 (PC6) is a critical regulator of decidualization both in the mouse and human.4,5 PC6 is dramatically and transiently up-regulated in the mouse uterus, specifically at the site and time of embryo implantation in decidualizing cells.6,7 Knockdown of PC6 during early pregnancy in vivo in the mouse uterus results in complete failure of implantation due to decidualization arrest.4 PC6 is also up-regulated in the human and rhesus monkey uterus * To whom correspondence should be addressed. Guiying Nie, Prince Henry’s Institute of Medical Research, PO Box 5152, Victoria, 3168, Australia. Phone, Tel: +61 3 9594 4380; fax, +61 3 9594 6125; e-mail, E-Mail: [email protected]. § These authors contributed equally to this work. 10.1021/pr900381a CCC: $40.75

 2009 American Chemical Society

specifically in the decidual cells and glandular epithelium at embryo implantation.4 Blocking PC6 production in human endometrial stromal cells (HESCs) significantly inhibits decidualization.5 PC6 belongs to a family of seven structurally related serine proteases, consisting of furin, PC1/PC3, PC2, PC4, PC6, PACE4, and PC7/PC8. The PCs play critical roles in post-translational protein activation by converting precursor proteins into their bioactive forms through limited proteolysis at the general consensus motif (K/R)-(X) n-(K/R)V, where n ) 0, 2, 4, or 6 and X is any amino acid (aa).8,9 PCs are hence important regulatory molecules in generating a large number of tissue-specific and functionally important bioactive proteins by temporal and spatial activation of their pro-proteins. These include growth factors, peptide hormones, neuropeptides, ECM proteins, adhesion molecules, proteolytic enzymes, and integral membrane proteins.10 PCs are therefore regarded as critical “master switch” molecules and recognized as promising targets for therapeutic applications.10-12 Two isoforms of PC6, one soluble (PC6A) and the other membrane-bound (PC6B), due to alternative splicing of one single gene PCSK5, have been identified in mammals.13 PC6A Journal of Proteome Research 2009, 8, 4983–4992 4983 Published on Web 09/18/2009

research articles is packaged into dense core granules and/or constitutively secreted into the extracellular milieu, while PC6B is transported to the cell membrane.9,14 PC6A interacts with heparin sulfate proteoglycans (HSPGs), thereby recruiting itself to the cell surface and activating a number of HSPG-bound substrates at the cell surface or in the extracellular space.9,15 Our studies suggest that the PC6A is the predominant form of PC6 involved in decidualization.6 Furthermore, PC6 is confirmed to be the only PC family member that is significantly up-regulated during decidualization both in the mouse and human.16,17 PC6 has been reported to process a number of precursor proteins, including receptor protein tyrosine phosphatase-µ,18 lefty,16,19 MT1-MMP,20,21 platelet derived growth factor,22,23 HIV-1 envelope glycoprotein gp160,24,25 integrin subunits R4, R5, R6 and Rv,26,27 prohepcidin,28 and Gdf11.29 However, apart from lefty, the proteins and pathways regulated by PC6 during decidualization remain unknown. To understand the molecular mechanism of PC6 action in the uterus for embryo implantation, it is critical to determine the physiological substrates of this important regulatory enzyme during decidualization. In this study, a proteomic approach was used to identify the in vivo substrates of PC6 in decidual cells in the human uterus. The overall strategy is schematically illustrated in Figure 1. Briefly, freshly isolated primary human endometrial stromal cells were decidualized in culture, equal amounts of decidual cell lysate were treated with or without recombinant human PC6 (rhPC6), and the protein profiles were then compared between the rhPC6treated and nontreated cell lysates by two-dimensional fluorescent differential gel electrophoresis (2D-DIGE). The differentially presented protein spots on the gel, representing proteins that were cleaved by PC6, were uncovered and their identities determined. The identified proteins were validated as PC6 substrates in decidual cells using the following approaches: determining consensus PC cleavage motifs on their primary sequences, examining changes in molecular weight and band patterns following PC6 cleavage by Western analysis using epitope-specific antibodies, and confirming their in vivo expression in decidual cells in the human uterus. A number of proteins were shown to be cleaved by rhPC6 in decidual cell lysates, one of which was caldesmon, a well-known protein important for actin filament reorganization. One critical feature of decidualization is reorganization of the cytoskeleton; we thus focused on caldesmon for further validation. We established that caldesmon isoform 1 is a physiological substrate of PC6 and that it is coexpressed with PC6 specifically in decidual cells in vivo in the human uterus. This study provides important insight into the molecular mechanism of PC6 action during decidualization. It also demonstrates for the first time the potential of a proteomic approach to identify PC-regulated proteins and pathways in physiological contexts.

Materials and Methods The overall experimental procedures are schematically illustrated in Figure 1. Each procedure is described in detail below. Patients and Human Uterine Tissue Collection. Ethical approval was obtained for all tissue collections from the Human Ethics Committee at Monash Medical Centre, Melbourne, Australia. Written informed consent was obtained from all patients prior to tissue collection. Human endometrial biopsies were obtained from women undergoing curettage following laparoscopic sterilization or assessment of tubal patency. Patients with uterine abnormalities such as leiomyomas or endometriosis, or those who had received steroid hormone therapy in the last 6 months, were excluded. Samples from days 4984

Journal of Proteome Research • Vol. 8, No. 11, 2009

Kilpatrick et al.

Figure 1. Schematic illustration of the proteomic strategy to identify physiological substrates of PC6 in endometrial decidual cells. Equal amounts of decidual cell lysate of primary HESCs were treated with (+PC6) or without (-PC6) recombinant human PC6 (rhPC6), and their protein profiles compared by 2D-DIGE. The proteins cleaved (| ) cleavage site) by PC6 were uncovered and their identities determined.

8-24 of the menstrual cycle were collected into Dulbecco’s modified Eagle’s medium (DMEM) and processed within 24 h for cell isolation. Tissues of proliferative (days 9-14) and secretory (days 19-28) phases of the menstrual cycle were fixed in 10% buffered formalin overnight, washed 3 times in Trisbuffered saline (TBS, pH 7.6) and processed to paraffin blocks. Decidualization of Human Endometrial Stromal Cells (HESCs) with cAMP. HESCs were isolated as previously published.5 The wells were decidualized with 500 µM 8-bromocyclic AMP (cAMP, Sigma-Aldrich, Castle Hill, NSW, Australia) for 3 days.4 The success of decidualization was confirmed by measuring the decidual marker prolactin in the conditioned medium by ELISA (Bioclone Aust. Pty Ltd., NSW, Australia). The cells were lysed on ice with 1 mL of sample solubilization buffer 1 (SB1: 7 M urea, 2 M thiourea, 40 mM Tris, 1% (w/v) C7BzO) and stored at -80 °C until use. Treatment of HESC Lysates with Recombinant Human PC6. HESC lysates were thawed and cell debris was pelleted by brief centrifugation. The supernatant was concentrated to 10-20 µL using a NanoSep 3 kDa spin column (Pall Life Sciences, MI),

Caldesmon Is a PC6 Substrate in Decidual Cells and resuspended in 1 mL of a 1:1 mix of DMEM/HamsF12, pH 7.4. Protein lysates were divided into two equal aliquots (500 µL each); one aliquot was treated with and the other without 10 units (an estimated amount in the decidual media) of hPC6-A (PhenoSwitch BioSciences, Inc., Quebec, Canada) for 2 h at 37 °C. To ensure that hPC6-A was biologically active in each treatment, 100 µM of a universal PC peptide substrate, the fluorogenic peptide pERTKR-MCA (Bachem, PA), was included in each treatment, and the peptide hydrolysis was monitored every 60 s at excitation/ emission of 355/460 at 37 °C (Wallac, Victor 2 spectrophotometer, PerkinElmer, MA). After the treatment, proteins were concentrated (NanoSep 3 kDa spin column), resuspended in 1 mL of SB1 buffer, reduced with 25 µL/mL of tributylphosphine (Sigma), and alkylated with 10 µL/mL of acrylamide alkylating reagent (Proteome Systems, NSW, Australia), before the addition of DTT (10 µL/mL, Sigma) and citric acid solution (80 µL/mL, Proteome Systems). Samples were briefly centrifuged and the supernatants precipitated with 10 vol of acetone (room temperature, 90 min). Protein pellets were air-dried and resuspended in CyDye labeling buffer (50 µL, 30 mM Tris, 7 M Urea, 2 M Thiourea, 1% (w/v) C7BzO, pH 8.0). Differential Protein Expression Analysis. Fluorescent labeling using minimal CyDyes (GE Healthcare Bioscience, Uppsala, Sweden), isoelectric focusing (IEF), second-dimension polyacrylamide gel electrophoresis (PAGE) and image acquisition were carried out as described.30 In brief, 50 µg of proteins from samples treated with or without hPC6-A was labeled with 200 pmol of Cy3 or Cy5 on ice, in the dark for 30 min. In addition, 25 µg of each sample (treated with or without hPC6-A) was pooled and labeled with 200 pmol of Cy2 (GE Healthcare) for use as the internal standard. Reactions were quenched by the addition of 1 µL of 10 mM lysine to each sample. Upon the completion of labeling, the Cy3 and Cy5 labeled samples were combined pairwise together with the Cy2-labeled internal standard. Each sample was then precipitated with acetone as described above, and the air-dried pellet was resuspended in 450 µL of solubilization buffer 2 (SB2: 7 M urea, 2 M thiourea, 1% (w/v) C7BzO). Samples were snap-frozen on dry ice and stored at -80 °C until use, typically within 1 day of labeling. Isoelectric focusing was carried out in a pH 3-10 immobilized pH gradient according to the following parameters: 60 µA current per strip, 100 V/90 min, 300 V/90 min, 500 V/3 h, gradient to 1000 V/4 h and then 8000 V/3 h, and finally constant 8000 V/overnight until reaching 60 000 Vh. Gel alignment, spot detection and expression analysis were performed using Progenesis PG240 SameSpots (NonLinear Dynamics, Newcastle Upon Tyne, U.K.). Fold changes were calculated based on normalized spot volumes, and all statistical analysis was carried out automatically by the software. PAGE gels were then fixed and stained with Sypro Ruby (Invitrogen, San Diego, CA) according to the manufacturer’s instructions, counterstained using Coomassie G250 and then scanned as above using a 670 nm laser and far-red filterset (FujiFilm, Japan). The acquired image was matched to the existing analytical gel images using the Samespots program. Protein spots of interest were then excised from the gel using a ProPicII robotic spot picker (Genomic Solutions, MI), based on X-Y coordinates exported from PG240 SameSpots using the spot picking function. Protein Identification by Mass Spectrometry. Protein identification was performed as described.30 Monoisotopic peak masses were extracted using GPS Explorer software (Applied Biosystems, version 3.0 build 311), and peak lists were searched against the nonredundant Swiss-Prot database 56.6 (405 506

research articles entries; http://www.expasy.org) using the online MASCOT search engine (updated 06-December-2008; http://www. matrixscience.com). Carbonylamide-cysteine (CAM) and oxidation of methionine residues were taken into account, allowing mass tolerance of 60 ppm and 1 missed cleavage, and the search was restricted to the human. The criteria for evaluating the search were the following: MOWSE score (>56), number and intensity of peptides matched, coverage of the candidate protein sequence and position on the 2D gel. Bioinformatic Validation of Identified Proteins as Potential PC Substrates. The amino acid sequence of each protein identified by MALDI TOF MS was screened for the presence of consensus PC cleavage sites using the publicly available program, Center for Biological Sequence Analysis (CBS) (http://www.cbs.dtu.dk/services).31 Validation of Caldesmon as a PC6 Substrate by Western Blot Analysis. HESC lysates (10 µg in CLB, treated with or without hPC6-A) were mixed with equal volumes of 2× reducing sample buffer [0.5 M Tris HCl, 10% (v/v) Glycerol, 10% (w/ v) SDS, β-mercaptoethanol and 0.5% bromophenol blue] and boiled for 5 min. Proteins were separated by 10% SDS-PAGE and blotted onto PVDF membrane. Membranes were blocked overnight at 4 °C (5% skim milk in TBS containing 0.2% Tween 20) and incubated for 1 h at room temperature with mouse monoclonal anti-caldesmon antibodies (1:500 dilution, Cald-5 and Cal-21, Sigma). Blots were washed and incubated with sheep anti-mouse-HRP antibody (1:10 000 dilution, Silenus, Victoria, Australia) for 1 h at room temperature. Chemiluminescence was detected using the ECL Plus Kit with ECL autoradiography film (Amersham Phamacia, Piscataway, NJ). Confirmation of Caldesmon mRNA Expression in HESCs by RT-PCR. HESCs were washed 3 times with PBS and lysed in Trizol reagent (Invitrogen), and RNA was extracted. RNA (1 µg) was reverse transcribed as previously described,32 and standard PCR was carried out using primers specific for human caldesmon isoform 1: forward, 5′ GAAACTACCGAGAAGGAAGA 3′, and reverse, 5′ TCTGCTTCCAACCTTGCCCT 3′, with an expected PCR product of 392 bp. PCR products were resolved on 1% agarose gels and stained with GelRed (Biotium, Hayward, CA). Immunolocalization of Caldesmon and PC6 in Decidual Cells in Vivo in the Human Uterus. The cellular localization of caldesmon and PC6 in the human uterus in the proliferative (days 9-14, n ) 5) and secretory (days 19-28, n ) 5) phases of the menstrual cycle was determined by immunohistochemistry. In brief, 5-µm paraffin sections were dewaxed in histosol (Sigma), rehydrated and then microwaved at high power (700 W) in 0.01 mL/L sodium citrate buffer (pH 6.0) for 5 min. Endogenous peroxidase activity was quenched with 6% H2O2 in 100% methanol (1:1 v/v) for 5 min at room temperature. Nonspecific binding was blocked with a blocking buffer (0.3 M NaCl in 50 mM Tris, pH 7.6, 0.1% Tween 20, and 15% horse serum) for 30 min at room temperature. The sections were incubated with caldesmon antibody (1:1000 dilution, clone Cal21, Sigma) at 37 °C for 30 min, washed and incubated with biotinylated horse anti-mouse IgG (1:200 dilution) for 30 min at room temperature. Signals were amplified with StreptABC/ HRP (DAKO Botany, NSW, Australia) and visualized with diaminobenzidine (DAKO). Sections were counterstained with Harris hematoxylin, dehydrated and mounted with DPX (ProSciTech, Thuringowa, QLD, Australia). Negative controls were included for each section using nonimmunized mouse immunoglobulin (IgG) (Dako) as the primary antibody. PC6 Journal of Proteome Research • Vol. 8, No. 11, 2009 4985

research articles

Figure 2. Cleavage of pERTKR-AMC by rhPC6 in decidual cell lysates. Cell lysates were incubated in the presence (9) or absence (O) of 10U of rhPC6 and assayed directly for the hydrolysis of 100 µM of PC fluorogenic substrate, pERTKR-AMC, over 50 min.

immunohistochemistry used an affinity purified sheep antibody with a protocol previously described.4

Results Cleavage of Cellular Proteins of Decidual HESCs with rhPC6. Our rationale to identify physiological substrates of PC6 in decidual cells was to search for decidual proteins that could be cleaved by rhPC6. Freshly isolated primary endometrial stromal cells (n ) 4) were decidualized in culture, and successful decidualization was confirmed by a significant increase in the secretion of the decidual marker prolactin (data not shown). The decidual cells were then lysed and equal amounts of cellular proteins were incubated with or without 10 units of rhPC6. To ensure that the rhPC6 was biologically active in each treatment, a widely used generic peptide PC substrate, the fluorogenic peptide pERTKR-MCA, was included in each sample and the activity of rhPC6 in hydrolyzing this exogenous substrate was monitored. Typical changes of fluorescence signals due to peptide hydrolysis in the presence or absence of rhPC6 in the HESC lysates are shown in Figure 2. A constant increase in fluorescence intensity over time was observed in every lysate to which rhPC6 was added, confirming that rhPC6 was biologically active and able to cleave PC6 substrates in the lysates. In contrast, no PC6 activity was detected in the control lysates without rhPC6, indicating that endogenous PC6 activity in the lysates was either below the level of detection or destroyed during the preparation of cellular proteins. Use of 2D DIGE To Identify Decidual Proteins That Were Cleaved by rhPC6. Following confirmation that rhPC6 was active in each treatment, the rhPC6-treated lysates were labeled with Cy5 and the untreated lysates with Cy3, or vice versa. The paired treated and untreated samples were combined together with the Cy2-labeled internal standard (a mix of equal amounts of treated and nontreated lysates) and separated on 2D PAGE over a pH range of 3-10. Each gel was sequentially scanned for each fluorescent label, and the images were overlaid digitally (Figure 3A). A total of 1678 protein spots were detected in decidual cell lysates. Following SameSpots 4986

Journal of Proteome Research • Vol. 8, No. 11, 2009

Kilpatrick et al.

Figure 3. Comparative 2D DIGE analysis of protein cleavage by hrPC6. Equal amounts of decidual cell lysates were treated with or without hrPC6, and the proteins were subsequently labeled using Cy2 (blue), Cy3 (green) or Cy5 (red). Labeled proteins were combined and separated by 2D PAGE prior to comparative analysis. (A). Representative 2D PAGE gel showing protein migration across a pH 3-10 gradient. Regions of interest are boxed. (B) Upper and (C) lower panels, respectively, as shown in panel A were enlarged to highlight protein spots of interest. Spots showing differences are indicated. Numbering refers to the protein spots identified by mass spectrometry (Table 1).

analysis, 18 protein spots were identified as present with significantly different abundance in the rhPC6-treated compared to the untreated samples. These changes were highly reproducible with multiple independent cell isolations (n ) 4). These spots were excised from the gels and further analyzed by MALDI-TOF mass spectrometry. Table 1 lists the 9 spots that were identified by mass spectrometry. The images of each of these spots on the gel at a high magnification are shown in Figure 3B,C, with reference to their exact positions on the overall image of the whole gel (Figure 3A). These proteins were identified as tropomyosin 2, tropomyosin 4, hypoxia up-regulated protein 1, and caldesmon, all of which contain consensus PC cleavage sites. Intriguingly, five of these spots (spots 4-8) were identified to be caldesmon (Table 1). These were further analyzed in detail, and the three-dimensional spot representations and normalized volumes of spots 4-8 in rhPC6treated and control samples are shown in Figure 4. The abundance of higher molecular weight (MW) spots 4, 5, and 6 was significantly reduced whereas that of lower MW spots 7 and 8 increased in rhPC6-treated samples compared to the controls (Figure 4). Spots 4-6 had a similar MW but slightly different pI values (Figure 3), suggesting that these spots may represent the same proteins with post-translational modifications (for example, alternate patterns of glycosylation). This was also true for spots 7-8. In addition, these spots had a slightly lower MW than spots 4-6, suggesting that rhPC6 cleaved the higher MW caldesmon (spots 4-6) and converted them into lower MW forms (spots 7-8). This is reflected also by a change in the amounts of these spots: an increase in lower MW spots 7-8 and a decrease in higher MW spots 4-6 (Figure 4).

research articles

Caldesmon Is a PC6 Substrate in Decidual Cells

Table 1. Proteins Identified As Differentially Cleaved by PC6 Using 2D DIGE and MALDI TOF Mass Spectrometrya spot

accession

protein name

MOWSE

matched eaksb

sequence coverage (%)

predicted mass (kDa)

predicted pI

fold change (+PC6 vs -PC6)

p-value (ANOVA)

1 2 3 4 5 6 7 8 9

P07951 P67936 Q9Y4L1 Q05682 Q05682 Q05682 Q05682 Q05682 Q9Y4L1

Tropomyosin beta chain Tropomyosin alpha-4 chain Hypoxia Up-regulated protein 1 Caldesmon Caldesmon Caldesmon Caldesmon Caldesmon Hypoxia Up-regulated protein 1

87c 62c 102 131 105 63 68 81 89

32 22 28 42 32 25 24 30 39

74% 60% 30% 44% 40% 33% 32% 39% 40%

32.9 28.6 111.5 93.2 93.2 93.2 93.2 93.2 111.5

4.66 4.67 5.16 5.63 5.63 5.63 5.63 5.63 5.16

–1.25 –1.26 1.78 –1.89 –2.00 –2.00 2.61 2.62 2.10

0.023 0.013 0.012 0.013 0.003 0.017 0.049 0.046 0.015

b

a Protein spots showing a significant change (p < 0.05; ANOVA) following PC6 cleavage were excised from the gel and submitted for MALDI TOF MS. MALDI TOF MS peak data is presented in the Supporting Information file. c Scores generated using up to 2 missed cleavages.

To determine which isoforms of caldesmon had been identified by mass spectrometry, the tryptic peptides identified by mass spectrometry were mapped against each of the caldesmon isoform amino acid sequences (Figure 6). In every case, several tryptic peptides derived from each protein spot identified as caldesmon were matched within the central region unique to isoform 1, as well as to the N- and C-terminal regions common to all five isoforms. In agreement with this, each spot identified as caldesmon displayed a MW well above 75 kDa on the 2D gel (Figure 3A,B) corresponding to the increased size of isoform 1 compared to isoforms 2-5. Therefore, the protein spots identified as caldesmon were all caldesmon isoform 1.

Figure 4. Example of proteomic expression changes observed following hrPC6 treatment. Protein expression was compared between hrPC6 treated or untreated samples, and spots of interest that showed increased (spots 4-6; upper panel) or decreased (spots 7-8; lower panel) expression were flagged for further analysis. Histograms (right-hand side) show the averaged normalized spot volumes (n ) 4) for each protein following incubation with (clear bars) or without (hatched bars) PC6. Spot volumes are also represented in 3-dimensional format (left-hand side). Spot numbers and treatments are as indicated.

Caldesmon was not previously known to be involved in decidualization, but it is known to be important for actin filament reorganization in other cell types. Because filament reorganization is an important feature of decidualization, we chose to focus on caldesmon for further analysis. Verification of Caldesmon Isoform 1 as One of the Decidual Proteins Cleaved by rhPC6. To date, five isoforms of human caldesmon resulting from alternate splicing of a single gene have been deposited in the Genbank database. However, no publication has clearly illustrated the relationship between these 5 isoforms. To clarify the molecular nature of the caldesmon spots identified, the primary sequence of all five known caldesmon isoforms was first aligned using the Australian National Genomic Information Service (ANGIS) (Figure 5). Caldesmon isoform 1 is the largest of the five (793 aa; MW ∼93 kDa) and possesses a unique ∼220 amino acid central region (residues ∼220-440) that is not present in isoforms 2-5. In addition, isoforms 4 and 5 have an identical N-terminal 18 amino acid sequence that is different to that found in isoforms 1-3 (Figure 5). These isoforms-specific features are also represented in Figure 6.

Bioinformatic Validation of Caldesmon Isoform 1 as a PC Substrate. To ascertain that caldesmon isoform 1 is a PC6 substrate, the primary sequence of human caldesmon isoform 1 was examined to determine whether it contains any consensus PC cleavage site. Three such sites were identified, one of which showed clear PC cleavage potential at amino acid position 110 (Figure 7). The sequence around this site (EERRQKRVLQ) agrees closely with the consensus PC cleavage motif [K/R-(X)n-(K/R)V], in accordance with caldesmon isoform 1 being a likely PC substrate. These potential PC cleavage sites are also mapped onto Figure 6 to illustrate their relationship with other parameters highlighted in that figure. Validation of Caldesmon as a PC6 Substrate in HESCs by Western Blot Analysis. To further validate that caldesmon isoform 1 is a PC6 substrate in decidual HESCs, cell lysates were treated with or without hPC6-A as for 2D DIGE analysis and subject to Western blotting. Two antibodies (Figure 6), one recognizing the N-terminal end of human caldesmon before the predicted main cleavage site of aa-110 (Cald-5), and the other recognizing the C-terminal end of the protein after the cleavage site (C-21), were used. It was predicted that antibody Cald-5 would recognize the full-length as well as the processed form of caldesmon isoform 1 and that antibody C-21 would detect only the processed form, should caldesmon be a substrate of PC6 and cleaved at the predicted cleavage site of aa-110 (Figure 6). The Western results using these two antibodies are shown in Figure 8. Antibody Cald-5 detected two clear bands around 100 kDa as predicted (Figure 8A). Treatment with rhPC6 reduced the intensity of the larger band (B1, the fulllength form) and increased the ratio of the smaller (B2, the processed form) to the larger band by 184%, highly consistent with the full-length caldesmon being converted into the short form by PC6. With identical samples, antibody C21 detected only one clear band (Figure 7B); its size is similar to the processed form demonstrated by antibody Cald-5 and its intensity was substantially increased by the rhPC6 treatment, Journal of Proteome Research • Vol. 8, No. 11, 2009 4987

research articles

Kilpatrick et al.

Figure 5. Alignment of the primary sequences of all 5 caldesmon isoforms. The protein sequence of each of the 5 known caldesmon isoforms was obtained from the NCBI protein database and aligned using ANGIS. Numbering indicates the amino acid position. Total amino acid length of each isoform is indicated at the end of each sequence. (*) identical residues; (-) alternate residues.

again in agreement with caldesmon isoform 1 being converted from its full-length (higher MW) to the processed form (lower 4988

Journal of Proteome Research • Vol. 8, No. 11, 2009

MW) by PC6 cleavage. These results therefore confirm that caldesmon is processed by PC6 as a substrate in decidual

Caldesmon Is a PC6 Substrate in Decidual Cells

research articles

Figure 6. Schematic illustration of caldesmon isoforms alignment, matched positions of the peptide sequences of the peptide sequences of spots 4-8, caldesmon antibody epitopes and the potential PC cleavage sites on caldesmon. Caldesmon isoforms were aligned by primary amino acid sequence (as in Figure 5). Regions of homology/identity with isoforms 1 are shown in blue. The unique N-terminal amino acid sequence of isoforms 4 and 5 is indicated in yellow. Epitopes recognized by anti-caldesmon antibodies (Cald5 and C-21) are indicated in gray. Potential PC cleavage sites are indicated by arrows; bold arrows indicate the most significant cleavage site identified (see text). The location of tryptic peptides observed following mass spectrometry analysis for each caldesmon spot identified is shown.

Figure 7. Analysis of caldesmon isoform-1 for potential PC cleavage sites. The three potential cleavage sites are shown; the one with the highest potential is shown by the bigger arrow. The inserted table shows the exact sequence around the each cleavage site.

HESCs, and that only caldesmon isoform 1 was involved in decidualization, as no lower MW bands corresponding to other isoforms were detected. To further verify that caldesmon isoform 1 was expressed in decidual HESCs, RT-PCR was employed and the expression of human caldesmon isoform 1 mRNA was clearly detected in decidual HESCs (Figure 8C). The cDNA from this band was sequenced and confirmed to be caldesmon isoform 1 (data not shown). Confirmation of Expression of Both Caldesmon and PC6 in Decidual Cells in Vivo in the Human Uterus. To further validate that caldesmon was expressed physiologically in vivo in decidual cells together with PC6, immunohistochemistry was performed to localize the both caldesmon and PC6 in the human uterus (Figure 9). During the proliferative phase of the menstrual cycle when no decidual cells are present, caldesmon was detected neither in the stroma nor the glands but only in a few scattered blood vessels (Figure 9A), whereas PC6 was detected only in the glandular epithelial cells but not in the stroma (Figure 9B). Thus, in the proliferative phase of the cycle, the two proteins are expressed in distinct cell types and neither PC6 nor caldesmon was detected in the nondecidualized stromal cells. In contrast, during the midlate secretory phase of the cycle when decidualization has occurred in the stromal compartment, high levels of

Figure 8. Confirmation of caldesmon as a PC6 substrate in decidual cells. (A and B) Western blotting of decidual cell lysates following treatment with (+PC6) or without (control) rhPC6 using two anti-caldesmon antibodies: an N-terminal antibody Cald-5 (A), and a C-terminal antibody C21 (B). The protein loading was checked by probing for β-actin. The change in intensity of bands (B1 and B2) in panel A following rhPC6 treatment is shown by the ratio between the two bands (B2/B1). (C) Confirmation of caldesmon expression in the decidual cells by RT-PCR. Lane M, molecular weight marker; lane 1, positive amplification of caldesmon isoform 1; lane 2, negative control in which reverse transcriptase was absent. Journal of Proteome Research • Vol. 8, No. 11, 2009 4989

research articles

Kilpatrick et al.

Figure 9. Immunohistochemical localization of caldesmon and PC6 in the human endometrium. (A and B) Midproliferative and (C-F) midlate secretory phase of the cycle was examined. The localization of caldesmon is shown in panels A, C, and D; the localization of PC6 is in panels B and E. A representative negative control for both antigens is shown in panel F. Very low levels of caldesmon were detected only in a few blood vessels in the proliferative phase (A), while its expressed was significantly up-regulated in the decidual cells and blood vessels in the secretory phase (C and D); (D) shows high magnification of panel C, highlighting strong expression in decidual cells and blood vessels, but no expression in the glands (inset in D). Low levels of PC6 were detected in the proliferative phase (A), but its expression was also dramatically increased in the secretory phase (E). Panels E and C are serial sections. Both caldesmon and PC6 are strongly expressed in the decidual cells in the secretory phase (C and E). ge, glandular epithelium; st, stroma; dec, decidual cells. Arrows show the blood vessels. Bar ) 50 µm.

caldesmon were detected particularly in the decidual cells in the stroma (Figure 9C,D). Importantly, on serial sections, strong expression of PC6 was also detected in the decidual cells (Figure 9E). These provide strong evidence that decidualization of stromal cells in vivo in the human uterus up-regulates the expression of both caldesmon and PC6. Interestingly, apart from the decidual cells, caldesmon and PC6 are localized in distinct cell types also in the secretory phase. Vascular smooth muscle cells showed strong imunoreactivity for caldesmon, whereas no signal was detected in the glandular epithelium (Figure 9D). In contrast, PC6 was highly expressed in the glands as previously published (Figure 9E). Together, these results suggest that caldesmon is a physiological substrate of PC6 specifically in the decidual cells. 4990

Journal of Proteome Research • Vol. 8, No. 11, 2009

Discussion The goal of this study was to investigate the mechanism of PC6 action during decidualization of uterine stromal cells for embryo implantation and for the establishment of pregnancy. To achieve this, we aimed to identify PC6 substrates in a physiologically relevant context, human uterine decidual cells, by searching for decidual proteins that could be cleaved by rhPC6 using the 2D DIGE technique. A number of candidate proteins were successfully identified, one of which was confirmed to be caldesmon. With several approaches, it was established that caldesmon isoform 1 is a physiological substrate of PC6 during decidualization of human uterine stromal fibroblasts. The study is novel and significant in a number of ways: it has identified previously

research articles

Caldesmon Is a PC6 Substrate in Decidual Cells unknown factors of importance for decidualization, it provides important insights into how PC6 functions during decidualization by regulating unconventional substrates such as cytoskeletal associated proteins, and it demonstrates a novel and unbiased approach to identify PC-regulated proteins and pathways of physiological relevance. Caldesmon is a thin-filament associated protein of smooth muscle and nonmuscle cells,33-35 with important functions in regulating cell motility and cytoskeletal remodelling.36 Historically, caldesmon is classified as high molecular weight (hcaldesmon) or low molecular weight (l-caldesmon) forms. It is now clear that h-caldesmon corresponds to isoform 1 (793 aa), whereas l-caldesmon is produced as isoforms 2-5 (532-564 aa) through alternative splicing of a single gene. The main difference between h-caldesmon and l-caldesmon is the lack of a central R-helical region (∼150 aa) in l-caldesmon.37 The N- and C-terminal regions harbor the major functional domains; the C-terminal region binds actin, tropomyosin and calmodulin, while the N-terminal region mainly interacts with myosin.37 The h-caldesmon binds to myosin and actin simultaneously and modulates the actomyosin interaction in smooth muscle cells, whereas l-caldesmon binds actin, stabilizes the actin stress fibers, and mediates the actin and nonmuscle myosin interaction in nonsmooth muscle cells.37 Decidualization of endometrial stromal cells requires major morphological, structural and functional changes, and is closely associated with cytoskeletal reorganization.38-40 Initiation of decidualization is accompanied by a significant down-regulation of R-smooth muscle actin,38 while disruption of actin filaments alters the course of decidualization.41 It was recently demonstrated that destabilization of the cytoskeleton accelerates decidualization, whereas increased phosphorylation of myosin light chain prevents decidualization.42 Given the importance of cytoskeletal restructuring during decidualization, it is not entirely surprising that caldesmon and its binding protein tropomyosin, the essential components of the cytoskeleton, are regulated during decidualization. Furthermore, caldesmon is important in the assembly of actin filaments as well as intermediate-sized filaments.34 The major components of the intermediate-sized filaments are desmin and vimentin,43 and desmin is a well-known marker of decidualization.44-48 It is thus likely that caldesmon functions through interaction with desmin and vimentin during decidualization. However, prior to this study, caldesmon was not even suspected to be a filament-regulating protein involved in decidualization. The discovery of caldesmon isoform 1 in nonmuscle cells (the decidual cells) is novel, as it is generally believed that h-caldesmon is preferentially expressed in smooth muscle cells.36 However, it is the N-terminal end of h-caldesmon that binds to myosin in smooth muscle cells.37 Interestingly, we showed that h-caldesmon was processed by PC6 specifically in decidual cells and its N-terminal end was removed. The decidual h-caldesmon is thus quite distinct from the conventional muscle h-caldesmon because of PC6 modification. PC6 was initially predicted to function like other PCs and regulate a number of growth factors.10,11,28 However, recently, PC6 was shown to be able to regulate extracellular proteins and cell surface proteins.9,15 The identification of caldesmon, a cytoskeletal protein, as a PC6 substrate suggests that the type of proteins PC6 regulates is much broader than previously thought. This is an important concept highlighting the need

to use an unbiased approach such as proteomics to uncover physiologically relevant substrates for PCs. This study also demonstrates that the substrates of PC6 are quite cell-specific, even in the same tissue. Although PC6 is expressed in both decidual and epithelial cells at the same time in the uterus, caldesmon is a PC6 substrate only in decidual cells. This underscores the diversity of PC substrates in different cell types, highlighting the need to identify cell-specific substrates of PCs in order to understand their unique roles in any specific cell type. In summary, this study employed a proteomic approach and identified a number of physiological substrates of PC6 during decidualization of human endometrial stromal cells. In particular, it has established that PC6 functions by regulating cytoskeletal protein caldesmon isoform 1 during decidualization. It therefore advances our understanding of the molecular mechanism of PC6 action in the uterus for embryo implantation and pregnancy establishment. This study has also demonstrated the power of proteomics to identify PC-regulated proteins in an unbiased manner.

Acknowledgment. This work was supported by a project grant (#441117) and Fellowships (#494808 to G.N. & #388901 to L.A.S.) from National Health and Medical Research Council of Australia, and an Ovarian Cancer Research Fellowship (A.N.S.). We thank Associate Professor David Robertson for his input to the proteomics approaches taken in this work. The authors declare no conflict of interest. Supporting Information Available: Supplemental data are provided for the mass spectra and matched peptides for all the identified protein spots shown in Table 1. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Robb, L.; Li, R.; Hartley, L.; Nandurkar, H. H.; Koentgen, F.; Begley, C. G. Infertility in female mice lacking the receptor for interleukin 11 is due to a defective uterine response to implantation. Nat. Med. 1998, 4, 303–308. (2) Wang, H.; Dey, S. K. Roadmap to embryo implantation: clues from mouse models. Nat. Rev. Genet. 2006, 7, 185–199. (3) Kurihara, I.; Lee, D.-K.; Petit, F. G.; Jeong, J.; Lee, K.; Lydon, J. P.; DeMayo, F. J.; Tsai, M.-J.; Tsai, S. Y. COUP-TFII Mediates Progesterone Regulation of Uterine Implantation by Controlling ER Activity. PLoS Genet. 2007, 3, e102. (4) Nie, G.; Li, Y.; Wang, M.; Liu, Y. X.; Findlay, J. K.; Salamonsen, L. A. Inhibiting uterine pc6 blocks embryo implantation: an obligatory role for a proprotein convertase in fertility. Biol. Reprod. 2005, 72, 1029–1036. (5) Okada, H.; Nie, G.; Salamonsen, L. A. Requirement for proprotein convertase 5/6 during decidualization of human endometrial stromal cells in vitro. J. Clin. Endocrinol. Metab. 2005, 90, 1028–1034. (6) Nie, G. Y.; Li, Y.; Minoura, H.; Findlay, J. K.; Salamonsen, L. A. Specific and transient up-regulation of proprotein convertase 6 at the site of embryo implantation and identification of a unique transcript in mouse uterus during early pregnancy. Biol. Reprod. 2003, 68, 439–447. (7) Wong, B. S.; Liu, S.; Schultz, G. A.; Rancourt, D. E. Subtilisin proprotein convertase-6 expression in the mouse uterus during implantation and artificially induced decidualization. Mol. Reprod. Dev. 2002, 61, 453–459. (8) Seidah, N.; Day, R.; Marcinkiewicz, M.; Chretien, M. Precursor convertases: an evolutionary ancient, cell-specific, combinatorial mechanism yielding diverse bioactive peptides and proteins. Ann. N.Y. Acad. Sci. 1998, 839, 9–24. (9) Seidah, N. G.; Mayer, G.; Zaid, A.; Rousselet, E.; Nassoury, N.; Poirier, S.; Essalmani, R.; Prat, A. The activation and physiological functions of the proprotein convertases. Int. J. Biochem. Cell Biol. 2008, 40, 1111–1125. (10) Seidah, N.; Chretien, M. Proprotein and prohormone convertases: a family of subtilases generating diverse bioactive polypeptides. Brain Res. 1999, 848, 45–62.

Journal of Proteome Research • Vol. 8, No. 11, 2009 4991

research articles (11) Rockwell, N. C.; Thorner, J. W. The kindest cuts of all: crystal structures of Kex2 and furin reveal secrets of precursor processing. Trends Biochem. Sci. 2004, 29, 80–87. (12) Scamuffa, N.; Siegfried, G.; Bontemps, Y.; Ma, L.; Basak, A.; Cherel, G.; Calvo, F.; Seidah, N. G.; Khatib, A. M. Selective inhibition of proprotein convertases represses the metastatic potential of human colorectal tumor cells. J. Clin. Invest. 2008, 118, 352–363. (13) Miranda, L.; Wolf, J.; Pichuantes, S.; Duke, R.; Franzusoff, A. Isolation of the human PC6 gene encoding the putative host proteases for HIV-1 gp160 processing in CD4+ T lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7695–7700. (14) De Bie, I.; Marcinkiewicz, M.; Malide, D.; Lazure, C.; Nakayama, K.; Bendayan, M.; Seidah, N. The isoforms of proprotein convertase PC5 are sorted to different subcellular compartments. J. Cell Biol. 1996, 135, 1261–1275. (15) Mayer, G.; Hamelin, J.; Asselin, M.-C.; Pasquato, A.; Marcinkiewicz, E.; Tang, M.; Tabibzadeh, S.; Seidah, N. G. The regulated cell surface zymogen activation of the proprotein convertase PC5A directs the processing of its secretory substrates. J. Biol. Chem. 2008, 283, 2373–2384. (16) Tang, M.; Mikhailik, A.; Pauli, I.; Giudice, L. C.; Fazelabas, A. T.; Tulac, S.; Carson, D. D.; Kaufman, D. G.; Barbier, C.; Creemers, J. W.; Tabibzadeh, S. Decidual differentiation of stromal cells promotes Proprotein Convertase 5/6 expression and lefty processing. Endocrinology 2005, 146, 60–70. (17) Freyer, C.; Kilpatrick, L. M.; Salamonsen, L. A.; Nie, G. Pro-protein convertases (PCs) other than PC6 are not tightly regulated for implantation in the human endometrium. Reproduction 2007, 133, 1189–1197. (18) Campan, M.; Yoshizumi, M.; Seidah, N.; Lee, M.-E.; Bianchi, C.; Haber, E. Increased proteolytic processing of protein tyrosine phosphate u in confluent vascular endothelial cells: the role of PC5, a member of the subtilisin family. Biochemistry 1996, 35, 3797–3802. (19) Ulloa, L.; Creemers, J. W.; Roy, S.; Liu, S.; Mason, J.; Tabibzadeh, S. Lefty proteins exhibit unique processing and activate the MAPK pathway. J. Biol. Chem. 2001, 276, 21387–21396. (20) Yana, I.; Weiss, S. J. Regulation of membrane type-1 matrix metalloproteinase activation by proprotein convertases. Mol. Biol. Cell 2000, 11, 2387–2401. (21) Stawowy, P.; Meyborg, H.; Stibenz, D.; Stawowy, N. B. P.; Roser, M.; Thanabalasingam, U.; Veinot, J. P.; Chretien, M.; Seidah, N. G.; Fleck, E.; Graf, K. Furin-like proprotein convertases are central regulators of the membrane type matrix metalloproteinase-promatrix metalloproteinase-2 proteolytic cascade in atherosclerosis. Circulation 2005, 111, 2820–2827. (22) Siegfried, G.; Basak, A.; Prichett-Pejic, W.; Scamuffa, N.; Ma, L.; Benjannet, S.; Veinot, J. P.; Calvo, F.; Seidah, N.; Khatib, A.-M. Regulation of the stepwise proteolytic cleavage and secretion of PDGF-B by the proprotein convertases. Oncogene 2005, 24, 6925– 6935. (23) Siegfried, G.; Khatib, A. M.; Benjannet, S.; Chretien, M.; Seidah, N. G. The proteolytic processing of pro-platelet-derived growth factor-A at RRKR(86) by members of the proprotein convertase family is functionally correlated to platelet-derived growth factorA-induced functions and tumorigenicity. Cancer Res. 2003, 63, 1458–1463. (24) Vollenweider, F.; Benjannet, S.; Decroly, E.; Savaria, D.; Lazure, C.; Thomas, G.; Chretien, M.; Seidah, N. Comparative cellular processing of the human immunodeficiency virus (HIV-1) envelope glycoprotein gp160 by the mammalian subtilisin/kexin-like convertases. Biochem. J. 1996, 314, 521–532. (25) Decroly, E.; Wouters, S.; Di Bello, C.; Lazure, C.; Ruysschaert, J.M.; Seidah, N. Identification of the paired basic convertases implicated in HIV gp160 processing based on in vitro assays and expression in CD4+ cell lines. J. Biol. Chem. 1996, 271, 30442– 30450. (26) Lissitzky, J.-C.; Luis, J.; Munzer, J. S.; Benjannet, S.; Parat, F.; Chretien, M.; Marvaldi, J.; Seidah, N. Endoproteolytic processing of integrin pro-R subunits involves the redundant function of furin and proprotein convertase (PC) 5A, but not paired basic amino acid converting enzyme (PACE) 4, PC5B or PC7. Biochem. J. 2000, 346, 133–138. (27) Bergeron, E.; Basak, A.; Decroly, E.; Seidah, N. G. Processing of alpha4 integrin by the proprotein convertases: His at position P6 regulates cleavage. Biochem. J. 2003, 373 (2), 475–484.

4992

Journal of Proteome Research • Vol. 8, No. 11, 2009

Kilpatrick et al. (28) Scamuffa, N.; Basak, A.; Lalou, C.; Wargnier, A.; Marcinkiewicz, J.; Siegfried, G.; Chretien, M.; Calvo, F.; Seidah, N. G.; Khatib, A. M. Regulation of prohepcidin processing and activity by the subtilisinlike proprotein convertases Furin, PC5, PACE4 and PC7. Gut 2008, 57, 1573–1582. (29) Essalmani, R.; Zaid, A.; Marcinkiewicz, J.; Chamberland, A.; Pasquato, A.; Seidah, N. G.; Prat, A. In vivo functions of the proprotein convertase PC5/6 during mouse development: Gdf11 is a likely substrate. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 5750– 5755. (30) Chen, J. C.; Hannan, N. J.; Mak, Y.; Nicholls, P. K.; Zhang, J.; Rainczuk, A.; Stanton, P. G.; Robertson, D. M.; Salamonsen, L. A.; Stephens, A. N. Proteomic characterization of midproliferative and midsecretory human endometrium. J. Proteome Res. 2009, 8 (4), 2032–2044. (31) Duckert, P.; Brunak, S.; Blom, N. Prediction of proprotein convertase cleavage sites. Protein Eng. Des. Sel. 2004, 17, 107–112. (32) Nie, G. Y.; Hampton, A.; Li, Y.; Findlay, J. K.; Salamonsen, L. A. Identification and cloning of two isoforms of human HtrA3, characterisation of its genomic structure and comparison of its tissue distribution with HtrA1 and HtrA2. Biochem. J. 2003, 371, 39–48. (33) Gunning, P.; O’Neill, G.; Hardeman, E. Tropomyosin-based regulation of the actin cytoskeleton in time and space. Physiol. Rev. 2008, 88, 1–35. (34) Deng, M.; Mohanan, S.; Polyak, E.; Chacko, S. Caldesmon is necessary for maintaining the actin and intermediate filaments in cultured bladder smooth muscle cells. Cell Motil. Cytoskeleton 2007, 64, 951–965. (35) Hai, C. M.; Gu, Z. Caldesmon phosphorylation in actin cytoskeletal remodeling. Eur. J. Cell Biol. 2006, 85, 305–309. (36) Sobue, K.; Hayashi, K.; Nishida, W. Expressional regulation of smooth muscle cell-specific genes in association with phenotypic modulation. Mol. Cell. Biochem. 1999, 190, 105–118. (37) Kordowska, J.; Huang, R.; Wang, C. L. Phosphorylation of caldesmon during smooth muscle contraction and cell migration or proliferation. J. Biomed. Sci. 2006, 13, 159–172. (38) Fazleabas, A. T.; Strakova, Z. Endometrial function: cell specific changes in the uterine environment. Mol. Cell. Endocrinol. 2002, 186, 143–147. (39) Salamonsen, L. A.; Dimitriadis, E.; Jones, R. L.; Nie, G. Complex regulation of decidualization: a role for cytokines and proteases-a review. Placenta 2003, 17, S76–85. (40) Cloke, B.; Huhtinen, K.; Fusi, L.; Kajihara, T.; Yliheikkila, M.; Ho, K. K.; Teklenburg, G.; Lavery, S.; Jones, M. C.; Trew, G.; Kim, J. J.; Lam, E. W.; Cartwright, J. E.; Poutanen, M.; Brosens, J. J. The androgen and progesterone receptors regulate distinct gene networks and cellular functions in decidualizing endometrium. Endocrinology 2008, 149, 4462–4474. (41) Kim, J. J.; Jaffe, R. C.; Fazleabas, A. T. Insulin-like growth factor binding protein-1 expression in baboon endometrial stromal cells: regulation by filamentous actin and requirement for de novo protein synthesis. Endocrinology 1999, 140, 997–1004. (42) Ihnatovych, I.; Hu, W.; Martin, J. L.; Fazleabas, A. T.; de Lanerolle, P.; Strakova, Z. Increased phosphorylation of myosin light chain prevents in vitro decidualization. Endocrinology 2007, 148, 3176– 3184. (43) Chang, L.; Goldman, R. D. Intermediate filaments mediate cytoskeletal crosstalk. Nat. Rev. Mol. Cell. Biol. 2004, 5, 601–613. (44) Tabanelli, S.; Tang, B.; Gurpide, E. In vitro decidualization of human endometrial stromal cells. J. Steroid Biochem. Mol. Biol. 1992, 42, 337–344. (45) Abrahamsohn, P. A.; Zorn, T. M. T. Implantation and decidualization in rodents. J. Exp. Zool. 1993, 266, 603–628. (46) Can, A.; Tekelioglu, M.; Baltaci, A. Expression of desmin and vimentin intermediate filaments in human decidual cells during first trimester pregnancy. Placenta 1995, 16, 261–275. (47) Oliveira, S. F.; Greca, C. P.; Abrahamsohn, P. A.; Reis, M. G.; Zorn, T. M. Organization of desmin-containing intermediate filaments during differentiation of mouse decidual cells. Histochem. Cell. Biol. 2000, 113, 319–327. (48) Korgun, E. T.; Cayli, S.; Asar, M.; Demir, R. Distribution of laminin, vimentin and desmin in the rat uterus during initial stages of implantation. J. Mol. Histol. 2007, 38, 253–260.

PR900381A