polymerization of insulin-like growth factor-binding protein-1 (igfbp-1) potentiates igf-i actions...

ORIGINAL ARTICLE 434J o u r n a l o fJ o u r n a l o f

CellularPhysiologyCellularPhysiology
Polymerization of Insulin-Like
Growth Factor-Binding Protein-1(IGFBP-1) Potentiates IGF-IActions in Placenta
HIROMI SHIBUYA,1 KEIJI SAKAI,1* MARYAM KABIR-SALMANI,2** YUICHI WACHI,1
AND MITSUTOSHI IWASHITA1

1Department of Obstetrics and Gynecology, Kyorin University School of Medicine, Tokyo, Japan2Department of Molecular Genetics, National Institute of Genetic Engineering and Biotechnology, Tehran, Iran

Insulin-like growth factor (IGF)-binding protein-1 (IGFBP-1), the main secretory protein of decidua that binds to IGFs and has been shownto inhibit or stimulate IGFs’ bioactivities. Polymerization, one of the posttranslational modifications of IGFBP-1, has been shown to lead toloss of inhibiting effect of IGFBP-1 on IGF-I actions. The current studies were undertaken to elucidate the effects of steroid hormones onIGFBP-1 polymerization in trophoblast cell cultures. Placental tissues were obtained during legal, elective procedures of termination ofpregnancy performed between 7 and 10 weeks of gestation, and primary trophoblast cells were separated. IGFBP-1 polymerization wasanalyzed by SDS–PAGE and immunoblotting. IGFBP-1 was polymerized when IGFBP-1 was added to trophoblast cell cultures.Polymerization of IGFBP-1 was inhibited by the addition of anti-tissue transglutaminase antibody into the culture media. There was anincrease in the intensity of polymerized IGFBP-1 bands with the addition of medroxyprogesterone acetate (MPA), while no such differencewas observed upon treatment with estradiol. MPA also increased the expression of tissue transglutaminase on trophoblast cellmembranes. IGF-I stimulated trophoblast cell migration, while IGFBP-1 inhibited this IGF-I-induced trophoblast response. Addition ofMPA attenuated the inhibitory effects of IGFBP-1 on IGF-I-induced trophoblast cell migration. IGFBP-1 was polymerized by tissuetransglutaminase on the cell surface of trophoblasts, and MPA increased tissue transglutaminase expression on the cell surface andfacilitated IGFBP-1 polymerization. These results suggest that progesterone might facilitate polymerization of decidua-secreted IGFBP-1and increase IGF-I actions at feto-maternal interface, thereby stimulating trophoblast invasion of maternal uterus.

J. Cell. Physiol. 226: 434–439, 2011. � 2010 Wiley-Liss, Inc.

Hiromi Shibuya, Keiji Sakai, and Maryam Kabir-Salmani contributedequally to this work.

*Correspondence to: Keiji Sakai, MD, Department of Obstetricsand Gynecology, Kyorin University School of Medicine, 6-20-2Shinkawa, Mitaka, Tokyo 181-8611, Japan.E-mail: [email protected]

**Correspondence to: Maryam Kabir-Salmani, Department ofMolecular Genetics, National Institute of Genetic Engineering andBiotechnology, Pajoohesh Blvd. Tehran-Karaj Highway, 17th Km.Tehran 14155-6343, Iran.

Received 29 October 2009; Accepted 15 July 2010

Published online in Wiley Online Library(wileyonlinelibrary.com), 29 July 2010.DOI: 10.1002/jcp.22349

During human placental development, cytotrophoblasts eitherdifferentiate as well as fuse to form syncytiotrophoblasts ordifferentiate into extravillous trophoblasts to form stratifiedcell columns (Bischof and Irminger-Finger, 2005). Appropriatetrophoblast invasion into maternal decidua is critical forplacental and fetal growth, and dysregulation of trophoblastinvasion results in several pregnancy-related complications,including abortion, intrauterine growth restriction, andpregnancy-induced hypertension (PIH; Kaufmann et al., 2003).Insulin-like growth factors (IGF-I and IGF- II) have been shownto play important roles in migration and proliferation oftrophoblast cells, and placental growth (Bauer et al., 1998;Kabir-Salmani et al., 2003). IGF-binding proteins (IGFBPs)present in extracellular fluid and modulate IGF’s actions in vivoand in vitro (Jones and Clemmons, 1995). IGFBP-1, one of theIGFBPs, is secreted from liver and decidua to maternalcirculation and has been shown to modulate IGF’s actionsduring pregnancy (Giudice et al., 1992; Rajkumar et al., 1996).Several posttranslational modifications have been shown toalter the affinity of IGFBP-1 for IGFs, including phosphorylation(Iwashita et al., 1998; Sakai et al., 2001a), proteolysis (Kabir-Salmani et al., 2005b), and polymerization (Busby et al., 1989;Sakai et al., 2001b).

Tissue transglutaminase is widely distributed in many cellsand tissues. Tissue transglutaminase is a cytosolic protein thathas also been observed in the nucleus and can be externalized tothe cell surface or extracellular matrix (Aeschlimann et al.,1995; Sakai et al., 2001b). It functions as a calcium-dependenttransamidating acyltransferase that catalyzes the formation ofisopeptide cross-links between glutamine and lysine residuesand also can attach primary amines to peptide-bonded

� 2 0 1 0 W I L E Y - L I S S , I N C .

glutamines (Mehta et al., 2006). Our previous studydemonstrated that IGFBP-1 is a substrate for tissuetransglutaminase and that it catalyzes the formation ofhigh-molecular weight covalently linked multimers, anddemonstrated that polymerization is an importantposttranslational modification of IGFBP-1 that regulates cellularresponses to IGF-I (Sakai et al., 2001b).

Progesterone is a steroid hormone with important growthfunctions in female reproductive tissues, including the breastand uterus (Rutanen, 2000; Mathews and Schneider, 2008). Itmay also modulate the actions of estrogen, IGF, and growthhormone (Etgen et al., 2006). We demonstrated previously thatestrogen stimulated the degrading actions of decidua-derivedIGFBP-1 and reduced IGFBP-1 phosphorylation, whereas

P O L Y M E R I Z A T I O N O F I G F B P - 1 I N P L A C E N T A 435

progesterone-induced decidual production of bothphosphorylated and non-phosphorylated IGFBP-1 (Kabir-Salmaniet al., 2005b).

The hypothesis that the embryo-maternal unit synthesizesconsiderable amounts of steroid hormones and that IGFBP-1and tissue transglutaminase are both controlled hormonally(Signorini et al., 1988; Lathi et al., 2005; Harada et al., 2006)prompted us to investigate the contribution of steroidhormones in regulating tissue transglutaminase-induced IGFBP-1polymerization at feto-maternal interface.

Materials and MethodsMaterials

Tissue culture media, fetal-calf serum (FCS), penicillin, andstreptomycin were purchased from Life Technologies, Inc. (GrandIsland, NY). IGF-I was a gift from Fujisawa Pharmaceutical (Osaka,Japan). Polyvinylidene difluoride (PVDF) transfer membranes(Immobilon-P) were obtained from Millipore (Bedford, MA).IGFBP-1 antibody was a gift from Dr. David Clemmons (Universityof North Carolina, Chapel Hill, NC). A monoclonal antibody totissue transglutaminase was purchased from NeoMarkar (Fremont,CA). All other chemicals were purchased from Sigma Chemical,Co. (St. Louis, MO).

Primary culture of villous trophoblasts

Placental tissues between 7 and 10 weeks of gestation wereobtained at legal, elective termination of pregnancy. All patientsgave informed consent for the collection and investigational use oftissues. This study was approved by the ethics committee of KyorinUniversity, School of Medicine, Tokyo, Japan. Trophoblasts wereobtained by a previously described method (Kabir-Salmani et al.,2004). Briefly, the tissues were rinsed several times with coldmedium 199 containing 100mg/ml streptomycin and 100 U/mlpenicillin. Villous tissue fragments were washed with medium 199supplemented with the above antibiotics and 10% FCS. Selectedtissues were cut into small pieces and other tissue componentssuch as blood vessels and decidual tissue were removed. The villousfragments were cultured in the same medium in tissue cultureflasks, which were precoated with 20mg/ml of fibronectin (Iwakiglass, Co., Chiba, Japan). Tissues were allowed to attach to thebottoms of the flasks for 2 h before adding the growth medium.Unattached tissues were removed after 2–3 days, and cultureswere continued for 1–2 weeks in medium 199 supplemented with10% FCS and antibiotics in an atmosphere of 5% CO2/95% air at378C and passaged after trypsinization using a split ratio of 1:4.Trophoblasts between passages 2 and 4 were utilized for theexperiments. The cells were identified as extravillous trophoblastsby immunohistochemical staining using anti-cytokeratin 7 and 8/18,anti-a5b1 and avb3 integrins, anti-vimentin, CD9, and factor VIII(Kabir-Salmani et al., 2004). More than 90% cells expressed all ofthese trophoblast markers.

Detection of IGFBP-1 polymerization in trophoblast cultures

To detect IGFBP-1 polymerization, human recombinant IGFBP-1(1mg/ml; Calbiochem Oncogene Research Products, San Diego,CA) was added to the serum-free media which was conditioned byincubating with primary trophoblast cells that had been pretreatedovernight with 17b-estradiol (E2, 0–10�5 M) ormedroxyprogesterone acetate (MPA, 0–10�5 M) in ethanol (finalconcentration <0.1%), and then incubated. The media werecollected and centrifuged at 1,200g for 5 min to remove cellularcomponents. Each conditioned medium (20ml) was added in 4�Laemmli buffer containing 400 mM dithiothreitol (DTT) and boiledfor 10 min. The proteins were analyzed by 10% SDS–PAGE andimmunoblotted for IGFBP-1 (7–9). Monomer and multimersIGFBP-1 were visualized by enhanced chemiluminescence(Super-Signal CL-H substrate system; Pierce, Rockford, IL) and

JOURNAL OF CELLULAR PHYSIOLOGY

exposed to Kodak X-AR film (Eastman Kodak, Rochester, NY).The experiments were performed in triplicate to ensurereproducibility. For Western ligand blotting, the proteins weretransferred to PVDF membranes, probed with 10 ng/ml of biotin-conjugated IGF-I (GroPep, Adelaide, SA, Australia) in TBST (10 mMTris–HCl, 150 mM NaCl, 0.1% Tween 20, pH 7.4) containing 0.5%bovine serum albumin (BSA) overnight at 48C, and then washedwith TBST containing 0.2% gelatin. The membranes were incubatedwith streptoavidine-horse raddish peroxidase (HRP) in TBSTcontaining 0.5% BSA for 3 h at room temperature. After incubation,membrane was washed three times for 10 min at roomtemperature in TBST with 0.2% gelatin. The membranes werevisualized by enhanced chemiluminescence as described above.Immunoblotting was performed in triplicate as described above.

To confirm the regulatory roles of MPA in IGFBP-1polymerization, cells were pretreated with 100 nM RU486 (Sigma),a specific progesterone receptor antagonist, before adding MPAand further treatment. Scanning densitometry was performedusing a scanning densitometer and the data were analyzed usingImage J (version 1.41).

Detection of progesterone receptors in trophoblast cultures

Trophoblast cells were grown to confluence in six-well cultureplates and the cells were solubilized in 100ml of SDS sample buffer,sonicated for 15 sec, and boiled for 10 min. Twenty microliters ofthe resulting lysate were analyzed by 10% SDS–PAGE andimmunoblotted using progesterone receptor antibody (C-20;Santa Cruz Biotechnology Co., Santa Cruz, CA).

Preparation of plasma membrane extract of primarytrophoblasts

Primary trophoblasts were grown to confluence in 10 cm tissueculture plates. The cultures were rinsed with serum-free medium199 and incubated with MPA (10�4–10�8 M) or E2 (10�5–10�9 M)for 0–24 h, and the plasma membrane extract was prepared aspreviously described (Sakai et al., 2001b). Cells were scraped fromthe plates using a cell scraper, washed twice using Ca2þ- and Mg2þ-free phosphate-buffered saline, and sonicated in buffer (20 mMHepes, pH 7.4, 100 mM NaCl, 5 mM KCl, 0.3 mM Na2HPO4, 1 mMNaHCO3, 250 mM sucrose, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, 1mg/ml pepstatin A, 1mg/ml leupeptin,1mg/ml aprotinin). The cell lysate was centrifuged at 16,000g for15 min, and the pellet was washed twice with the same buffer,solubilized in the same buffer containing 10 mM 3-((3-cholamidopropyl) dimethylammonio)-1-propanesulfonate(CHAPS), and centrifuged at 16,000g for 15 min. The supernatantwas collected and immunoprecipitated with anti-tissuetransglutaminase antibody and analyzed by SDS–PAGE andimmunoblotting. Scanning densitometry was performed using ascanning densitometer and the data were analyzed using Image J(version 1.41).

Cell migration assay

Extravillous trophoblast cells were plated at 1.5� 104/cm2 inmedium 199 containing 10% FBS and grown to confluence over3 days. Wounding was performed using a single razor blade toscrape cells off the culture plate, leaving a denuded area and a sharpvisible demarcation line at the wound edge, as described previously(Kabir-Salmani et al., 2004). The wounded monolayers were rinsedtwice with serum-free medium 199 and inspected immediatelyafter wounding. Sections of the wounds to be used for quantifyingmigration were selected according to previously described criteria(Kabir-Salmani et al., 2004), marked, and numbered. The cells wereincubated for 24 h with medium 199 containing 0.1% FCS in theabsence or presence of 100 ng/ml IGF-I with or without IGFBP-1and other reagents indicated in figures and figure legends. At theend of incubation, more than 95% cells were viable as assessed byTrypan blue dye exclusion. The cells were rinsed in PBS, fixed,

436 S H I B U Y A E T A L .

stained using Diff-Quick kit (International Reagents, Co., Kobe,Japan) and examined by phase-contrast microscopy. The migratedtrophoblast cells were counted in 1 mm long sections, keeping a20-mm gap from the demarcation line to minimize the possiblephysical effects of cell movement resulting from cell proliferation.Using a calibrated eyepiece grid, the mean number of migratedcells was determined. The mean of 10 sections per test substancefor each experiment was calculated, and the number of migratedcells was expressed as mean� SEM. Statistical analysis wascarried out using Stat View 5.0 (SAS Institute Inc., Cary, NC), andalternate Welch’s t-test and the Mann–WhitneyU-test were used tocompare the differences between the control and test groups(P< 0.05 was considered statistically significant). The experimentswere repeated at least four times in each group to assessreproducibility.

ResultsIGFBP-1 polymerization in primary trophoblast cultures

Because IGFBP-1 polymerization has been previouslydemonstrated in cell culture systems, we determined if theexposure of IGFBP-1 to trophoblast cultures would allowpolymerization. Recombinant IGFBP-1 was incubated withtrophoblast cell cultures in serum-free medium. IGFBP-1 waspolymerized within 48 h and was seen as 66 and 97 kDa bandsafter IGFBP-1 (1mg/ml) was added to medium of trophoblastcells (Fig. 1A). IGFBP-1 formed covalent multimers introphoblast cell cultures that were not altered with reducingreagents.

To determine the binding capacity of polymerized IGFBP-1to IGF-I, conditioned media of trophoblasts containing IGFBP-1were analyzed by SDS–PAGE following immunoblotting andligand blotting under non-reducing condition. No polymerizedIGFBP-1 band was observed if the same membrane wasreprobed with biotin-conjugated IGF-I under non-reducingconditions, suggesting that polymerized IGFBP-1 lost its abilityto bind to IGF-I (Fig. 1B).

Fig. 1. Polymerization of IGFBP-1 incubated with primary trophoblastsPrimary trophoblasts were grown to 80% confluence on 24-well cell cultureIGFBP-1(1mg/ml) for48h.Conditionedmedia(20ml)werecollectedandanThe figure shows a representative blot of three independent experimentsIGFBP-1.B:Primary trophoblastsweregrown to80%confluenceon24-welcontaining IGFBP-1 (1mg/ml) for 24 h. Conditioned media (20ml) were coreducing conditions. Themembrane was probedwith IGFBP-1 antibody fobiotin-conjugated IGF-I for ligandblotting.Thefigure showsarepresentativarrow denotes the monomeric form of IGFBP-1.


Effects of progesterone on IGFBP-1 polymerization inprimary trophoblast cultures

To determine the effects of steroid hormones on IGFBP-1polymerization by trophoblasts, conditioned media of MPA(10�5 and 10�7 M)-treated, E2 (10�5 and 10�7 M)-treated, anduntreated cells were analyzed by SDS–PAGE under reducingconditions, followed by IGFBP-1 immunoblotting. IGFBP-1polymerization in MPA-treated cells was found to be higherthan that in E2-treated and untreated cells (Fig. 2A).

To investigate the effects of MPA on IGFBP-1polymerization, time–course experiments were performedwith MPA-treated cells in the presence of IGFBP-1. The resultsrevealed that MPA treatment increased the intensity ofpolymerized IGFBP-1 band in a time-dependent manner(Fig. 2B). However, preincubation with RU486 (100 nM) beforethe addition of MPA (10�5 M) led to a decrease in polymerizedIGFBP-1 (Fig. 2C).

Since the polymerization of IGFBP-1 has been previouslyshown to occur in the presence of tissue transglutaminase, themonoclonal antibody for tissue transglutaminase (10 nM) waspreincubated before the addition of IGFBP-1. Exposure of cellsto the tissue transglutaminase antibody, inhibited IGFBP-1polymerization (Fig. 3). This result demonstrated that tissuetransglutaminase on trophoblast cell surface facilitated thepolymerization of IGFBP-1.

Effects of steroid hormones on tissue transglutaminaseexpression on the cell surface

Dose–response (Fig. 4A) and time–course (Fig. 4B)experiments were performed to detect the effects of steroidhormones on tissue transglutaminase expression onextravillous trophoblast cell membranes. Immunoblottingperformed in dose–response experiments produced consistentfindings of a dose-dependent increase in cell surface tissuetransglutaminase concentrations in MPA-treated cells(Fig. 4A, lower part), while no difference was observed in theselevels upon treatment with E2 (Fig. 4A, upper part). The results

in cultures and analysis by ligand blotting and immunoblotting. A:plates. Cells were incubatedwith serum-freemedium199 containingalyzedbySDS–PAGEandimmunoblottingunderreducingconditions.that gave similar results. The arrow denotes the monomeric form ofl cell cultureplates.Cellswere incubatedwith serum-freemedium199llected and analyzed by SDS–PAGE and ligand blotting under non-r immunoblotting, and then, the samemembrane was reprobedwitheblotof three independentexperiments thatgavesimilar results.The

Fig. 2. Polymerization of IGFBP-1with trophoblasts incubatedwithplacental steroid hormones. A: Primary trophoblasts were grown to80% confluence on 24-well cell culture plates. Cells were incubatedwith serum-free medium 199 containing E2 (0–10

�5M) or MPA(0–10�5M) overnight, and then IGFBP-1 (1mg/ml) was added andincubation was continued for 12h. Conditioned media (20ml) werecollected and analyzed by SDS–PAGE and immunoblotting underreducing conditions. The arrow denotes the monomeric form ofIGFBP-1. B: Trophoblast cells were incubated with serum-freemedium 199 containingMPA (10�5M) overnight. Thereafter, IGFBP-1(1mg/ml) was added and incubation was continued for 60h.Conditionedmedia (20ml) were collected and analyzed by SDS–PAGEand immunoblotting under reducing conditions. The arrow denotesthemonomeric formof IGFBP-1.C:Trophoblast cellswere incubatedwith serum-free medium 199 containing 100nM RU486 for 2 h. MPA(0 or 10�5M) was added and incubation was continued overnight.Thereafter, IGFBP-1 (1mg/ml) was added and incubation wascontinued for 12h. Conditioned media (20ml) were collected andanalyzed by SDS–PAGE and immunoblotting under reducingconditions. The figure shows a representative blot of threeindependent experiments that gave similar results. The arrowdenotes themonomeric formof IGFBP-1.The scanningdensitometryvalues for the bands of monomeric form shown in the left part of (A)were, from left to right: 19541, 19393, and19470; and for the right partof (A), they were: 21,761, 19,794, and 10,771; and for B, they were:18,032, 15,608, 14,223, 13,647, 13,388, 12,039, 8,018, and 1,741. Allvalues represent the mean of three separate experiments.

Fig. 3. Effect of anti-tissue transglutaminase antibody onpolymerization of IGFBP-1 in trophoblasts. Primary trophoblastsweregrown to80%confluence in24-well cell cultureplates.Cellswereincubated with serum-free medium 199 containing 10nM anti-tissuetransglutaminase antibody for 2 h. MPA (10�5M) was added andincubation was continued overnight. Thereafter, IGFBP-1 (1mg/ml)was added and incubation was continued overnight. Conditionedmedia (20ml) were collected and analyzed by SDS–PAGE andimmunoblotting under reducing conditions. The figure shows arepresentative blot of three independent experiments that gavesimilar results. The arrow denotes the monomeric form of IGFBP-1.


of time–course experiments revealed that while incubationwith 10�6 M E2 (0–24 h) had almost no effect on expressionlevels of cell surface tissue transglutaminase (Fig. 4B, upperpart), transglutaminase levels increased with 10�5 M MPA


treatment in a time-dependent manner (Fig. 4B, lower part).The increase of transglutaminase concentrations in trophoblastcell membrane were inhibited by the addition of RU486(Fig. 4C), since extravillous trophoblast cells expressprogesterone receptor A and B (Fig. 4D).

Progesterone-mediated improvement in IGF-I-inducedmigration of primary trophoblasts

Since the critical role of extravillous trophoblast cells is invasioninto maternal deciduas, a cell wounding migration assay wasperformed to detect the effects of steroid hormones on IGF-I-induced cell migration. The number of migrated cellssignificantly increased with IGF-I treatment compared tocontrol (P< 0.05), while addition of the IGF-I/IGFBP-1 complexsignificantly reduced the number compared to IGF-I-treatedcells (Fig. 5). This inhibitory effect of IGFBP-1 was significantlyinhibited by the addition of MPA (P< 0.05). Therefore, MPAattenuated the inhibitory actions of IGFBP-1 on IGF-I-inducedtrophoblast cell migration by promoting IGFBP-1polymerization.

Discussion

Locally acting molecules, including IGF and IGFBP-1, arerequired for placentation to occur during the first trimester(Gleeson et al., 2001). IGFBP-1, a major product of decidualized

Fig. 4. Effect of placental steroid hormones on tissuetransglutaminase expression on cell membranes of primarytrophoblasts. Primary trophoblasts were grown to confluence on10 cm tissue culture plates. A: The cultures were incubated in serum-free medium 199 with MPA (10�4–10�8M) or E2 (10

�5–10�9M) forovernight. B: The cultures were incubated in serum-freemedium199withMPA(10�5M)orE2 (10

�6M) for indicatedperiod.C:Thecultureswere incubated in serum-free medium 199 with MPA (0 or 10�5M)and RU486 (0 or 10�7M) for overnight. Membrane extract wereprepared and immunoprecipitated with anti-tissue transglutaminaseantibody (1mg). This was followed by analysis with SDS–PAGE andimmunoblotting using the same antibody. D: Trophoblast cells weregrown to confluence in six-well culture plates and the cells weresolubilized in 100ml of SDS-sample buffer, sonicated for 15 sec, andboiled for 10min. Cell lysate were analyzed by 10% SDS–PAGE andimmunoblotted using progesterone receptor antibody (PRA:progesterone receptor A, PRB: progesterone receptor B). The figureshows a representative blot of three independent experiments thatgave similar results. The scanning densitometry values for the bandsshown in the top part of (A) were, from left to right: 7,321, 8,125,7,925, 7,097, and 8,191; and for the bottom part of (A) they were:6,594, 6,386, 14,241, 24,561, and 26,952; and for the top part of (B),they were: 6,712, 6,597, 7,053, 7,123, and 5,949; and for the bottompart of A, they were: 5,524, 7,890, 12,388, 14,812, and 16,542. For C,they were: 5,005, 6,109, and 17,599. All values represent the mean ofthree separate experiments.

Fig. 5. Effects of IGFBP-1 polymerization on IGF-I-inducedmigration of primary trophoblasts. Primary trophoblastswere grownto confluence in six-well culture plates. The cultures were woundedwith a single razor blade and treated with IGF-I (0 or 100ng/ml),IGFBP-1 (0 or 1mg/ml), MPA (0 or 10�5M), E2 (0 or 10

�6M), or RU486(0 or 10�7M). The number of cells that migrated across the woundwas counted as described in Materials and Methods Section. Theresults are shown asmeanWSE (nU 8–20 replicates per experiment).The representative experiment shownwas repeated three timeswithsimilar results.

438 S H I B U Y A E T A L .

human endometrial stromal cells, is spatially and temporallypositioned to play a key role in the regulation of IGFsbioavailability at the embryo-maternal interface (Giudice et al.,1992). It has been shown that posttranslational modifications ofIGFBP-1 alter cellular response to IGFs. Polymerization ofIGFBP-1, one of the posttranslational modifications, was foundfor the first time in amniotic fluid, although the physiologicalsignificance of IGFBP-1 polymerization during pregnancy wasnot elucidated (Busby et al., 1989). Recently, we reported that


steroid hormones regulate posttranslational modification ofIGFBP-1 (phosphorylation and degradation) by decidual cells(Kabir-Salmani et al., 2005b). In this study, we determined theeffects of steroid hormones on tissue transglutaminaseexpression and the consequent tissue transglutaminase-induced IGFBP-1 polymerization by trophoblasts. Our findingsdemonstrated, for the first time, that MPA increases tissuetransglutaminase-induced IGFBP-1 polymerization on thetrophoblast cell surface. Polymerization, which is an importantposttranslational modification of IGFBP-1, plays a significantphysiological role in regulating the bioavailabilities andbioactivities of IGFs at the feto-maternal interface. The additionof steroid hormones to the culture media revealed an increasein the ratio of polymerized to unpolymerized IGFBP-1 in MPA-treated cells compared to that observed in E2-treated andcontrol cells (Fig. 1). Since tissue transglutaminase facilitatesIGFBP-1 polymerization in other cell types (Sakai et al., 2001b),we examined the effects of steroid hormones on tissuetransglutaminase expression, which could potentiate IGFBP-1polymerization by trophoblast cells. The results of time–courseand dose–response immunoblot analyses revealed that MPAincreased tissue transglutaminase expression on the cellmembrane in a time- and dose-dependent manner, while E2 hadalmost no effect (Fig. 4). These findings are in agreement withthose of Deng et al. (2003), who demonstrated markedupregulation of tissue transglutaminase in the secretory phase,indicating its regulation by progesterone. Tissuetransglutaminase mRNA has been reported to be consistentlyinduced in a dose-dependent manner after the addition ofprogesterone to uterine stromal cell cultures (Fujimoto et al.,1996). However, controversies exist regarding the effect ofestrogen treatment on tissue transglutaminase expression indifferent tissues. While consistent with our observations, Denget al. demonstrated that estrogen treatment did not affecthuman endometrial tissue transglutaminase concentration. Onthe other hand, a marked increase in the transglutaminaseactivity was reported in rat vaginal epithelial cells (Vijayalakshmiand Gupta, 1994).


Consistent with the effects of steroid hormones on tissuetransglutaminase expression, the results of time–course anddose–response immunoblots for IGFBP-1 polymerizationshowed that pretreatment of cells with MPA before incubationwith recombinant IGFBP-1 increased the formation ofcovalently linked IGFBP-1 multimers that were stable duringSDS–PAGE under reducing conditions (Figs. 2 and 3); however,no change was observed with E2 treatment (Fig. 2). To confirmthat MPA specifically induces IGFBP-1 polymerization,trophoblast cells were pretreated with RU486, a progesteronereceptor antagonist, before the addition of MPA. Our resultsshowed that MPA-induced IGFBP-1 polymerization wasattenuated when progesterone receptors were blocked (Fig. 2).Although a previous study reported that tissuetransglutaminase catalyzes the formation of IGFBP-1 multimerson fibroblast cell membranes (Sakai et al., 2001b), the currentfindings show, for the first time, that MPA stimulates tissuetransglutaminase-induced IGFBP-1 polymerization.

Our results showed that MPA significantly enhanced(P< 0.05) the migration of primary trophoblasts cultured withthe IGF/IGFBP-1 complex compared to E2-treated or controlcultures (Fig. 5). Human trophoblastic transglutaminase hasbeen demonstrated to regulate the different functions (Jensenet al., 1993; Hager et al., 1997; Kabir-Salmani et al., 2005a), andin the current study, we identified its migration promotingfunction at the embryo-maternal interface. Several studies havealso reported the separate and combined effects of IGF-I, IGF-II,and IGFBP-1 on proliferation and migration of human villoustrophoblasts and extravillous trophoblasts (Kabir-Salmani et al.,2003, 2004; Hills et al., 2004). However, this is the first reportto determine the role of steroid hormones in trophoblastmigration through modulation of the IGF/IGFBP-1 complex bytissue transglutaminase. More interestingly, IGFBP-1, whichshares an Arg-Gly-Asp (RGD) motif with other IGFBPs in its C-terminal domain, modulates cell motility by binding to a5b1

integrin. It can also modify the mitogenic effects of epidermalgrowth factor at the embryo-maternal interface (Hamiltonet al., 1998; Irwin and Giudice, 1998; Hills et al., 2004; Cavailleet al., 2006). Considering that polymerized IGFBP-1 has higheraffinity for a5b1 integrin, it is tempting to suggest that MPAtreatment may increase trophoblast migration not only throughthe release of IGF from the IGF/IGFBP-1 complex, but also viaIGFBP-1 polymerization and subsequent integrin activationthrough IGFBP-1 binding. Taken together, we propose thatMPA-stimulated tissue transglutaminase-induced IGFBP-1polymerization increases trophoblast migration as a result ofdecrease in affinity of polymerized IGFBP-1 for IGF andsubsequent activation of IGF and integrin receptors.

Acknowledgments

We would like to thank Dr. D.R.Clemmons (University of NorthCarolina at Chapel Hill) for his valuable suggestion and review.

Literature Cited

Aeschlimann D, Kaupp O, Paulsson M. 1995. Transglutaminase-catalyzed matrix cross-linkingin differentiating cartilage: Identification of osteonectin as a major glutaminyl substrate.J Cell Biol 129:881–892.

Bauer MK, Harding JE, Bassett NS, Breier BH, Oliver MH, Gallaher BH, Evans PC, WoodallSM, Gluckman PD. 1998. Fetal growth and placental function. Mol Cell Endocrinol140:115–120.

Bischof P, Irminger-Finger I. 2005. The human cytotrophoblastic cell, a mononuclearchameleon. Int J Biochem Cell Biol 37:1–16.


Busby WH, Hossenlopp P, Binoux M, Clemmons DR. 1989. Purified preparations of theamniotic fluid-derived insuline-like growth factor-binding protein contain multimericforms that are biologically active. Endocrinology 125:773–777.

Cavaille F, Neau E, Vouters M, Bry-Gauillard H, Colombel A, Milliez J, Le Bouc Y. 2006. IGFBP-1 inhibits EGF mitogenic activity in cultured endometrial stromal cells. Biochem BiophysRes Commun 345:754–760.

Deng L, Shipley GL, Loose-Mitchell DS, Stancel GM, Broaddus R, Pickar JH, Davies PJ. 2003.Coordinate regulation of the production and signaling of retinoic acid by estrogen in thehuman endometrium. J Clin Endocrinol Metab 88:2157–2163.

Etgen AM, Gonzalez-Flores O, Todd BJ. 2006. The role of insulin-like growth factor-Iand growth factor-associated signal transduction pathways in estradiol andprogesterone facilitation of female reproductive behaviors. Front Neuroendocrinol27:363–375.

Fujimoto M, Kanzaki H, Nakayama H, Higuchi T, Hatayama H, Iwai M, Kaneko Y, Mori T, FujitaJ. 1996. Requirement for transglutaminase in progesterone-induced decidualization ofhuman endometrial stromal cells. Endocrinology 137:1096–1101.

Giudice LC, Dsupin BA, Irwin JC. 1992. Steroid and peptide regulation of insulin-like growthfactor-binding proteins secreted by human endometrial stromal cell is dependent onstromal differentiation. J Clin Endocrinol Metab 75:1235–1241.

Gleeson LM, Chakraborty C, Mckinnon T, Lala PK. 2001. Insulin-like growth factor bindingprotein 1 stimulates human trophoblast migration by signaling through a5b1 integrin viamitogen-activated protein kinase pathway. J Clin Endocrinol Metab 86:2484–2493.

Hager H, Jensen PH, Hamilton-Dutoit S, Neilsen MS, Birckbichler P, Gliemann J. 1997.Expression of tissue transglutaminase in human bladder carcinoma. J Pathol 183:398–403.

Hamilton GS, Lysiak JJ, Han VK, Lala PK. 1998. Autocrine–paracrine regulation of humantrophoblast invasiveness by insulin-like growth factor (IGF)-II and IGF-binding protein(IGFBP)-1. Exp Cell Res 244:147–156.

Harada M, Osuga Y, Takemura Y, Yoshino O, Koga K, Hirota T, Morimoto C, Yano T,Taketani Y. 2006. Mechanical stretch upregulates IGFBP-1 secretion from decidualizedendometrial stromal cells. Am J Physiol Endocrinol Metab 290:E268–E272.

Hills FA, Elder MG, Chard T, Sullivan MH. 2004. Regulation of human villous trophoblast byinsulin-like growth factors and insulin-like growth factor-binding protein-1. J Endocrinol183:487–496.

Irwin JC, Giudice LC. 1998. Insulin-like growth factor binding protein-1 binds to placentalcytotrophoblast alpha5beta1 integrin and inhibits cytotrophoblast invasion intodecidualized endometrial stromal cultures. Growth Horm IGF Res 8:21–31.

Iwashita M, Sakai K, Kudo Y, Takeda Y. 1998. Phosphoisoforms of insulin-like growth factorbinding protein-1 in appropriate-for-gestational age and small-for gestational age fetuses.Growth Horm IGF Res 8:487–493.

Jensen PH, Lorand L, Ebbesen P, Gliemann J. 1993. Type-2 plasminogen-activator inhibitor is asubstrate for trophoblast transglutaminase and factor XIIIa. Transglutaminase-catalyzedcross-linking to cellular and extracellular structures. Eur J Biochem 214:141–146.

Jones JI, Clemmons DR. 1995. Insulin-like growth factors and their binding proteins: Biologicalactions. Endocr Rev 6:3–34.

Kabir-Salmani M, Shiokawa S, Akimoto Y, Sakai K, Nagamatsu S, Sakai K, Nakamura Y, Lotfi A,Kawakami H, Iwashita M. 2003. Alpha v beta 3 integrin signaling pathway is involved ininsulin-like growth factor I-stimulated human extravillous trophoblast cell migration.Endocrinology 144:1620–1630.

Kabir-Salmani M, Shiokawa S, Akimoto Y, Sakai K, Iwashita M. 2004. The role ofalpha(5)beta(1)-integrin in the IGF-I-induced migration of extravillous trophoblast cellsduring the process of implantation. Mol Hum Reprod 10:91–97.

Kabir-Salmani M, Shiokawa S, Akimoto Y, Sakai K, Sakai K, Iwashita M. 2005a. Tissuetransglutaminase at embryo-maternal interface. J Clin Endocrinol Metab 90:4694–4702.

Kabir-Slmani M, Shimizu Y, Sakai K, Iwashita M. 2005b. Posttranslational modifications ofdecidual IGFBP-1 by steroid hormones in vitro. Mol Hum Reprod 11:667–671.

Kaufmann P, Black S, Huppertz B. 2003. Endovascular trophoblast invasion: Implicationsfor the pathogenesis of intrauterine growth retardation and preeclampsia. BiolReprod 69:1–7.

Lathi RB, Hess AP, Tulac S, Nayak NR, Conti M, Giudice LC. 2005. Dose-dependentinsulin regulation of insulin-like growth factor binding protein-1 in human endometrialstromal cells is mediated by distinct signaling pathways. J Clin Endocrinol Metab90:1599–1606.

Mathews L, Schneider SS. 2008. Insulin-like growth factor-I inhibits growth regulatoryresponses engaged by estrogen and progesterone in the mouse mammary gland. Eur JCancer Prev 17:297–305.

Mehta K, Fok JY, Mangala LS. 2006. Tissue transglutaminase: From biological glue to cellsurvival cues. Front Biosci 11:173–185.

Rajkumar K, Dheen T, Krsek M, Murphy LJ. 1996. Impaired estrogen action in the uterus ofinsulin-like growth factor binding protein-1 transgenic mice. Endocrinology 137:1258–1264.

Rutanen EM. 2000. Insulin-like growth factor and insulin-like growth factor bindingproteins in the endometrium. Effect of intrauterine levonorgestrel delivery. HumReprod 15:173–181.

Sakai K, D’Ercole AJ, Murphy LJ, Clemmons DR. 2001a. Physiological differences in insulin-likegrowth factor binding protein-1 (IGFBP-1) phosphorylation in IGFBP-1 transgenic mice.Diabetes 50:2–8.

Sakai K, Busby WH, Clarke JB, Clemmons DR. 2001b. Tissue transglutaminase facilitates thepolymerization of insulin-like growth factor-binding protein-1 (IGFBP-1) and leads to lossof IGFBP-1’s ability to inhibit insulin-like growth factor-I-stimulated protein synthesis. J BiolChem 276:8740–8745.

Signorini M, Pansini F, Bonaccorsi G, Mollica G, Ferrari C, Bergamini CM. 1988. Regulationof endometrial transglutaminase activity during the menstrual cycle. Biochem Int 16:7–82.

Vijayalakshmi V, Gupta PD. 1994. Estradiol-regulated transamidation of keratins vaginalepithelial cell transglutaminase. Exp Cell Res 214:358–366.

polymerization of insulin-like growth factor-binding protein-1 (igfbp-1) potentiates igf-i actions...

Documents