{"paper_id":"9e578db7-27d5-41e2-8119-4745ac98eb33","body_text":"MOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011 | MILINGOS ET AL. | 21\nINTRODUCTION\nEndometriosis is a common benign gy-\nnecological condition that is defined as\nthe presence of endometrial stromal and\nglandular cells outside the endometrial\ncavity and has a prevalence of 6–10% in\nwomen of reproductive age (1). This\nprevalence is approximately 20% in\nwomen with infertility (2). Despite ex-\ntensive research and numerous theories\nproposed, the pathogenesis of en-\ndometriosis is yet to be determined, and\nseveral immunological and growth fac-\ntors have been investigated in the estab-\nlishment and maintenance of endometri-\notic lesions.\nIn addition, it has been suggested that\nectopic endometrial cells undergo de-\ncreased apoptosis compared with eutopic\nendometrial cells (3) and that insulinlike\ngrowth factor 1 (IGF-1) is one of the con-\ntributing factors that inhibits apoptosis\nand acts mitogenically on endometrial\ncells in vitro (4). Furthermore, increased\nlevels of IGF-1 have been found in the\nperitoneal fluid of women with endo -\nmetriosis compared with controls (5) and\nwere associated with increased proteoly-\nsis of IGF-binding protein 3 (IGFBP-3)\nmediated by urokinase-type plasmino-\ngen activator (uPA) (6,7).\nThe igf-1 gene contains six exons that,\nin humans, give rise to three igf-1 gene\ntranscripts by alternative splicing,\nnamely IGF-1Ea, IGF-1Eb and IGF-1Ec\n(which has also been named mechano\ngrowth factor [MGF]). The resulting\nIGF-1 isoforms undergo posttranslational\ncleavage to produce a common biologi-\ncally active product, namely the mature\nIGF-1, which is encoded by exons 3 and\nInsulinlike Growth Factor-1Ec (MGF) Expression in Eutopic\nand Ectopic Endometrium: Characterization of the \nMGF E-Peptide Actions \nIn Vitro\nDimitrios S Milingos,1 Anastassios Philippou,1 Athanassios Armakolas,1 Efstathia Papageorgiou,1\nAntigone Sourla,2 Athanassios Protopapas,3 Anthi Liapi,3 Aris Antsaklis,3 Minas Mastrominas,4 and \nMichael Koutsilieris1\n1Department of Experimental Physiology, Medical School, National and Kapodistrian University of Athens, Goudi-Athens, Greece;\n2Endo/OncoResearch Medical Laboratories, Ampelokipi-Athens, Greece; 3First Department of Obstetrics and Gynecology,\n“ALEXANDRAS” General Hospital, Medical School, National and Kapodistrian University of Athens, Ampelokipi-Athens, Greece;\nand \n4Embryogenesis, Inc., Maroussi, Athens, Greece\nThe transcription of the insulinlike growth factor 1(igf-1) gene generates three mRNA isoforms, namely IGF-1Ea, IGF-1Eb and IGF-\n1Ec (or MGF [mechano growth factor]). Herein, we analyzed the expression of IGF-1 isoforms in eutopic and ectopic endometrium\n(red lesions and endometriotic cysts) of women with endometriosis, and we characterized the actions of a synthetic MGF E-peptide\non KLE cells. Our data documented that all three \nigf-1 gene transcripts are expressed in the stromal cells of the eutopic and ec-\ntopic endometrium; however, endometriotic cysts contained significantly lower IGF-1 isoform expression, both at the mRNA and\nprotein level, as was shown using semiquantitative PCR and immunohistochemical methods. In addition, the glandular cells of the\neutopic endometrium did not express any of the IGF-1 isoforms; however, the glandular cells of the ectopic endometrium (red le-\nsions) did express the IGF-1Ec at mRNA and protein level. Furthermore, synthetic MGF E-peptide, which comprised the last 24\namino acids of the MGF , stimulated the growth of the KLE cells. Experimental silencing of the type 1 IGF receptor (IGF-1R) and in-\nsulin receptor expression of KLE cells (siRNA knock-out methods) did not alter the mitogenic action of the synthetic MGF E-peptide,\nrevealing that MGF E-peptide stimulates the growth of KLE cells via an IGF-1R–independent and insulin receptor–independent\nmechanism. These data suggest that the IGF-1Ec transcript might generate, apart from mature IGF-1 peptide, another posttrans-\nlational bioactive product that may have an important role in endometriosis pathophysiology.\n© 2011 The Feinstein Institute for Medical Research, www.feinsteininstitute.org\nOnline address: http://www.molmed.org\ndoi: 10.2119/molmed.2010.00043\nAddress correspondence and reprint requests to Michael Koutsilieris, MD, Department of\nExperimental Physiology, Medical School, University of Athens, 75 Micras Asias, Goudi,\nAthens, 115 27, Greece. Phone: 0030210-7462507; Fax: 0030210-7462571; E-mail:\nmkoutsil@med.uoa.gr.\nSubmitted March 29, 2010; accepted for publication September 12, 2010; Epub\n(www.molmed.org) ahead of print September 14, 2010.\n\n22 | MILINGOS ET AL. | MOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011\nMGF EXPRESSION IN EUTOPIC AND ECTOPIC ENDOMETRIUM\nposttranslational bioactive E-peptide of\nthe IGF-1Ec isoform may be involved in\nthe pathophysiology of endometriosis.\nMATERIALS AND METHODS\nEthical Approval\nA written informed consent was ob-\ntained by all the volunteers to participate\nin this study, which was approved by the\nEthics Committee of the National and\nKapodistrian University of Athens, and\nall experimental procedures conformed\nto the Declaration of Helsinki.\nSubjects\nThe subjects were women of reproduc-\ntive age undergoing laparoscopy for en-\ndometriosis. Median age of the women\nwas 35.7 years (range 28–49), and none\nhad received any form of hormone ther-\napy up to 3 months before the operation.\nzation of the intracellular signaling of\nMGF E-domain vis-à-vis IGF-1 signaling\nin endometrial-like cells (20,22).\nHerein, we report that all IGF-1 iso-\nforms are expressed in both eutopic and\nectopic endometrium, which is, however,\nsignificantly lower in endometriotic cysts\ncompared with either eutopic en-\ndometrium or red lesions. In addition,\nwe report for the first time that the glan-\ndular cells of eutopic endometrium and\nendometriotic cysts are deprived of any\nexpression of the IGF-1 isoforms,\nwhereas the glandular cells of red lesions\nexpress the IGF-1Ec isoform. Further-\nmore, our data documented that a syn-\nthetic MGF E-peptide can stimulate the\nproliferation of human KLE cells, an en-\ndometrial carcinoma cell line with a phe-\nnotype of endometrial-like cells, via an\nIGF-1R–independent and IR-independent\nmechanism. These data suggest that a\n4, and it is responsible for binding with\nthe IGF receptors and different E-domain\nproducts (Figure 1), which contain differ-\nent parts of exon 5 and/or exon 6 (8–11)\nand have been proposed to act au-\ntonomously (8,12).\nIGF-1 mediates its actions through\nbinding to specific receptors, such as the\ntype 1 IGF receptor (IGF-1R), the insulin\nreceptor (IR), and several atypical recep-\ntors such as the hybrid IR/IGF-1R. IGF-\n1R and IR are cell surface heterotetrameric\ntyrosine kinase receptors that are coupled\nto intracellular signaling pathways, such\nas the ras-raf-MAPK-ERKs and PI3K-AKT\nsignaling cascades (13).\nExcept for binding IGF-1, IGF-1R can\nalso bind insulinlike growth factor 2\n(IGF-2). This is a small peptide that\nshares approximately 60% of amino acids\nwith IGF-1 and 40% with pro-insulin,\nand by its binding to IGF-1R, IGF-2 regu-\nlates cell proliferation, survival and dif-\nferentiation. The affinity of IGF-2 for\nbinding IGF-1R is far less than IGF-1 and\nso it is for insulin (14,15). Although IGF-2\ncan bind all three receptors (IGF-1R,\nIGF-2R and IR), its mitogenic and meta-\nbolic actions are mediated primarily by\nbinding to IGF-1R. In contrast to IGF-1R,\nIGF-2R is a transmembrane single-chain\nglycoprotein known as the cation-\n independent mannose-6-phosphate re-\nceptor (16).\nThe distinctive biological roles of the\nIGF-1 isoforms and the mechanisms that\nregulate their expression have not been\nclearly documented. Several studies have\ninvestigated the expression patterns of\nthese IGF-1 transcripts in skeletal muscle\n(17–19), and there is growing interest vis-\nà-vis the potential role of MGF expression\nin skeletal and cardiac muscle regenera-\ntion and hypertrophy after exercise-\n induced skeletal muscle damage (20) and\nmyocardial infarction (21,22). In addi-\ntion, we have previously reported pre-\nliminary data on the expression of IGF-1\nisoforms in endometriosis at mRNA level\n(23). However, there is little information\nregarding the IGF-1Ec (MGF) expression\nin stromal and glandular epithelium of\nendometriotic lesions and the characteri-\nFigure 1. Human IGF-1 alternative splicing and encoded propeptides. The igf-1 gene\ngives rise to multiple mRNA transcripts by alternative splicing. The different IGF-1 mRNA\ntranscripts encode several precursor proteins, which differ by the length of the amino-\n terminal (signal) peptide and the structure of the extension peptide (E-peptide) on the\ncarboxy-terminal end. The mature IGF-1 peptide results from posttranslational cleavage\nof all precursor polypeptides, by which the signal and the E-peptide are removed. Exons\n5 and 6 encode distinct portions of the E-peptide (called the E-domain) with alternative\ncarboxy-terminal sequences of the extension peptide. The IGF-1Ec splice variant is an\nexon 4-5-6 variant that produces an E-peptide, termed Ec-peptide. The synthetic \nMGF E-peptide that comprises the last 24 C-terminal amino acids (aa) of Ec-peptide is\nshown.  \n\nRESEARCH ARTICLE\nMOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011 | MILINGOS ET AL. | 23\nembedded and processed for paraffin\nsections. The sections were incubated\nwith the same primary antibodies used\nfor the Western blot analyses (i.e., the\npolyclonal anti-MGF antibody at a dilu-\ntion of 1:1,000 in phosphate-buffered\nsaline (PBS) and the monoclonal\nanti–IGF-1) (1:50 dilution, MS-1508;\nThermo Scientific) overnight at 4°C.\nAfter repeated PBS buffer washing,\n secondary biotinylated goat antirabbit\nIgG or goat antimouse IgG (DAB; Dako\nReal EnVision, Glostrup, Denmark)\n antibody was added for 25 min at room\ntemperature, followed again by re-\npeated PBS buffer washes. Visualization\nof the immunocomplex was obtained\nby incubating the sections in a solution\nof 3,3-diaminobenzidine (DAB) in PBS\nfor 10 min. Tissue sections were visual-\nized under light microscopy, and images\nwere captured on a PENTAX ASAHI\ndigital color camera mounted on the\nmicroscope. A qualitative analysis of\nthe tissue sections was then performed\nin the form of positive or negative\nstaining. Negative control staining pro-\ncedures were included in all immuno-\nhistochemical analyses, as described\nelsewhere (25).\nCell Cultures\nHuman KLE cells were obtained by\nthe American Type Culture Collection\n(ATCC, Bethesda, MD, USA) and main-\ntained as subconfluent monolayers in\nculture using Dulbecco’s modified\nEagle’s medium (DMEM/F-12; Cambrex,\nWalkersville, MD, USA) supplemented\nwith 10% fetal bovine serum (FBS; Invit-\nrogen) at 37°C in a humidified atmos-\nphere with 5% CO\n2, with culture media\nbeing replaced every 2–3 d. KLE cells\nwere treated with 0.5 ng/mL up to\n30 ng/mL insulin (Novo Nordisk,\nBagsværd, Denmark), with 0.5 ng/mL\nup to 50 ng/mL of mature IGF-1 peptide\n(rhIGF-1; Chemicon, Temecula, CA, USA)\nand with 0.5 ng/mL up to 50 ng/mL of a\nsynthetic MGF peptide (which comprises\nthe last 24 amino acids of the E-domain\nof human MGF, synthesized and vali-\ndated as previously described [25]; see\ntranscripts have been described else-\nwhere (23).\nProtein Extraction and Western\nAnalysis of IGF-1 and MGF\nThe extracts were analyzed for total\nprotein concentration using the Bradford\nprocedure (Bio-Rad Protein Assay; Bio-\nRad, Hercules, CA, USA). Samples were\nstored in aliquots at –80°C until Western\nblot analysis as previously described (25).\nThe following primary antibodies were\nused for the immunodetection of IGF-1Ec\n(MGF) and IGF-1: MGF, a rabbit antihu-\nman MGF polyclonal antibody (1:10,000\ndilution), which was raised against a syn-\nthetic peptide corresponding to the last 24\namino acids of the E-domain of human\nMGF (IGF-1Ec) and characterized in our\nlaboratory, as has been described else-\nwhere (22); and IGF-1, a mouse mono-\nclonal anti–IGF-1 (1:1,000 dilution) (MS-\n1508; Thermo Scientific, Fremont, CA,\nUSA; molecular weight of antigen:\n~7.6 kDa). After the overnight incubation\nof blots with the primary antibodies,\nmembranes were incubated with a horse-\nradish peroxidase–conjugated secondary\nantirabbit IgG (goat antirabbit, 1:2,000 di-\nlution; Santa Cruz Biotechnology, Santa\nCruz, CA, USA) or antimouse IgG goat\nantimouse (1:2,000 dilution; Santa Cruz\nBiotechnology) for 1 h at room tempera-\nture. Glyceraldehyde 3-phosphate dehy-\ndrogenase (GAPDH) was used as an in-\nternal control to correct for potential\nvariation in the protein loading and to\nnormalize the protein measurements on\nthe same immunoblot. Blots were incu-\nbated with a mouse monoclonal primary\nantibody for GAPDH (1:2,000 dilution;\nSanta Cruz Biotechnology) and with a\nhorseradish peroxidase–conjugated sec-\nondary antimouse IgG (goat antimouse,\n1:2,000 dilution; Santa Cruz Biotechnol-\nogy), and specific band(s) were visual-\nized as described elsewhere (20).\nImmunohistochemical Analysis\nFormaldehyde-fixed eutopic and\n ectopic endometrium (red lesions and\nendometriotic cysts) samples from all\npatients’ biopsies were paraffin wax\nLaparoscopy was performed during the\nproliferative phase of the menstrual cycle\n(fifth to tenth day after menstruation).\nTissue Sampling\nTissue sampling was from normal en-\ndometrium (eutopic), red lesions and/or\nendometriotic cysts. We analyzed 15 tis-\nsue biopsies of endometriotic peritoneal\nlesions (red lesions) and 20 tissue biop-\nsies of endometriotic cysts from 15 and\n20 patients, respectively. From the same\nwomen, normal endometrium was aspi-\nrated using the Cornier device (Labora-\ntoire C.C.D., Paris, France). All patients\nhad stage III–IV endometriosis according\nto revised American Fertility Society\n(rAFS) classification. Tissue biopsies for\nRNA and protein extraction were snap-\nfrozen in liquid nitrogen and then stored\nat –80°C until analysis, whereas biopsies\nfor immunohistochemistry were trans-\nferred to formaldehyde 9%. The diagno-\nsis of endometriosis was confirmed with\nhistological examination of related tissue\nbiopsies. The proliferative phase of the\nmenstrual phase was determined based\non the last menstrual period and con-\nfirmed with histological examination of\nthe eutopic endometrium using the\nNoyes’ criteria (24).\nRNA Extraction and Relative\nQuantitative PCR Analysis\nThe expression of IGF-1 transcripts in\neutopic and ectopic endometrium (red\nlesions and/or endometriotic cysts) and\nin KLE endometrial-like cells was as-\nsessed as previously described (23).\nBriefly, each endometriotic tissue sample\nwas homogenized and total RNA was\nextracted using Trizol Reagent (Invitro-\ngen, Carlsbad, CA, USA) according to\nthe manufacturer’s recommendations.\nThe RNA samples were used for the de-\ntermination of the mRNA of specific\nIGF-1 transcripts by reverse transcription\n(RT) and semiquantitative RT–polymerase\nchain reaction (PCR) procedures. Both\nthese RT and PCR methods have been\ndescribed and extensively validated else-\nwhere (19). Primer sets and PCR condi-\ntions used for the assessment of IGF-1\n\n24 | MILINGOS ET AL. | MOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011\nMGF EXPRESSION IN EUTOPIC AND ECTOPIC ENDOMETRIUM\nFigure 1) in a time-dependent manner\n(i.e., for 24 and 48 h).\nTrypan Blue Assay\nKLE cells were plated at a cell density\nof about 2.3 × 104 cell/well in 24-well\nplates and grown with DMEM/F-12 con-\ntaining 10% FBS. Twenty-four h after\nplating, the media were changed to\nDMEM/F-12 containing 0.5% FBS, and\nmitogens under investigation were\nadded in a dose-dependent manner (ma-\nture IGF-1, MGF E-peptide and insulin).\nThe actual living KLE cell number was\nmeasured at different time intervals (24\nand 48 h) using the Trypan Blue exclu-\nsion assays, as previously described (26).\nIGF-1R and IR siRNA Knock Out\nTo investigate if the synthetic MGF E-\npeptide acts on KLE cells via the IGF-1R–\nor IR-mediated pathway, IGF-1R and IR\nexpression was silenced in KLE cells\nusing the commercially available Stealth\nsiRNA technology (Invitrogen). Three\ndifferent 25-mer siRNA molecules were\nexamined in each case for their potential\nto knock out (KO) the expression of IR\nand that of IGF-1R in KLE cells. It was\ndetermined that the most efficient KO of\nthe IR was obtained by using the\nACAAACUGCCCGUUGAUGACGGUGG\nsiRNA duplex at a concentration of\n40 pmol by using the reverse transfection\nmethod. In the case of IGF-1R KO, the\nmolecule of choice was the UCUUC\nAAGGGCAAUUUGCUCAUUAA\nsiRNA duplex, at a concentration of\n50 pmol, again by using reverse transfec-\ntion according to the manufacturer’s in-\nstructions. As a negative control, we\nused a universal negative control stealth\nsiRNA (Invitrogen). In brief, KLE cells\nwere grown in 10% DMEM/F-12 media.\nThe transfection mixture was obtained\nby diluting the 40 pmol of the siRNA du-\nplex in 100 μL OptiMem serum-free\n medium (Invitrogen) in a well of a 24-well\nplate, followed by the addition of 2 μL\nlipofectamine RNAiMAX (Invitrogen).\nAfter 20 min, 500 μL of the trypsinized\nKLE cells was added to the mixture.\nForty-eight hours after the KO, the\nmedia switched to DMEM 0.5% FBS, and\nafter 24 h, the IR KO cells were exposed\nto either insulin or MGF E-peptide,\nwhereas the IGF-1R KO cells were ex-\nposed to mature IGF-1 or MGF E-peptide\nfor 24 and 48 h in triplicate determina-\ntions. The viable cells were counted\nusing the Trypan Blue exclusion assay.\nExpression of IGF-1R and IR\nThe expression level of IGF-1R and IR\ntranscripts, after the siRNA IGF-1R KO and\nIR KO in KLE cells, was assessed by quanti-\ntative real-time PCR (qRT-PCR). The KO\nmRNA levels were determined 48 h after the\nsiRNA KO according to the manufacturer’s\ninstructions. As an internal control, we used\nGAPDH and β-actin. The validation of the\nproduct identity was obtained by the melt-\ning curve. Quantitative RT-PCR to examine\nthe levels of expression of IGF-1R and IR\nwas carried out before and after the siRNA\nIGF-1R or IR KO in KLE cells. Briefly, RT-\nPCR data quantification analysis was carried\nout in the forms of melting and amplifica-\ntion curves, cycle threshold (Ct) values and\nnormalized gene expression (Delta Delta Ct\n[ddCt]), using the Bio-Rad IQ5 optical soft-\nware 2.0. The primers used in the reactions\nwere generated using the FastPCR program\nand were as follows: IGF-1R forward:\n ACCCGGAGTACTTCAGCGC; IGF-1R re-\nverse: CACAGAAGCTTCGTTGAGAA; IR\nforward: ACTCTCAGATCCTGA\n AGGAGCTGGA; IR reverse: AGTGT\nTGGGGAAAGCTGCCAC. The set of\nprimers for IR was designed to detect\nboth IR isoforms in a single PCR. The\nPCR conditions were the same in both\ncases: 95°C for 30 s × 1 cycle, 94°C for\n20 s, 60°C for 30 s, 72°C for 30 s × 35 cy-\ncles and 72°C for 5 min.\nStatistical Analysis\nChanges in cell numbers were assessed\nusing analysis of variance (ANOVA)\n(SPSS v. 11 statistical package; SPSS, Chi-\ncago, IL, USA). Where significant F ratios\nFigure 2. Expression of the different IGF-1 transcripts (IGF-1 Ea, IGF-1 Eb and IGF-1 Ec\n[MGF]) in eutopic endometrium (EU), peritoneal red lesion (PE) and ovarian endometriotic\ncyst (OvE). Representative PCR gel images demonstrate the differential mRNA expression\nof the IGF-1 transcripts in PE and in OvE compared with EU (upper panel). In the lower\npanel, PCR relative quantification is presented. Values of PE and OvE were normalized to\neach corresponding ribosomal 18S and expressed as percentage differences (%) from EU\nlevels (means ± SD, PE: n = 15, OvE: n = 20). *Significantly different from EU (\nP < 0.01). \n\nRESEARCH ARTICLE\nMOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011 | MILINGOS ET AL. | 25\nwere found (P < 0.05), the means were\ncompared using Tukey post hoc tests. A\nStudent t test was used to evaluate tran-\nscriptional and translational differences\nin IGF-1 isoform expression between eu-\ntopic endometrium and endometriotic\ncysts or red lesions. All data are pre-\nsented as mean ± SD. The level of signifi-\ncance was set at P < 0.05.\nRESULTS\nThe expression of IGF-1 mRNA tran-\nscripts was found to be significantly\nlower in the endometriotic cysts com-\npared with that of the eutopic en-\ndometrium and that of the red lesions, as\nassessed by semiquantitative PCR meth-\nods (Figure 2). Similar patterns of the IGF-\n1 transcripts translation were also de-\ntected in red lesions and endometriotic\ncysts compared with eutopic endo -\nmetrium at the protein level (Figure 3A,\nB). The immunohistochemical analysis re-\nvealed that in the eutopic endo metrium\n(Figure 4A, E) and the endometriotic cysts\n(Figure 4C, G), the IGF-1 transcripts were\nexpressed only in the stromal cells and\nnot in the glandular epithelium, whereas\nin the red lesions, the IGF-1 transcripts\nwere expressed not only in the stroma but\nalso in the glandular cells (Figure 4B, F).\nAll 15 red lesion biopsies were positive\nfor glandular MGF/ IGF-1 expression,\nwhereas all 15 biopsies from the eutopic\nendometrium (of the same women) were\nnegative for glandular MGF/IGF-1 ex-\npression. The stroma was steady positive\nfor MGF/ IGF-1 expression in all eutopic\nand ectopic endometrial biopsies; how-\never, endometriotic cysts did express con-\nsiderably lower MGF/IGF-1 levels, as\nnoted by immunohistochemical analysis.\nIn order to characterize the IGF-1Ec\nposttranslational products (mature IGF-1\npeptide and synthetic MGF E-peptide) in\nvitro, we initially characterized the KLE\ncells. We documented that the KLE en-\ndometrial-like cells express all three IGF-\n1 mRNA transcripts, which are certainly\ntranslated to pro–IGF-1 and pro–IGF-1Ec\n(MGF) products at protein level (Fig-\nure 5A, B).\nBecause the actions of IGF-1 can be me-\ndiated not only via its high-affinity IGF-1R\nbut also via IRs as well as hybrid IGF-\n1R/IR, we experimentally engineered KLE\ncells with silenced IGF-1R and IR expres-\nsion, using siRNA methods, to further\nFigure 3. Representative Western blots\ndemonstrating the expression of IGF-1\nand IGF-1Ec (MGF) in (A) peritoneal red\nlesion (PE) and in (B) ovarian endometri-\notic cyst (OvE) samples examined in rela-\ntion to eutopic endometrium (EU).\nFigure 4. (A–D) Cytoplasmic localization of IGF-1 in stromal cells (SC) in eutopic en-\ndometrium (A), endometriotic lesion (B) and endometriotic cyst (C). Note the absence of\nstaining in glandular epithelium (GE) in eutopic endometrium as opposed to the positive\nstaining of glandular epithelium in endometriotic lesions. (D) Negative control. (E–H) Cyto-\nplasmic localization of MGF (IGF-1Ec) in stromal cells (SC) of tissue biopsies from eutopic\nendometrium (E), endometriotic lesion (F) and endometriotic cyst biopsies (G). Note the\nabsence of MGF staining in glandular epithelium biopsies (GE) of the eutopic en-\ndometrium and endometriotic cyst as opposed to the positive staining of glandular ep-\nithelium in endometriotic lesion. (H) Negative control. Solid arrows represent stromal cells;\nhollow arrows represent glandular epithelium. \nFigure 5. Expression of the differentigf-1\ngene transcripts in KLE endometrial-like\ncells. (A) PCR products (that is, amplified\ntarget cDNAs) from the different primer\nsets and PCR conditions used for the de-\ntection of IGF-1 transcripts at the mRNA\nlevel. An equal amount of each PCR prod-\nuct was loaded onto a 2% agarose gel\nand separated by electrophoresis. (B)\nTranslational products of the different IGF-1\nmRNA transcripts were detected by West-\nern blot analysis using antibodies specific\nfor anti–IGF-1 and anti–IGF-1Ec (MGF). \n\n26 | MILINGOS ET AL. | MOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011\nMGF EXPRESSION IN EUTOPIC AND ECTOPIC ENDOMETRIUM\ncharacterize MGF E-peptide  actions in\nKLE cells. Thus, we generated transfectans\nof siRNA IGF-1R KO KLE cells and siRNA\nIR KO KLE cells. Indeed, we achieved ap-\nproximately 60–80% reduction of IR\nmRNA expression (for both IR transcripts;\nIR-A and IR-B) compared with the respec-\ntive expression levels assessed in control\nKLE cells (Figure 6A, D, G). Similar results\nwere obtained in the siRNA IGF-1R KLE\ncells (Figure 6B, E, F). Analyses of β-actin\n(Figure 6C) and GAPDH expression (not\nshown) were used as internal controls for\nnormalization in all cases.\nUsing these KLE transfectans, we were\nable to show that exogenous IGF-1 and\ninsulin administration did not stimulate\nthe IGF-1R KO and IR KO KLE cells. On\nthe contrary, IGF-1 and insulin stimu-\nlated the growth of parental KLE cells\n(Table 1). Interestingly, MGF E-peptide\nstimulated the proliferation of parental\nand IGF-1R KO and IR KO KLE cells (see\nTable 1).\nFigure 6. Characterization of the degree of reduction of IR expression in IR KO KLE cells (A) and of IGF-1R expression in IGF-1R KO KLE\ncells (B) by qRT-PCR. IR KO and IGF-1R KO lines represent the melting curves of IR (IR-A and IR-B isoforms) and IGF-1R in IR KO  KLE and\nIGF-1R KO KLE cells, respectively, compared with the lines of IR and IGF-1R in control KLE cells. The amplification curves and Cts (D, E) as\nwell as normalized expression (ddCt) charts (F , G) are also shown. The degree of reduction of IR and IGF-1R expression was from 60% up\nto 80% in this cell line. Normalization in all the cases was carried out by β-actin (C). \n\nRESEARCH ARTICLE\nMOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011 | MILINGOS ET AL. | 27\nDISCUSSION\nIn this study, we documented that eu-\ntopic and ectopic (endometriotic cysts and\nred lesions) endometrium obtained from\nwomen with endometriosis as well as\nhuman KLE endometrial-like cells express\nIGF-1 transcripts. We have defined that\nthe IGF-1Ec transcript is expressed both at\nthe mRNA and protein level. This particu-\nlar IGF-1 transcript has been associated\nwith regeneration mechanisms of skeletal\nmuscle and myocardial cells (20–22).\nSemiquantitative analysis of the IGF-1\ntranscript expression using PCR methods\nrevealed that endometriotic cysts ex-\npressed IGF-1 transcripts at a significantly\nlower level than eutopic endometrium\nand red lesions. These findings were in-\nline with our preliminary data previously\npublished (23). Our findings could be ex-\nplained by the fact that even though en-\ndometriotic cysts represent a feature of\nadvanced disease, they are characterized\nby the presence of fibrosis and low levels\nof active endometriotic tissue. This result\nis consistent with the natural history of\nthe disease, during which active en-\ndometriotic tissue is substituted by fi-\nbrotic tissue accounting for the increased\nscarring and adhesion formation found in\nlate stages of endometriosis (23). This re-\nsult is also consistent with the results of\nour previous studies, where we docu-\nmented increased expression of other\ncomponents of the IGF bioregulatory sys-\ntem, which includes IGFs/uPA/plas-\nmin/IGFBP-3 expression (6,7,27).\nIn this study, we used specific antibod-\nies to identify the expression of IGF-1 and\nMGF in endometriotic biopsies by im-\nmunohistochemical and Western blot\nanalyses. Because IGF-1 peptide is a com-\nmon product of all three IGF-1 transcripts,\nanti–IGF-1 antibody can detect the expres-\nsion of pro–IGF-1 peptide from any IGF-1\ntranscript. On the contrary, our anti-MGF\nantibody identifies the expression of the\nIGF-1Ec (MGF) transcript only. The im-\nmunohistochemical analysis of IGF-1\ntranscripts posttranslational products re-\nvealed that in eutopic endo metrium and\nendometriotic cysts, IGF-1 and IGF-1Ec\n(MGF) were expressed only in stroma\ncells but not in glandular cells. In contrast,\nin red lesions, there was positive staining\nnot only in stroma cells but also in glan-\ndular epithelium. Even though histologi-\ncal diagnosis of endo metriosis requires\nthe presence of stroma and glandular cells\nin tissue biopsies, the proportion of\nstroma/glands in endometriotic tissue is\nnot constant, and it has been suggested\nthat lesions related to more active forms\nof endo metriosis (for example, red le-\nsions) present a higher proportion of glan-\ndular cells (28). This was evident in our\nbiopsies as well, as histological examina-\ntion showed increased proportion of glan-\ndular epithelium in red lesions compared\nwith endometriotic cysts. This could ac-\ncount for the increased IGF-1 transcripts\nexpression (although not significant) in\nred lesions as it was documented by semi-\nquantitative PCR analysis in our study.\nThe expression of IGF-1 and IGF-1Ec in\nthe glandular epithelium of only en-\ndometriotic lesions and not in eutopic en-\ndometrium and endometriotic cysts could\nfavor our hypothesis that IGF-1 and IGF-\n1Ec isoforms are associated with active\nendometriosis, and their action in ectopic\nendometriotic cells could be involved in\nthe progression of the disease and evolu-\ntion of endometriotic lesions.\nThe IGF-1 stimulates the growth and\ndifferential function of endometrial cells\nvia the IGF-1R, and possibly via several\natypical receptors, including the hybrid\nIR/IGF-1R. The latter is composed of an\nIR hemi-receptor linked to an IGF-IR\nhemi-receptor and has been reported to\nhave an important role in cancer biology\n(29–31). Recently, the two IR isoforms\n(IR-A and IR-B) have been reported that\nare overexpressed in cancer tissues (32),\nwhereas the expression of IGF-1R has\nbeen previously characterized in KLE\ncells in our laboratory (7). Therefore, aim-\ning to the characterize the MGF E-peptide\nactions in KLE cells, we performed a se-\nries of silencing experiments of these\nmajor receptors involved in the IGF-\n mediated actions. Our data suggested that\nsilencing of the IGF-1R and IR expression\nin KLE cells did not have an important ef-\nfect on the proliferative activity of the ex-\nogenous MGF E-peptide in vitro, thus sug-\ngesting that synthetic MGF E-peptide\naction is apparently mediated via an IGF-\n1R–independent, IR- independent mecha-\nnism. Because the IR/IGF-1R hybrid re-\nTable 1. The effects of 48 h of treatment with mature IGF-1, insulin and synthetic MGF E-peptide on KLE cell proliferation, as assessed by\nTrypan blue exclusion assays (cell number × 104). \nIGF-1 (50 ng/mL) in MGF (50 ng/mL) in IGF-1 (50 ng/mL) in MGF (50 ng/mL) in Control siRNA-transfected\nuntransfected KLE cells untransfected KLE cells IGF1R siRNA KLE cells IGF1R siRNA KLE cells KLE cells\n112.5 ± 8.66 103.33 ± 5.20 76.25 ± 5.30 95.83 ± 3.81 73.75 ± 5.30\nab c\nInsulin (30 ng/mL) in MGF (50 ng/mL) in Insulin (30 ng/mL) in MGF (50 ng/mL) in Control siRNA-transfected\nuntransfected KLE cells untransfected KLE cells IR siRNA KLE cells IR siRNA KLE cells KLE cells\n32.9 ± 5.49 34.37 ± 5.15 12.5 ± 2.5 28.87 ± 4.73 11.25 ± 2.5\nba b\nThe mitogenic activity of the IGF-1 and insulin was blocked in IGF-1R siRNA KLE cells and in the IR siRNA KLE cells, respectively, whereas\nMGF E-peptide mitogenic actions were not affected in IGF-1R siRNA KLE cells and in the IR siRNA KLE cells. These data suggested that\nMGF actions are possibly mediated via an IGF-1R–independent, IR-independent and hybrid IGF-1R/IR–independent mechanism in KLE\nendometrial-like cells. Significantly different from control-siRNA transfected KLE cells: \naP < 0.001; bP < 0.01; cP < 0.05.\n\n28 | MILINGOS ET AL. | MOL MED 17(1-2)21-28, JANUARY-FEBRUARY 2011\nMGF EXPRESSION IN EUTOPIC AND ECTOPIC ENDOMETRIUM\nceptor consists of IR and IGF-1R hemi-re-\nceptors, the silencing of the IR or the IGF-\n1R is expected to block the formation of\nthe hybrid receptor. Therefore, our experi-\nments suggested that mitogenic activity of\nthe synthetic MGF E-peptide is mediated\nvia another receptor molecule.\nFurther evidence for such autonomous\nactions of the synthetic MGF E-peptide\nwas provided by our recent data, which\nrevealed that MGF E-peptide activated\nERK1/2 phosphorylation but did not ac-\ntivate AKT phosphorylation in skeletal\nmuscle–like and myocardial-like cells\n(20,22). This particular phosphorylation\npattern generated by the MGF E-peptide\nis in agreement with the trypan blue ex-\nclusion assays in KLE cells, thus suggest-\ning that MGF E-peptide activity is via an\nIGFR/IR-independent mechanism and\nvia an as yet unidentified molecule.\nIn conclusion, our data suggest the pos-\nsible role of IGF-1Ec (MGF) expression in\nendometriosis. This is supported by the\npreferential expression of this IGF-1 tran-\nscript in glandular epithelial cells in ec-\ntopic endometrium only (red lesions).\nConceivably, this preferential MGF ex-\npression generates posttranslational prod-\nucts IGF-1 and MGF E-peptide, with the\nlatter being capable of stimulating the\nproliferation of endometrial-like cells via\nan IGF-1R–independent, IR-independent\nand hybrid IGF-1R/IR–independent\nmechanism. These data suggest that there\nmay be a role for MGF in the pathogene-\nsis of endometriosis that is autonomous\nand independent from the IGF system.\nDISCLOSURE\nThe authors declare that they have no\ncompeting interests as defined by Molec-\nular Medicine, or other interests that\nmight be perceived to influence the re-\nsults and discussion reported in this\npaper.\nREFERENCES\n1. Giudice LC, Kao LC. (2004) Endometriosis. Lancet\n364:1789–99.\n2. Gao X, Outley J, Botteman M, Spalding J, Simon\nJA, Pashos CL. (2006) Economic burden of en-\ndometriosis. Fertil. Steril. 86:1561–72.\n3. Gebel HM, Braun DP , Tambur A, Frame D, Rana N,\nDmowski WP . (1998) Spontaneous apoptosis of en-\ndometrial tissue is impaired in women with en-\ndometriosis. Fertil. Steril.69:1042–7.\n4. Koutsilieris M, Mastrogamvrakis G, Lembessis P ,\nSourla A, Miligos S, Michalas S. (2001) Increased\ninsulin-like growth factor 1 activity can rescue\nKLE endometrial-like cells from apoptosis. Mol.\nMed. 7:20–6.\n5. Kim JG, Suh CS, Kim SH, Choi YM, Moon SY, Lee\nJY. (2000) Insulin-like growth factors (IGFs), IGF-\nbinding proteins (IGFBPs), and IGFBP-3 protease\nactivity in the peritoneal fluid of patients with and\nwithout endometriosis. Fertil. Steril. 73:996–1000.\n6. Koutsilieris M, Lavergne E, Lemay A. (1997) Asso-\nciation of protease activity against IGFBP-3 with\nperitoneal fluid mitogens: possible implications for\nthe ectopic growth of endometrial cells in women\nwith endometriosis. Anticancer Res.17:1239–44.\n7. Lembessis P , Kalariti N, Koutsilieris M. (2004)\nGlucocorticoid receptor function suppresses in-\nsulin-like growth factor 1 activity in human KLE\nendometrial-like cells. In Vivo 18:43–7.\n8. Siegfried JM, Kasprzyk PG, Treston AM, Mulshine\nJL, Quinn KA, Cuttitta F. (1992) A mitogenic pep-\ntide amide encoded within the E peptide domain\nof the insulin-like growth factor IB prohormone.\nProc. Natl. Acad. Sci. U. S. A. 89:8107–11.\n9. Gilmour RS. (1994) The implications of insulin-\nlike growth factor mRNA heterogeneity. J. En-\ndocrinol. 140:1–3.\n10. Chew SL, Lavender P , Clark AJ, Ross RJ. (1995)\nAn alternatively spliced human insulin-like\ngrowth factor-I transcript with hepatic tissue ex-\npression that diverts away from the mitogenic\nIBE1 peptide. Endocrinology 136:1939–44.\n11. Shavlakadze T, Winn N, Rosenthal N, Grounds\nMD. (2005) Reconciling data from transgenic\nmice that overexpress IGF-I specifically in skele-\ntal muscle. Growth Horm. IGF Res. 15:4–18.\n12. Kuo YH, Chen TT. (2002) Novel activities of pro-\nIGF-I E peptides: regulation of morphological\ndifferentiation and anchorage-independent\ngrowth in human neuroblastoma cells. Exp. Cell\nRes. 280:75–89\n13. Samani AA, Yakar S, LeRoith D, Brodt P . (2007)\nThe role of the IGF system in cancer growth and\nmetastasis: overview and recent insights. Endocr.\nRev. 28:20–47.\n14. Jones JI, Clemmons DR. (1995) Insulin-like\ngrowth factors and their binding proteins: bio-\nlogical actions. Endocr. Rev. 16:3–34.\n15. LeRoith D, Werner H, Beitner-Johnson D, Roberts\nCT. (1995) Molecular and cellular aspects of the\ninsulin-like growth factor I receptor. Endocr. Rev.\n16:143–63.\n16. Kornfeld S. (1992) Structure and function of the\nmannose-6-phosphate/insulinlike growth factor\nII receptors. Annu. Rev. Biochem. 61:307–30.\n17. Hameed M, Orrell RW, Cobbold M, Goldspink G,\nHarridge SD. (2003) Expression of IGF-I splice\nvariants in young and old human skeletal muscle\nafter high resistance exercise. J. Physiol.547:247–54.\n18. Bickel CS, Slade J, Mahoney E, Haddad F, Dudley\nGA, Adams GR. (2005) Time course of molecular\nresponses of human skeletal muscle to acute bouts\nof resistance exercise.J. Appl. Physiol. 98:482–8.\n19. Kim JS, Cross JM, Bamman MM. (2005) Impact of\nresistance loading on myostatin expression and\ncell cycle regulation in young and older men and\nwomen. Am. J. Physiol. Endocrinol. Metab.\n288:E1110–9.\n20. Philippou A, et al. (2009) Expression of IGF-1 iso-\nforms after exercise-induced muscle damage in\nhumans: characterization of the MGF E peptide\nactions in vitro. In Vivo 23:567–75.\n21. Carpenter V , et al. (2008) Mechano-growth factor\nreduces loss of cardiac function in acute myocar-\ndial infarction. Heart Lung Circ. 17:33–9.\n22. Stavropoulou A, Halapas A, Sourla A, Philippou\nA, Papageorgiou E, Papalois A, Koutsilieris M.\n(2009) IGF-1 expression in infarcted myocardium\nand MGF E peptide actions in rat cardiomy-\nocytes in vitro. Mol. Med. 15:127–35.\n23. Milingos D, et al. (2006) Insulin-like growth fac-\ntor-1 isoform mRNA expression in women with\nendometriosis: eutopic endometrium versus en-\ndometriotic cyst. Ann. N. Y. Acad. Sci.\n1092:434–9.\n24. Noyes RW, Hertig AT, Rock J. (1975) Dating the en-\ndometrial biopsy. Am. J. Obstet. Gynecol.122:262–3.\n25. Philippou A, Stavropoulou A, Sourla A, Pissimis-\nsis N, Halapas A, Maridaki M, Koutsilieris M.\n(2008) Characterization of a rabbit antihuman\nmechano growth factor (MGF) polyclonal anti-\nbody against the last 24 amino acids of the E do-\nmain. In Vivo 22:27–35.\n26. Koutsilieris M, Rabbani SA, Goltzman D. (1987)\nEffects of human prostatic mitogens on rat bone\ncells and fibroblasts. J. Endocrinol. 115:447–54.\n27. Koutsilieris M, Sourla A, Pelletier G, Doillon CJ.\n(1994) Three-dimensional type I collagen gel sys-\ntem for the study of osteoblastic metastases pro-\nduced by metastatic prostate cancer. J. Bone\nMiner. Res. 9:1823–32.\n28. Donnez J, Squifflet J, Casanas-Roux F, Pirard C,\nJadoul P , van Langendonckt A. (2003) Typical\nand subtle atypical presentations of endometrio-\nsis. Obstet. Gynecol. Clin. N. Am. 30:83–93.\n29. Bogdanos J, Karamanolakis D, Tenta R, Tsintavis\nA, Milathianakis C, Mitsiades C, Koutsilieris M.\n(2003) Endocrine/paracrine/autocrine survival\nfactor activity of bone microenvironment partici-\npates in the development of androgen ablation\nand chemotherapy refractoriness of prostate can-\ncer metastasis in skeleton. Endocr. Relat. Cancer\n10:279–89.\n30. Koutsilieris M, Mitsiades C, Sourla A. (2000) In-\nsulin-like growth factor I and urokinase-type\nplasminogen activator bioregulation system as a\nsurvival mechanism of prostate cancer cells in\nosteoblastic metastases: development of anti-\n survival factor therapy for hormone-refractory\nprostate cancer. Mol. Med. 6:251–67.\n31. Koutsilieris M. (1992) Pathophysiology of uterine\nleiomyomas. Biochem. Cell Biol. 70:273–8.\n32. Cox ME, et al. (2009) Insulin receptor expression\nby human prostate cancers. Prostate 69:33–40.","source_license":"CC0","license_restricted":false}