UBC Faculty Research and Publications

HOX cofactors expression and regulation in the human ovary Ota, Takayo; Asahina, Haruka; Park, Se-Hyung; Huang, Qing; Minegishi, Takashi; Auersperg, Nelly; Leung, Peter C Oct 30, 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12958_2008_Article_455.pdf [ 1.56MB ]
JSON: 52383-1.0220749.json
JSON-LD: 52383-1.0220749-ld.json
RDF/XML (Pretty): 52383-1.0220749-rdf.xml
RDF/JSON: 52383-1.0220749-rdf.json
Turtle: 52383-1.0220749-turtle.txt
N-Triples: 52383-1.0220749-rdf-ntriples.txt
Original Record: 52383-1.0220749-source.json
Full Text

Full Text

ralReproductive Biology and ssBioMed CentEndocrinologyOpen AcceResearchHOX cofactors expression and regulation in the human ovaryTakayo Ota1, Haruka Asahina2, Se-Hyung Park1, Qing Huang1, Takashi Minegishi2, Nelly Auersperg1 and Peter CK Leung*1Address: 1Department of Obstetrics and Gynecology, the University of British Columbia (UBC), Vancouver, BC, V6H 3V5, Canada and 2Department of Gynecology and Reproductive Medicine, Gunma University Graduate School of Medicine, Gunma, 371-8511, JapanEmail: Takayo Ota - takayo.ota@gmail.com; Haruka Asahina - harukaasahina7@hotmail.com; Se-Hyung Park - sehyungpark@hotmail.com; Qing Huang - dr.qinghuang@gmail.com; Takashi Minegishi - tminegis@showa.gunma-1.ac.jp; Nelly Auersperg - auersper@interchange.ubc.ca; Peter CK Leung* - peleung@interchange.ubc.ca* Corresponding author    AbstractBackground: HOX cofactors enhance HOX binding affinities and specificities and increase HOX'sunique functional activities. The expression and the regulation of HOX cofactors in human ovariesare unknown.Methods: In this study, the expression of HOX cofactors, PBX1, PBX2, and MEIS1/2, wereexamined by using RT-PCR, immunofluorescence in cultured immortalized human granulosa(SVOG) cells. The distribution of these HOX cofactors in human ovaries was examined byimmunohistochemistry. The effects of growth differentiation factor-9 (GDF-9) and follicle-stimulating hormone (FSH) on PBX2 in SVOG cells were investigated by western blot analysis.Binding activities of HOXA7 and PBX2 to the specific sequences in granulosa cells weredetermined by electrophoretic mobility shift assay (EMSA).Results and conclusion: In SVOG cells, PBX1, PBX2 and MEIS1/2 were expressed during cellculture. In normal human ovaries, PBX1 and MEIS1/2 were expressed in granulosa cells atessentially all stages of follicular development. These cofactors were expressed in the nuclei of thegranulosa cells from the primordial to the secondary follicles, whereas beyond multilayered follicleswas observed in the cytoplasm. The co-expression of PBX1 and MEIS1/2 in granulosa cells innormal human ovaries suggested that MEIS1/2 might control PBX1 sublocalization, as seen in othersystems. PBX2 was not expressed or weakly expressed in the primordial follicles. From the primaryfollicles to the preovulatory follicles, PBX2 expression was inconsistent and the expression wasfound in the granulosa cell nuclei. The PBX2 expression pattern is similar to HOXA7 expressionin ovarian follicular development. Furthermore, FSH down-regulated, GDF-9 did not change PBX2expression, but co-treatment of the granulosa cells with FSH and GDF-9 up-regulated PBX2expression. These results implicated a role for PBX2 expression in the steroidogenic activities ofgranulosa cells in humans. Moreover, PBX2 and HOXA7 bound together to the Pbx sequence, butnot to the EMX2 promoter sequence, in SVOG cells. Our findings indicate that HOX cofactorsexpression in normal human ovary is temporally and spatially specific and regulated by FSH andGDF-9 in granulosa cells. HOX proteins may use different HOX cofactors, depending on DNAPublished: 30 October 2008Reproductive Biology and Endocrinology 2008, 6:49 doi:10.1186/1477-7827-6-49Received: 18 August 2008Accepted: 30 October 2008This article is available from: http://www.rbej.com/content/6/1/49© 2008 Ota et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 9(page number not for citation purposes)sequences that are specific to the granulosa cells.Reproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49BackgroundTranscription factors play important roles in oogenesisand folliculogenesis [1,2]. Numerous studies haveemphasized the importance of ovarian-specific transcrip-tion factor genes during ovarian follicular development.However, non-ovarian-specific genes are also expressedduring ovarian follicular development, and their potentialfunctions in folliculogenesis, such as steroidogenesis, arenot known [3-5].Homeobox genes are transcription factors that encodehomeodomain-containing DNA-binding proteins thatspecify the anterior-posterior orientation of a variety oforgans during embryonic development and regulate dif-ferentiation in the adult tissues. The homeodomain ishighly conserved in Hox proteins. Hox share similar DNAbinding specificities in vitro. In vivo, Hox cofactorsenhance their target DNA binding affinities and specifici-ties [6,7]. Homeobox domains in Hox cofactors contain athree-amino acid loop extension (TALE), which can beclassified into two groups: Pbx and Meis/Prep. Pbx bindsto groups 1 to 10 (out of 13 paralogous Hox groups) andMeis proteins bind to groups 9–13. However, recent stud-ies have shown that Meis1 also can bind to anterior Hoxproteins [8] in vitro. In addition, Meis proteins interactdirectly with Pbx and participate in the DNA bound Hoxcomplex and regulate the sublocalization of Pbx [7].Pbx has four subclasses: Pbx1–4. Pbx1 was first identifiedin acute pre-B-cell leukemias [9]. The Pbx2, Pbx3, Pbx4genes were subsequently identified [10,11]. In the femalereproductive system, Pbx1 is expressed in the coelomicepithelium [12]. Pbx1 encodes two spliced variants Pbx1aand Pbx1b. Pbx1a is not expressed in the developingfemale reproductive systems, whereas Pbx1b is expressedin the Müllerian ducts at embryonic day 14.5 (E14.5)[13]. Pbx1-deficient mice die at E15.5-E16.5, with axialskeletal malformations and abnormalities of multipleorgans [14], including improper differentiation of thegonads and an absence of Müllerian ducts [12]. In devel-oping embryos, at E14.5, weak Pbx3 expression wasobserved throughout the ovary [15]. Pbx3-deficient micesurvive to term but die within a few hours of birth [16],and there is no description about reproductive systemanomalies thus far. The expression patterns of Pbx1 andPbx3 are overlapped during embryonic development, andthey could have redundant functions. In contrast, Pbx2 isnot crucial for development, which implies possible com-pensation by other Pbx isoforms [17].Three Meis isoforms (Meis1, Meis2 and Meis 3) are allexpressed in newborn and 10-week old mice in the ovary[8]. Meis1 null mice die at E14.5 and are deficient in nor-reports. Immunohistochemical studies have shown that,in newborn female reproductive systems, Meis1 isexpressed throughout the ovary, uterus, cervix and vagina[8]. Meis2 is not expressed in the gonad primodium atE9.5–11.5, but, in adult mice, the Meis2c and Meis2d iso-forms are expressed [20]. MEIS1 and MEIS2 are expressedin the nuclei of normal adult human ovarian surface epi-thelium; however, their functions in the ovary have notbeen elucidated [21].In this study, we examined the immunohistochemical dis-tribution of HOX cofactors in normal human premeno-pausal ovaries. We also examined PBX1, PBX2 andMEIS1/2 expression in immortalized granulosa cells invitro. Furthermore, because PBX2 expression wasobserved to be similar to HOXA7 as seen in our previousstudy [3], we focused on PBX2 and investigated the regu-lation of PBX2 by FSH and GDF-9 treatment and on theDNA binding activities of PBX2 and HOXA7 in granulosacells.MethodsMaterialsThe antibodies for PBX1 (sc-889), PBX2 (sc-890), MEIS1/2 (sc-10599), MEIS1/2 blocking peptide (sc-10599p) andactin (C-11) were purchased from Santa Cruz Biotechnol-ogy Ltd. (Santa Cruz, CA). The epitopes of the MEIS1/2antibody was designed for MEIS1, however, has 94% sim-ilarities with MEIS2. Human recombinant FSH was pur-chased from Dr. A.F. Parlow (National Hormone andPituitary Program, CA). GDF-9 was kindly provided byDr. A. J.W. Hsueh (Stanford University, CA).TissuesThe use of human tissues and cells was approved by thecommittee for ethical review of research involving humansubjects of the University of British Columbia. Paraffinsections were obtained from grossly and histologicallynormal ovaries of 7 premenopausal women, collected bythe Department of Pathology (UBC) with patient consent.Premenopausal status was confirmed by the presence offollicles or corpora lutea. Carcinoma was excluded in allof the ovarian samples used in this study by histopatho-logical assessment.Cell cultureSV40 Tag-immortalized human granulosa cells (SVOG)were generated, as described previously [22], by transfec-tion of the SV40 early genes Tag/tag into human granulosacells obtained from IVF procedures. SVOG cells weremaintained in 199/MCDB 105/10% FBS with 0.4 μg/mlhydrocortisone (BD Biosciences Clontech, MountainView, CA). A cervical cancer cell line, Hela, and an eryth-Page 2 of 9(page number not for citation purposes)mal capillaries, megakaryotes and eyes [18,19], but therewas no description about the reproductive system in theseroleukemic cell line, K562, were purchased from theAmerican Type Culture Collection (ATCC). Ishikawa cellsReproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49were provided by Dr. Taylor (Yale University). Hela cellswere maintained in DMEM (Invitrogen, Burlington, ON)supplemented with 10% FBS (Hyclone Laboratories Ltd.,Logan, UT). K562 cells were maintained in RPMI 1640medium (Invitrogen) with 15% FBS. Ishikawa cells weremaintained in DMEM/F-12 with 5% FBS. All cells weregrown in a 5% CO2/air atmosphere, and Hela, Ishikawaand SVOG cells were passaged using 0.06% trypsin/0.01%EDTA in Ca2+-, Mg2+-free Hanks' BSS.RT-PCRTotal RNA was extracted from cultured cells using theRNeasy kit (Qiagen Inc., Mississauga, ON) according tothe manufacturer's procedure. Complementary DNA wastranscribed from 1.5 μg total RNA using a First StrandcDNA synthesis kit (Amersham, Oakville, Ont., Canada)and used as a template for polymerase chain reaction(PCR). All PCR primers span introns in order to detectspecific mRNA sequences. The forward and reverse prim-ers used were as follows: PBX1 (NM_002585): 5'-CCACGTGATGAATCTCCTGCGAGAG-3' and 5'-TCACT-GTATCCTCCTGTCTGGCTGA-3', PBX2 (NM_002586):5'-CTGGTTTGGCAACAAGAGGATTCGC-3' and 5'-TGGAGGTATCAGAGTGAACACTCCC-3' and MEIS1(BC043503): 5'-AAGGTGATGGCTTGGACAA-3' and 5'-GGCTGCACTATTCTTCTCCG-3'. The expected size ofPCR products using these sets of primers are 627 bp forPBX1a, 414 bp for PBX1b, 411 bp for PBX2, and 259 bpfor MEIS1. The amplification reaction was carried out inthe linear range of the logarithmic phase unless otherwisespecified. Each cycle consisted of denaturation at 95°C for30 s, primer annealing at 55°C for 30 s, extension at 72°Cfor 60 s and a final extension at 72°C for 5 min in a DNAthermal cycler (Mini cyclerTM, PTC-100 TM, MJ-Research,Bio-Rad). PCR products were verified by sequence analy-sis.Western blot analysisCells were washed with ice-cold PBS and lysed by coldlysis buffer (1% Triton X-100, 0.5% sodium deoxycholate,0.1% SDS in PBS, pH 7.4), including freshly added pro-tease inhibitor (Sigma, St. Lois, MO). The extracts wereplaced on ice for 10 min, vortexed briefly and centrifugedfor 15 min at 4°C to remove cell debris. The total proteinconcentration was determined using a Bradford assay(Bio-Rad Laboratories, Mississauga, ON). The sampleswere boiled for 10 min before running the gels. SDS poly-acrylamide gel electrophoresis (SDS-PAGE) was per-formed using a 10% separating gel. Twenty-five to 30 μgof protein was then electrotransferred to a nitrocellulosemembrane (Bio-Rad Laboratories). The membrane wasblocked with 5% skim milk in PBS. PBX2 and actin weredetected by primary antibodies in 5% skim milk/TBS forsecondary antibodies conjugated with horse radish perox-idase and visualized by ECL (Pierce, Rockford, IL).ImmunofluorescenceFor PBX1 and MEIS1/2, cells were grown on glass cover-slips, fixed in freshly prepared 4% paraformaldehyde inPBS for 10 min at 4°C, and permeabilized with 0.1% Tri-ton X-100 in PBS for 10 min at 25°C. For PBX2, cells onthe coverslips were fixed in cold methanol and postfixedin cold methanol/acetone (1:1). The cells were incubatedwith primary antibodies, anti-PBX1 (1:20), anti-PBX2(1:50) and anti-MEIS1/2 antibody (1:50). The binding ofprimary antibodies was followed by goat anti-mouse anti-body (Alexa Fluor 594, Molecular Probes, Eugene, Ore-gon) or rabbit anti-goat antibody (Alexa Fluor 488,Molecular Probes). Control experiments were done in theabsence of primary antibodies and verified their no or lit-tle background. For Hoechst staining, after incubation ofthe secondary antibodies, cells were incubated for 1 minwith 5 μg/ml Hoechst 33258.ImmunohistochemistryIshikawa, Hela and K562 cell pellets were solidified bycollagen gel, fixed in 4% paraformaldehyde, and embed-ded in paraffin. Sections of these cells and of the ovarieswere deparaffinized, rehydrated, and submitted to antigenretrieval by a steamer for 30 min in 10 mM citric buffer(PBX2) or Target Retrieval Solution (DAKO, Mississauga,ON; PBX1 and MEIS1/2). Endogenous peroxide wasdiminished with 3% H2O2 for 30 min. Nonspecific bind-ing was blocked with 3% BSA in PBS (PBX2) or ProteinBlock (DAKO; PBX1 and MEIS1/2) for 30 min, followedby incubation with primary antibodies overnight at 4°C.The dilution of the primary antibodies was the same asthat described for the immunofluorescence. Sampleswithout primary antibodies were used as a negative con-trol. There was no or little background without primaryantibodies. The sections were then incubated with Bioti-nylated link universal (DAKO) for 15 min and streptavi-din (DAKO) for 25 min at 25°C. Slides were developed indiaminobenzine (DAKO) or in NovaRED substrate (Vec-tor Laboratories, Inc., Burlington, ON) and counter-stained with hematoxylin.Electrophoretic Mobility-Shift Assay (EMSA)Nuclear extracts were prepared with a nuclear extract kit(Panomics, Inc., Redwood City, CA) according to themanufacturer's instructions. DNA-binding reactions wereperformed with equal amounts of nuclear proteins (10μg) and 32P-labeled probes at 25°C for 10 min. The fol-lowing synthetic double-strand oligonucleotide probeswere used: Pbx, 5'-CGAATTGATTGATGCACTAATTGGAG-3', and EMX2, 5'-AGGAAGCTGTTTATGTGATCCCCG-3',Page 3 of 9(page number not for citation purposes)1 hr at 25°C. Subsequently, the signals were detected with which have been previously shown to contain the consen-sus recognition sequences for the HOX-Pbx complexesReproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49[23,24]. For cold competition assays, a 100-fold excess ofunradiolabeled oligonucleotides was added to the reac-tions prior to the 32P-labeled probes. Anti-HOXA7 andanti-PBX2 antibodies used in the supershift experimentswere added to the nuclear extract at 25°C for 30 minbefore the addition of labeled probe. Protein-DNA com-plexes were resolved in 5% polyacrylamide gel containing1× TBE (Tris-borate-EDTA: 0.09 M Tris-borate and 2 mMEDTA, pH 8.0). Before loading of the samples, the gel waspre-run for 90 min at 100 V at 4°C. Electrophoresis wascarried out at 120 V at 4°C. The gel was then dried undervacuum and exposed to Kodak X-OMAT AR film (Kodak,Rochester, NY).ResultsExpression of HOX cofactors in immortalized granulosa cellsRecent studies have identified a number of HOX genes inthe oocytes, and we have previously shown the distribu-tion of HOXA7 in ovarian follicular development [3]. Toexamine the expression of HOX cofactors in normalhuman ovary, we focused on PBX1, PBX2 and MEIS1/2. Inaddition, to determine the expression of HOX cofactors inhuman granulosa cells, we performed RT-PCR followingthe culture of SVOG cells. Figure 1A showed that PBX1,PBX2 and MEIS1 are transcribed in proliferating SVOGcells. Two transcripts corresponding to PBX1a and PBX1bwere detected, and PBX1b, which lacks the C-terminaldomain in PBX1a, was the more predominant form inSVOG cells (Figure 1A). The specificities of the HOXcofactors antibodies were confirmed by using paraffin-embedded Ishikawa, Hela and K562 cells (data notshown). By immunofluorescence, PBX1 was observed tobe expressed both in the nuclei and in the cytoplasm,which is consistent with the results reported in Hela orIshikawa cells (Figure 1B) [25]. In SVOG cells, PBX2 andMEIS1/2 were expressed in the nuclei.ImmunohistochemistryThe distribution of PBX1, PBX2 and MEIS1/2 was exam-ined by immunohistochemistry on paraffin sections ofpremenopausal ovaries. Five ovaries were used for PBX1and 6 ovaries were used for MEIS1/2 and PBX2. Immuno-histochemical specificity of the anti-MEIS1/2 antibodywas confirmed by using MEIS1/2 blocking peptide (datanot shown). As shown in Figure 2, oocyte nuclei were neg-ative, and oocyte cytoplasms were weakly positive, at allstages for PBX1, PBX2 and MEIS1/2. PBX1 staining wasobserved in the nuclei of granulosa cells from primordialto secondary follicles. In 50% of primordial follicles, gran-ulosa cells were partially positive (Figure 2A; Table 1). Incontrast, in primary and secondary follicles, almost allgranulosa cells were completely positive (Figure 2B, C;Expression of HOX cofactors in immortalized human granu-losa (SVOG) cells in ultureFigure 1Expression of HOX cofactors in immortalized human granulosa (SVOG) cells in culture. (a) mRNA levels of PBX1, PBX2 and MEIS1. control, no cDNA. (b). Levels of PBX1, PBX2 and MEIS1/2 as shown by immunofluorescence microscopy. The staining of nuclei with Hoechst is shown in the right column. (scale bar = 20 μm).Page 4 of 9(page number not for citation purposes)Table 1). In multilayered and Graffian follicles, similar toHOXA7, PBX1 had mainly translocated to the cytoplasmReproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49(Figure 2J). The theca interna were positive, and the thecaexterna were weakly positive (Figure 2J). In the theca, theexpression was also in the cytoplasm. MEIS1/2 expressionwas quite similar to PBX1 expression (Figure 2G, H, I, L).In primordial follicles, the percentage of MEIS1/2 nega-tive granulosa cells was more compared to PBX1 (Table1). Other differences between PBX1 and MEIS1/2 stain-ings were that the theca interna were positive but showedboth nuclear and cytoplasmic staining for MEIS1/2 (Fig-ure 2L).PBX2 expression in normal human ovaries was com-pletely different from PBX1 and MEIS1/2 expression.Interestingly, the expression pattern was similar toHOXA7 expression as we previously observed [3]. Granu-losa cells in primordial follicles were negative or weaklypositive (Figure 2D; Table 1). From the primary to Graf-fian follicles, the intensity of PBX2 expression in granu-losa cells became higher when compared to theprimordial follicles and the expression pattern was mixed,i.e., negative and positive (Figure 2E, F, K; Table 1). Thecells of the theca interna were positive, and the cells of thetheca externa were negative, which was again similar toHOXA7 expression. PBX2 expression in the theca internawas in the nuclei.Effects of FSH and GDF-9 on PBX2 expressionSince PBX2 expression in human normal ovaries wasobserved to be similar to HOXA7 expression, we focusedon the regulation of PBX2 expression. In our previousstudy, HOXA7 is regulated by GDF-9 [3]. In the presentstudy, to investigate whether PBX2 expression is regulatedin SVOG cells, the cells were treated with FSH (100 ng/ml)and/or GDF-9 (50 or 100 ng/ml). Our results showed thatPBX2 expression was down-regulated by treatment withregulated the expression of PBX2 when compared to thecontrol (Figure 3).HOXA7 and PBX2 bind to the Pbx consensus sequence asa heterodimer The HOX-PBX heterodimeric complex canbind to the Pbx and EMX2 consensus sequences[23,24,26]. Therefore, we investigated the possibility thatHOXA7 and PBX2 may cooperatively bind to these twoDNA sequences by EMSA. PBX and EMX2 binding activi-ties were observed in SVOG cells (Lane 3, Figure 4A, B).The specificity of binding to the respective probes wasdetermined by using a 100-fold molar excess of unlabeledoligonucleotide as a competitor, which indicated that thebound complexes resulted from sequence specific DNA-protein interactions (Lane 1, Figure 4A, B).To examine whether HOXA7 and PBX2 were present inthe complexes, supershift analysis was performed usingantibodies directed against the two proteins. Strikingly,the addition of anti-HOXA7 and/or anti-PBX2 antibodiescaused supershifts of the complexes bound to the Pbxsequence, resulting in two bands with apparently highermolecular weights (Lanes 4, 5 and 6, Figure 4A). In con-trast, whereas addition of the anti-HOXA7 antibody pro-duced a characteristic shift in the complexes formed withthe EMX2 probe, the PBX2 antibody alone did not resultin a marked supershift (Lanes 4, 5, Figure 4B). These datarevealed a differential recruitment of PBX2 to the Pbx andEMX2 consensus sequences. HOXA7 may bind to Pbx as aheterodimer with PBX2 and to EMX2 with other cofac-tor(s).DiscussionIn this study, we demonstrated for the first time theexpression of HOX cofactors in normal human ovariesTable 1: Expression of HOX cofactors in granulosa cells during follicular developmentnegative positive Strongly positive totalPBX1primordial 1 (6%) 9 (50%) 8 (44%) 18primary 1 (2.5%) 1 (2.5%) 41 (95%) 43secondary 0 (0%) 0 (0%) 2 (100%) 2multi, Graffian 4 (100%) 0 (0%) 0 (0%) 4PBX2primordial 4 (21%) 15 (79%) 0 (0%) 19primary 3 (10%) 20 (67%) 7 (23%) 30secondary, multi, Graffian 0 (0%) 5 (83%) 1 (17%) 6MEIS1/2primordial 8 (36%) 12 (55%) 2 (9%) 22primary 6 (18%) 27 (79%) 1 (3%) 34secondary 0 (0%) 0 (0%) 4 (100%) 4multi, Graffian 5 (100%) 0 (0%) 0 (0%) 5Page 5 of 9(page number not for citation purposes)FSH, but not changed by GDF-9 alone (Figure 3). Interest-ingly, concomitant treatment with FSH and GDF-9 up-and, particularly, in human granulosa cells. During cellculture, PBX1 was observed in the nuclei and the cyto-Reproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49Page 6 of 9(page number not for citation purposes)Immunohistochemical staining for PBX1, PBX2 and MEIS1/2 in human ovariesFigure 2Immunohistochemical staining for PBX1, PBX2 and MEIS1/2 in human ovaries. (A, B, C, J) PBX1, (D, E, F, K) PBX2 and (G, H, I, L) MEIS1/2. (A, D, G): primordial follicles. (B, E, H): primary follicles. (C, F, I): secondary follicles. (J, K, L): preovu-latory follicles. Figure 2E contains both primordial and primary follicles, and Figure 2I contains both primordial and secondary follicles (scale bars: A,B,D,E,G,H = 9 μm, C,F,I = 40 μm, J,K,L = 350 μm).Reproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49plasm, whereas PBX2 and MEIS1/2 were observed only inthe nuclei in proliferating SVOG cells. In normal humanovaries in situ, PBX1 and MEIS1/2 were expressed in thenuclei in granulosa cells from primordial to secondary fol-licles. In Graffian follicles, the expression of both PBX1and MEIS1/2 was translocated from the nuclei to the cyto-plasm of the granulosa cells. There is a correlationbetween PBX1 expression and the cell cycle [25]. PBX1expression in the nucleus from the primordial to the sec-ondary follicles might show that granulosa cells are at theG2/M phase, while the cytoplasmic PBX1 expression inpreovulatory follicles might indicate that these granulosacells are not dividing. Compared to ovaries in mice, PBX1expression in human ovaries was more evident, and PBX2expression was similar [5]. The discrepancies might bedue to species differences. MEIS1 and MEIS2 control PBX1nuclear localization where the PBX1 gene is functional[27], whereas recent studies have shown that PBX1nuclear localization does not require both MEIS proteins[25]. In our in vitro experiments, MEIS1/2 was consist-ently observed in the nuclei of SVOG cells, while PBX1was expressed both in the nuclei and in the cytoplasm. Incontrast, during ovarian follicular development, PBX1and MEIS1/2 were expressed in the same pattern in situ,cellular localization differences in PBX1 and MEIS1/2 invitro and in vivo is not known.Similar to HOXA7, PBX2 expression was negative in pri-mordial follicles but positive from the stage of primaryfollicles. Furthermore, the mixed negative and positiveexpression pattern was also similar to HOXA7. One differ-ence is in preovulatory follicles; PBX2 was expressed in thenuclei, whereas HOXA7 is expressed mainly in the cyto-plasm [3]. Thus far, the role of different sublocalization ofHOXA7 is not known. These results support the notionthat HOXA7 and PBX2 might act together in granulosacells, especially when they are co-expressed in the nuclei.By FSH and/or GDF-9 treatment, PBX2 expression in gran-ulosa cells might be reversely related to steroidogenicactivities. The immortalized granulosa (SVOG) cells usedin this study are steroidogenic, as demonstrated byincreased progesterone production following treatmentwith cAMP or pregnenolone in our earlier report [22]. It iswell known that FSH and cAMP up-regulates StAR expres-sion and stimulates steroidogenesis in human granulosacells [28]. In the present study, treatment of SVOG cellswith FSH alone decreased PBX2 expression. GDF-9 on itsown did not change StAR [28], and, in our study, GDF-9had no or weak inhibitory effects on progesterone produc-tion in SVOG cells (data not shown), and it did notchange PBX2 expression. FSH or cAMP-stimulated proges-terone production, which is mediated through StAR, waspotently inhibited by GDF-9 co-application in humangranulosa cells [28]. In the present study, PBX2 expressionwas increased by the co-application of FSH and GDF-9.Hence, PBX2 expression might be inversely related to ster-oidogenesis in human granulosa cells. This notion is inaccordance with our demonstration that PBX2 wasexpressed in the granulosa and theca interna cells thatproduce estrogen and androgen, respectively, but not intheca externa that lack steroidogenic activity.To validate the PBX2 and HOXA7 protein-protein interac-tion in SVOG cells, we examined their binding to two tar-get sequences: PBX and EMX2. Previous studies showedthat complexes of Pbx1 and HOX1–4 display optimalbinding to the target sequence 5'-CGAATTGATTGAT-GCACTAATTGGAG-3' [23] and Pbx2 is also known tobind to this sequence [26]. TGAT is a Pbx binding site andTNAT is a HOX site. The TAAT, the HOX binding sitewhich was used in this study, is accepted to bind the mid-dle paralog groups 3–8 [29]. In addition, PBX2 andHOXA10 interactions with EMX2 were demonstrated inendometrial cancer cell lines [24]. EMX2 is expressed inthe epithelial components of the urogenital system duringdevelopment, and, as shown by its knockout studies, thisEffects of FSH and GDF-9 on PBX2 in SVOG cellsFigure 3Effects of FSH and GDF-9 on PBX2 in SVOG cells. The cells were treated with FSH (100 ng/ml) or/and GDF-9 (50 or 100 mg/ml). For co-treatment, cells were pre-treated with FSH for 4 hr and then treated with GDF-9 for 24 hr.Page 7 of 9(page number not for citation purposes)which indicates that MEIS1/2 might control PBX1 nuclearlocalization in vivo. The explanation underlying the sub-gene is essential for the development of the female repro-ductive system [30]. In this study, HOXA7 and PBX2 com-Reproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49plexes bound to the Pbx sequence, but not to the EMX2sequence. The results indicate that a HOXA7 and PBX2interaction occurs in granulosa cells. EMX2 can be a targetof HOXA7, but it did not bind to PBX2 in granulosa cells.to the EMX2 promoter in granulosa cells, different cofac-tors might be used to enhance the HOXA7 binding specif-icity and strength.HOXA7 and PBX2 bind as a heterodimer to the Pbx sequence, but not to EMX2Figure 4HOXA7 and PBX2 bind as a heterodimer to the Pbx sequence, but not to EMX2. EMSA analysis was performed with 32P-labeled oligonucleotides containing the (A) Pbx consensus sequence and (B) EMX2 consensus sequence. Supershift assays were also performed with antibodies to HOXA7 and PBX2. Lanes are designated as follows: lane 1, negative control using excess unlabeled cold probe (100×) with nuclear extract; lane 2, 32P-labeled probe without nuclear extract, showing the migration of free probe in absence of nuclear extract; lane 3, 32P-labeled probe with nuclear extract; lane 4, supershift reaction with 32P-labeled probe, nuclear extract and anti-PBX2 antibody; lane 5, supershift reaction with 32P-labeled probe, nuclear extract and anti-HOXA7 antibody; lane 6, supershift reaction with 32P-labeled probe, nuclear extract, anti-Pbx2 and anti-HOXA7 antibodies; lane 7, 32P-labeled probe without nuclear extract but with anti-PBX2 and anti-HOXA7 antibodies. Abbre-viation: n.s., non-specific complex.+++----++-+----++++-++++++++------++++----++-+----++++-++++++++------+Unlabeled probe32P-labeled probeNuclear extractAnti-Pbx2 AbAnti-HOXA7 AbUnlabeled probe32P-labeled probeNuclear extractAnti-Pbx2 AbAnti-HOXA7 AbA. Pbx probe B. EMX2 probe1   2   3   4   5   6   7 Free probe Free proben.s.n.s.complexesn.s.n.s.complexessupershiftssupershift1   2   3   4   5   6   7 Page 8 of 9(page number not for citation purposes)These results suggest that HOXA7 and PBX2 can makedimers in granulosa cells. However, when HOXA7 bindsReproductive Biology and Endocrinology 2008, 6:49 http://www.rbej.com/content/6/1/49ConclusionIn summary, this study demonstrated, for the first time,the expression and regulation of HOX cofactors in normalhuman ovaries. The results indicate that PBX1 and MEIS1/2 might act together and that PBX2 and HOXA7 acttogether at specific promoter regions in human granulosacells. It appears that PBX2 expression is closely related tosteroidogenic activities in granulosa cells and that it playsa role in ovarian follicular development.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsTO and HA designed the study and performed the molec-ular genetic studies, the immuno-technical studies andparticipated in discussion of the results and drafted themanuscript. SP performed the EMSA and QH performedimmunoassays and both of them participated in discus-sion of the results and critical revision of the manuscript.TM, NA and PCKL were responsible for supervision of thiswork. PCKL was responsible for the conception, design,discussion of the results, drafting and critical revision ofthe manuscript. All authors read and approved the finalmanuscript.AcknowledgementsWe thank the B.C. Foundation for Non-Animal Research and Ovarian Can-cer Canada for graduate scholarships to T.O. We thank Dr. C. Blake Gilks (UBC) for the ovarian tissues. We also thank Ms. Song Ling Poon and Dr. Christian Klausen (UBC) for their invaluable assistance. This work was sup-ported by grants from the Canadian Institutes of Health Research to PCKL (MOP-82739).References1. Barnett KR, Schilling C, Greenfeld CR, Tomic D, Flaws JA: Ovarianfollicle development and transgenic mouse models.  HumReprod Update 2006, 12:537-555.2. Pangas SA, Rajkovic A: Transcriptional regulation of early oog-enesis: in search of masters.  Hum Reprod Update 2006, 12:65-76.3. Ota T, Choi KB, Gilks CB, Leung PC, Auersperg N: Cell type- andstage-specific changes in HOXA7 protein expression inhuman ovarian folliculogenesis: possible role of GDF-9.  Dif-ferentiation 2006, 74:1-10.4. Huntriss J, Hinkins M, Picton HM: cDNA cloning and expressionof the human NOBOX gene in oocytes and ovarian follicles.Mol Hum Reprod 2006, 12:283-289.5. Villaescusa JC, Verrotti AC, Ferretti E, Farookhi R, Blasi F: Expres-sion of Hox cofactor genes during mouse ovarian folliculardevelopment and oocyte maturation.  Gene 2004, 330:1-7.6. Laurent A, Bihan R, Omilli F, Deschamps S, Pellerin I: PBX proteins:much more than Hox cofactors.  Int J Dev Biol 2008, 52:9-20.7. Moens CB, Selleri L: Hox cofactors in vertebrate development.Dev Biol 2006, 291:193-206.8. Williams TM, Williams ME, Innis JW: Range of HOX/TALE super-class associations and protein domain requirements forHOXA13:MEIS interaction.  Dev Biol 2005, 277:457-471.9. Kamps MP, Murre C, Sun XH, Baltimore D: A new homeoboxgene contributes the DNA binding domain of the t(1;19)translocation protein in pre-B ALL.  Cell 1990, 60:547-555.10. Monica K, Galili N, Nourse J, Saltman D, Cleary ML: PBX2 and11. Wagner K, Mincheva A, Korn B, Lichter P, Pöpperl H: Pbx4, a newPbx family member on mouse chromosome 8, is expressedduring spermatogenesis.  Mech Dev 2001, 103:127-131.12. Schnabel CA, Selleri L, Cleary ML: Pbx1 is essential for adrenaldevelopment and urogenital differentiation.  Genesis 2003,37:123-130.13. Schnabel CA, Selleri L, Jacobs Y, Warnke R, Cleary ML: Expressionof Pbx1b during mammalian organogenesis.  Mech Dev 2001,100:131-135.14. Selleri L, Depew MJ, Jacobs Y, Chanda SK, Tsang KY, Cheah KS,Rubenstein JL, O'Gorman S, Cleary ML: Requirement for Pbx1 inskeletal patterning and programming chondrocyte prolifer-ation and differentiation.  Development 2001, 128:3543-3557.15. Di Giacomo G, Koss M, Capellini TD, Brendolan A, Pöpperl H, SelleriL: Spatio-temporal expression of Pbx3 during mouse organo-genesis.  Gene Expr Patterns 2006, 6:747-757.16. Rhee JW, Arata A, Selleri L, Jacobs Y, Arata S, Onimaru H, Cleary ML:Pbx3 deficiency results in central hypoventilation.  Am J Pathol2004, 165:1343-1350.17. Selleri L, DiMartino J, van Deursen J, Brendolan A, Sanyal M, Boon E,Capellini T, Smith KS, Rhee J, Pöpperl H, Grosveld G, Cleary ML: TheTALE homeodomain protein Pbx2 is not essential for devel-opment and long-term survival.  Mol Cell Biol 2004,24:5324-5331.18. Hisa T, Spence SE, Rachel RA, Fujita M, Nakamura T, Ward JM,Devor-Henneman DE, Saiki Y, Kutsuna H, Tessarollo L, Jenkins NA,Copeland NG: Hematopoietic, angiogenic and eye defects inMeis1 mutant animals.  EMBO J 2004, 23:450-459.19. Azcoitia V, Aracil M, Martínez-A C, Torres M: The homeodomainprotein Meis1 is essential for definitive hematopoiesis andvascular patterning in the mouse embryo.  Dev Biol 2005,280:307-320.20. Oulad-Abdelghani M, Chazaud C, Bouillet P, Sapin V, Chambon P,Dollé P: Meis2, a novel mouse Pbx-related homeobox geneinduced by retinoic acid during differentiation of P19 embry-onal carcinoma cells.  Dev Dyn 1997, 210:173-183.21. Crijns AP, de Graeff P, Geerts D, Ten Hoor KA, Hollema H, Sluis Tvan der, Hofstra RM, de Bock GH, de Jong S, Zee AG van der, deVries EG: MEIS and PBX homeobox proteins in ovarian can-cer.  Eur J Cancer 2007, 43:2495-2505.22. Lie BL, Leung E, Leung PC, Auersperg N: Long-term growth andsteroidogenic potential of human granulosa-lutein cellsimmortalized with SV40 large T antigen.  Mol Cell Endocrinol1996, 120:169-176.23. Charboneau A, East L, Mulholland N, Rohde M, Boudreau N: Pbx1 isrequired for Hox D3-mediated angiogenesis.  Angiogenesis2005, 8:289-296.24. Sarno JL, Kliman HJ, Taylor HS: HOXA10, Pbx2, and Meis1 pro-tein expression in the human endometrium: formation ofmultimeric complexes on HOXA10 target genes.  J Clin Endo-crinol Metab 2005, 90:522-558.25. Dintilhac A, Bihan R, Guerrier D, Deschamps S, Bougerie H, WatrinT, Bonnec G, Pellerin I: Pbx1 intracellular localization is inde-pendent of Meis1 in epithelial cells of the developing femalegenital tract.  Int J Dev Biol 2005, 49:851-858.26. Neuteboom ST, Murre C: Pbx raises the DNA binding specifi-city but not the selectivity of antennapedia Hox proteins.Mol Cell Biol 1997, 17:4696-4706.27. Saleh M, Huang H, Green NC, Featherstone MS: A conformationalchange in PBX1A is necessary for its nuclear localization.Exp Cell Res 2000, 260:105-115.28. Yamamoto N, Christenson LK, McAllister JM, Strauss JF 3rd: Growthdifferentiation factor-9 inhibits 3'5'-adenosine monophos-phate-stimulated steroidogenesis in human granulosa andtheca cells.  J Clin Endocrinol Metab 2002, 87:2849-2856.29. Shen WF, Rozenfeld S, Lawrence HJ, Largman C: The Abd-B-likeHox homeodomain proteins can be subdivided by the abilityto form complexes with Pbx1a on a novel DNA target.  J BiolChem 1997, 272:8198-8206.30. Miyamoto N, Yoshida M, Kuratani S, Matsuo I, Aizawa S: Defects ofurogenital development in mice lacking Emx2.  Development1997, 124:1653-1664.Page 9 of 9(page number not for citation purposes)PBX3, new homeobox genes with extensive homology to thehuman proto-oncogene PBX1.  Mol Cell Biol 1991, 11:6149-6157.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items