Carcinogenesis vol.36 no.1 pp.32–40, 2015 doi:10.1093/carcin/bgu222 Advance Access publication October 24, 2014
Syndecan-1 regulates adipogenesis: new insights in dedifferentiated liposarcoma tumorigenesis Laure-Emmanuelle Zaragosi1,6, Bérengère Dadone2,3,4, JeanFrançois Michiels2,3, Marion Marty5, Florence Pedeutour3,4, Christian Dani1,†, Laurence Bianchini3,*,† 1
Institute of Biology Valrose, UMR7277 CNRS/UMR1091 INSERM/University of Nice-Sophia Antipolis, 06108 Nice, France, 2Department of Pathology, Nice University Hospital, 06202 Nice, France, 3Institute for Research on Cancer and Aging of Nice, CNRS UMR 7284/INSERM U1081, University of NiceSophia Antipolis, 06107 Nice, France, 4Laboratory of Solid Tumor Genetics, Nice University Hospital, 06107 Nice, France and 5Department of Pathology, Bordeaux University Hospital, 33076 Bordeaux, France 6 Present address: CNRS, Institut de Pharmacologie Moléculaire et Cellulaire, UMR7275, University of Nice- Sophia Antipolis, 06560 Sophia Antipolis, France
Syndecan-1 (SDC1/CD138) is one of the main cell surface proteoglycans and is involved in crucial biological processes. Only a few studies have analyzed the role of SDC1 in mesenchymal tumor pathogenesis. In particular, its involvement in adipose tissue tumors has never been investigated. Dedifferentiated liposarcoma, one of the most frequent types of malignant adipose tumors, has a high potential of recurrence and metastastic evolution. Classical chemotherapy is inefficient in metastatic dedifferentiated liposarcoma and novel biological markers are needed for improving its treatment. In this study, we have analyzed the expression of SDC1 in well-differentiated/dedifferentiated liposarcomas and showed that SDC1 is highly overexpressed in dedifferentiated liposarcoma compared with normal adipose tissue and lipomas. Silencing of SDC1 in liposarcoma cells impaired cell viability and proliferation. Using the human multipotent adiposederived stem cell model of human adipogenesis, we showed that SDC1 promotes proliferation of undifferentiated adipocyte progenitors and inhibits their adipogenic differentiation. Altogether, our results support the hypothesis that SDC1 might be involved in liposarcomagenesis. It might play a prominent role in the dedifferentiation process occurring when well-differentiated liposarcoma progress to dedifferentiated liposarcoma. Targeting SDC1 in these tumors might provide a novel therapeutic strategy.
Introduction Syndecan-1 (SDC1) is a cell surface heparan sulphate-bearing proteoglycan, also named CD138, that plays an important role in regulating the initiation and progression of some cancers, including breast cancer and myeloma (1–3). High serum levels of SDC1 in patients affected by myeloma correlate with poor prognosis. Studies in animal models indicate that SDC1 is a potent stimulator of myeloma tumor growth and metastasis and functional analysis confirmed the importance of SDC1 in myeloma pathobiology (4). SDC1 binds growth factors, such as fibroblast growth factors (FGFs), and can potentiate their intracellular signalling pathways, probably mediating the effects of SDC1 in myeloma. Altogether, these data suggest that SDC1 could be a marker of some cancers. Abbreviations: DDLPS, dedifferentiated liposarcomas; FACS, fluorescenceactivated cell sorting; hMADS, human multipotent adipose-derived stem; NSAT, normal subcutaneous adipose tissue; SDC1, Syndecan-1; FGFs, fibroblast growth factors; WDLPS, well-differentiated liposarcomas. †
These authors contributed equally to this work.
Materials and methods WDLPS and DDLPS specimens Eighteen WDLPS and 17 DDLPS were retrieved from the database of the Laboratory of Solid Tumor Genetics (Nice, France). In all cases, the diagnosis of WDLPS or DDLPS was established according to the World Health Organization Classification of Tumors (5). Control samples Normal subcutaneous adipose tissue (NSAT) samples from five non-obese patients who underwent surgery for non-malignant disease were used as controls. WDLPS cell lines 93T449, 94T778 and 95T1000 are WDLPS cell lines that have been established in our laboratory in 1993, 1994 and 1995, respectively. 93T449 and 94T778 were established from a primary WDLPS and its recurrence, respectively. 95T1000 was established from a WDLPS recurrence. These cell lines were extensively described previously by our laboratory (6,7). The three cell lines have been passaged for more than 50 times and regular testing showed no changes in typical WDLPS/DDLPS characteristics: (i) presence of surpernumerary ring or giant rod marker chromosomes (tested by karyotyping), (ii) strong amplification of MDM2 (tested by fluorescence in situ hybridization). Array-CGH analyses are also performed on a regular basis for a more accurate characterization of genomic stability of these cell lines. Cells were last tested a week before all experiments were performed. Isolation and culture of hMADS cells hMADS cells are primary cells that were obtained from the stroma of human adipose tissue in 2003 and were extensively described previously by our laboratory (8). Adipose tissue was collected, with the informed consent of the parents and approbation of the Centre Hospitalier Universitaire de Nice Review Board, as scraps from surgical specimens. The cell populations that have been studied in this work were isolated from the pubic region fat pad of a 5-year old (hMADS2) and a 4-month old (hMADS3) male donors, respectively. After 10 passages of each population, cells were tested in our laboratory by karyotyping and no chromosomal abnormalities were detected. A liquid-nitrogen frozen stock of hMADS2 and hMADS3 cells was established. Upon thawing, cells were maintained for a maximum of 15 passages before all experiments were performed. As these cells are primary cells that are never passaged more than a total of 25 times before being used in experiments, no re-authentication was performed. Moreover, these cells are cultivated in a separate biological safety cabinet to avoid any cross contamination. Proliferation medium for routine
© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
32
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
*To whom correspondence should be addressed. Laboratoire de Génétique des Tumeurs Solides, Faculté de Médecine, 28 avenue de Valombrose, 06107 Nice, Cedex 02, France. Tel: +33 4 93 37 70 12; Fax: +33 4 93 37 70 07; Email: [email protected]
Disrupting the activation or decreasing expression of SDC1 might be a valid therapeutic approach for SDC1 expressing cancers. The expression and the functional role, if any, of SDC1 in tumors of soft tissues have not been reported yet. Adipose tissue tumors are the most frequent soft tissue tumors. Three types of adipose tissue tumors are distinguished by the current international classification upon their degree of malignancy: (i) benign: mainly lipomas; (ii) intermediate: well-differentiated liposarcomas (WDLPS) also called ‘atypical lipomatous tumors’ when peripherally located; (iii) malignant: dedifferentiated liposarcomas (DDLPS), myxoid liposarcoma and pleomorphic liposarcoma. WDLPS are composed mostly of mature fat whereas DDLPS contain both a WDLPS component and a non-lipogenic sarcoma component which in most cases is high grade. More than 40% of DDLPS will relapse locally. A significant proportion of patients will remain with a non-resectable disease that will metastasize in 15–20% of DDLPS cases with a mortality of 30% at 5-year followup. Standard chemotherapy is poorly efficient and alternative options are so far limited. Identification of new therapeutic targets is therefore urgent. We have analyzed the expression and function of SDC1 in WDLPS/DDLPS primary tumors as well as in WDLPS cell lines. In addition, we have used the human multipotent adipose-derived stem (hMADS) cell model of human adipogenesis to investigate the potential role of SDC1 in adipocyte differentiation. Our results provide evidence that high levels of SDC1 may block adipogenic differentiation of liposarcoma cells and maintain the undifferentiated phenotype.
SDC1 in adipocyte progenitors and in liposarcoma
maintenance of hMADS cells is composed of DMEM (low glucose) containing 10% fetal calf serum, 10 mM HEPES, 100 U/ml penicillin and streptomycin and supplemented with FGF2 (PreproTech) as previously reported (9). After reaching 80% confluence, adherent cells were dissociated in 0.25% trypsin EDTA and seeded at 4500 cells/cm2. Cultures were maintained at 37°C in a humidified gassed incubator, 5% CO2 in air. FGFR inhibitor PD173074 was from Selleckchem. hMADS cell differentiation Adipocyte differentiation was performed as described previously (10). Basically, confluent cells were cultured in DMEM/Ham’s F12 media supplemented with transferrin (10 µg/ml), insulin (0.86 µM), triiodothyronine (0.2 nM), dexamethasone (1 µM), isobutyl-methylxanthine (100 µM) and rosiglitazone (500 nM). Three days later, the medium was changed (dexamethasone and isobutylmethylxanthine were omitted). Neutral lipid accumulation was assessed by Oil red O staining (11). Glycerol-3-phosphate dehydrogenase activity was performed in triplicate wells using the method described previously (11). Glycerol-3-phosphate dehydrogenase is an enzyme that is required for the formation of triglycerides.
siRNA and SDC1-expressing vector transfection hMADS and WDLPS cells were transfected with small interfering RNA (siRNA) duplexes using HiPerfect reagent (Qiagen) during the exponential growth cell phase. siRNAs against SDC1 were from Ambion SDC1 si1 sense: GGACUUCACCUUUGAAACCtt, SDC1 si1 antisense: GGUUUCAAAGGUGAAGUCCtt; SDC1 si2 sense: GGAGGAAUUCUAUGCCUGAtt, SDC1 si2 antisense: UGUUUCUUUCAU UGCAUUUtt and siRNA control duplex was from Eurogentec. Cells were transfected with 8 nM siRNA in medium composed of 60% Dulbecco's modified Eagle's medium low glucose, 40% MCDB-201, insulin (10 µg/ml), transferrin (5 µg/ml), selenium (50 ng/ml), dexamethasone (10–9 M), ascorbic sodium acid (50 µg/ml), 2.5 ng/ml FGF2, and supplemented with 0.5% fetal calf serum. For mSDC1 expressing vector (gift of R. Sanderson, University of Alabama at Birmingham, USA), transfection of hMADS cells was performed by nucleofection as we described previously (12) and stable clones were selected using neomycin. FACS analysis Fluorescence-activated cell sorting (FACS) analyses were performed on living hMADS cells after trypsin dissociation and a single phosphate-buffered saline wash as described previously (12). Syndecan-1 antibody was from Santa Cruz Biotechnology: Syndecan-1 (DL-101): sc-12765 and was used according to the manufacturer. Immunohistochemistry Immunohistochemical staining was performed on representative tissue sections from each case, (6 WDLPS cases and 12 DDLPS cases) using a 48-link autostainer (Dako). Sections were heated for 20 min at 97°C in buffer solution (pH 6.0) (PT-link Dako device) and incubated for 20 min with mouse monoclonal antibody to CD138/syndecan-1 (Dako clone MI 15 1:100 dilution) followed by 20-min incubation with the secondary antibody linked to a peroxidase. 3′,3′-diaminobenzidine tetrahydrochloride was used as the chromogen. Slides were counterstained with hematoxylin. The staining score was evaluated according to the intensity of the labelling (+: weak staining, ++: moderate staining and +++: strong staining) and the percentage of positive cells. Plasmocytes served as positive internal controls. When plasmocytes were absent, the technique was validated using an external control. Real time quantitative RT-PCR analysis qRT-PCR analysis was done in WDLPS and DDLPS cases for which frozen material was available (16 WDLPS cases and 7 DDLPS cases). Total RNA was extracted using either the RNeasy lipid tissue mini kit (Qiagen) for tumor samples of primary WDLPS/DDLPS, lipomas, NSAT and hMADS cells or TrizolTM (6) (Invitrogen) for WDLPS cell lines 93T449, 94T778, 95T1000 and short-term culture of normal adipose tissue. The quality of the isolated RNA was evaluated using the 2100 Bioanalyzer and the RNA 6000 Nano kit (Agilent Technologies). The RNA samples were treated by DNA-free™ (Applied Biosystems). One microgram of total RNA was reverse-transcribed into cDNA using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems). Each qRT-PCR experiment was performed three times in duplicate with the ABI PRISM 7500 Detection System and FAM dyes (Applied Biosystems) (WDLPS/DDLPS experiments) or SYBR Green (Applied Biosystems) (hMADS cells experiments) according to the manufacturer’s
Results To analyze the role of SDC1 in adipocytic tumors, we first investigated SDC1 expression in WDLPS and DDLPS. Expression of SDC1 in WDLPS/DDLPS We used qRT-PCR to analyze the expression levels of SDC1 in WDLPS, DDLPS and lipomas. When compared with NSAT, we detected an overexpression of SDC1 in all the 16 WDLPS cases (Figure 1A). This overexpression was moderate in all the WDLPS cases except one which showed a high overexpression level; mean level for WDLPS compared with NSAT was: 8.96 ± 2.8, range 3.75– 50. Strikingly, SDC1 overexpression was found to be much higher in DDLPS than in WDLPS: mean level compared with NSAT: 183.8 ± 90, range 19–730 (Figure 1A). To check whether SDC1 overexpression was characteristic of malignant adipose tumors, we analyzed SDC1 expression in a series of 38 lipomas. We did not detect any significant change of SDC1 expression in lipomas compared with NSAT, with a mean level of 2.22 ± 3.10, range 0.3–14 (Supplementary Figure 1, available at Carcinogenesis Online). Higher SDC1 expression in DDLPS compared with WDLPS was confirmed by microarray re-analysis from GEO datasets, when comparing gene expression of 47 WDLPS and 36 DDLPS cases (Supplementary Figure 2A, available at Carcinogenesis Online). Interestingly, the same type of analysis showed that DDLPS also expressed higher SDC1 levels compared with two other liposarcoma subtypes: pleomorphic and myxoid liposarcomas; with all three liposarcomas expressing higher SDC1 levels than normal fat (Supplementary Figure 2B, available at Carcinogenesis Online). To determine whether SDC1 overexpression in WDLPS/DDLPS resulted from genomic amplification or rearrangement of the SDC1 gene, we performed FISH analysis using, respectively, a BAC probe covering the SDC1 gene (RP11-202B22) and SDC1 ‘break-apart’ BAC probes (SDC1 gene 5′ region: RP11-327N17 and SDC1 gene 3′ region: RP11-152M21 on a series of WDLPS (19 cases) and DDLPS cases (8 cases) including some of the cases we analyzed by qRT-PCR. None of the cases showed any amplification or rearrangement of the SDC1 gene indicating that SDC1 overexpression was not consecutive to amplification or gross structural rearrangement of the gene (data not shown). We performed expression analysis of the SDC1 protein using immunohistochemistry in 6 WDLPS cases and 12 DDLPS cases. As illustrated in Figure 1B, all WDLPS cases were negative for SDC1 staining. In contrast, most of the DDLPS cases (10 of 13; 77%) exhibited cytoplasmic positive staining for SDC1: weak staining (1+) was observed in 46% (6 of 13) of the cases while the staining was moderate (2+) to strong (3+) in 30% (4 of 13) of the cases. Role of SDC1 in proliferation of WDLPS-derived cell lines We analyzed SDC1 expression in WDLPS-derived cell lines. SDC1 expression was higher in all WDLPS-derived cell lines than in NSAT (Figure 1C), suggesting that SDC1 expression is upregulated in adipocytic tumor cells.
33
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Cell proliferation assays Cells were plated at 2.8 × 104 cells per well for hMADS cells and 1.5 × 104 per well for WDLPS cells. After the appropriate time, cells were trypsinized and counted with a Coulter counter.
protocol. RPLP0 and TBP were used as endogenous controls for normalization. NSAT and short-term cultures of normal adipose tissue were used as the reference samples for experiments on WDLPS/DDLPS tumors and WDLPS cell lines, respectively. The reaction mix consisted of 10 µl of TaqMan master mix 2×, 1 µl of TaqMan gene expression mix and 5 µl of 1/10 cDNA in a final volume of 20 µl. The following TaqMan gene expression assays (Applied Biosystems) were used: Hs99999902_m1 (RPLP0) and Hs00896423_m1 (Syndecan-1). For SYBR Green experiments, final reaction volume was 25 µl, including specific primers (0.4 µM), 5 ng of reverse transcribed RNA and 12.5 ml SYBR green mastermix (Applied Biosystems). Primer sequences were for SDC1 PFd: 5′-TTTGCCCCCTGAAGATCAAGA-3′/ PRv: 5′-GAGCTGCGTGTCCTCCAAGT-3′ and for TBP PFd: 5′-CACGAACCACGGCACTGATT-3′/PRv: 5′-TTTTCTTGCTGCCAGTC TGGAC-3′. PCR conditions were as described previously (13). The comparative Ct (threshold cycle) method was used to achieve relative quantification of gene expression.
L.-E. Zaragosi et al.
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Fig. 1. Analysis of SDC1 expression in WDLPS and DDLPS primary tumors and cell lines. (A) Boxplot of qRT-PCR analysis measuring SDC1 expression in WDLPS (16 cases) and DDLPS (7 cases). Gene expression was normalized to levels of RPLP0 mRNA and expressed as fold changes relative to normal subcutaneous adipose tissue. (B) Analysis of SDC1 protein by immunohistochemistry in one WDLPS case and two DDLPS cases. Negative staining for SDC1 in a WDLPS case and positive staining for SDC1 in two DDLPS cases showing, respectively, weak positive staining (+) and strongly positive staining (+++). (C) QRT-PCR analysis measuring SDC1 expression in cultured cells derived from the WDLPS cell lines 93T449, 94T778 and 95T1000 and in cells from short term cultures of normal subcutaneous adipose tissue (NSAT). Gene expression was quantified by qRT-PCR, normalized to levels of TBP mRNA and expressed as fold changes relative to NSAT cells.
To study the function of SDC1 in WDLPS cells, we analyzed the effects of its knockdown by RNA interference in 94T778 and 95T1000 cells. We found that viable cell numbers were significantly reduced by transfection of a siRNA targeting SDC1 compared with a scrambled siRNA, as shown in Figure 2. Expression of SDC1 in human adipocyte progenitors To further understand the role of SDC1 in WDLPS/DDLPS tumorigenesis, we analyzed the putative involvement of SDC1 in human
34
adipogenesis using the hMADS cell model. These adipocyte progenitors have been isolated from the stromavascular fraction of infant adipose tissues (8). They exhibit the characteristics of mesenchymal stem cells, i.e. the capacity to self-renew while maintaining a normal diploid karyotype and the potential to undergo differentiation into various cell types including adipocytes (14,15). We had previously analyzed the transcriptome of hMADS cells to identify factors that control human adipogenesis and we had screened for genes that were differentially expressed in adipocytes compared with their progenitors. Analysis of
SDC1 in adipocyte progenitors and in liposarcoma
these data indicated that SDC1 indeed belonged to the subset of genes which expression was down-regulated during adipogenic differentiation (16). qRT-PCR analysis validated microarray results showing that SDC1 was down-regulated early during adipogenesis (Figure 3A), as its expression was already inhibited 24 h after treatment of hMADS cells with the adipogenic cocktail. This result was confirmed by studying expression of SDC1 protein at the cell surface by FACS analysis. As shown in Figure 3B, the number of SDC1-positive hMADS cells dramatically decreased upon exposure of the cells to the adipogenic medium. The percentage of SDC1-positive cells decreased from 47 ± 8% in undifferentiated hMADS to 18% at day 1 of differentiation; 14% at day 5 and 3.7% at day 10. We have previously reported that FGF2 plays an autocrine and critical role in proliferation and self-renewal of hMADS cells (9). Interestingly, we noticed that inhibition of SDC1 and of FGF2 gene expression were concomitant during adipocyte differentiation (Supplementary Figure 3, available at Carcinogenesis Online). This observation prompted us to investigate the contribution of FGF pathway to the control of SDC1 expression. Flow cytometry analysis revealed that SDC1 expression was upregulated in undifferentiated hMADS cells upon addition of exogenous FGF2. As shown in Figure 4A, FGF2 greatly induced expression of SDC1 gene. FACS analysis confirmed that the number of SDC1positive cells increased upon addition of FGF2 (Figure 4B). FGF2 treatment increased SDC1-positive cell fraction from 19 ± 1.8% to 52 ± 3%. Conversely, in the absence of exogenous FGF2, addition of a specific inhibitor of FGF receptor, the PD173074 compound, inhibited SDC1 expression (Figure 4B). About 19% of the cells were SDC1positive in control conditions versus 13 ± 1.4% in PD173074-treated cells. Altogether, these data strongly suggest that SDC1 expression is
under the control of the FGF pathway and that SDC1 could play a role in proliferation and differentiation of hMADS cells. SDC1 plays a critical role in proliferation and differentiation of hMADS cells We addressed the role of SDC1 in hMADS cell proliferation using gene silencing and overexpression approaches. hMADS cells were transiently transfected with SDC1-siRNA duplexes (see Supplementary Figure 4, available at Carcinogenesis Online, for the efficiency of SDC1 silencing). The consequences of SDC1 depletion on the number of undifferentiated cells were analyzed. In contrast to the treatment of hMADS cells with the scrambled siRNA, SDC1 knockdown significantly reduced hMADS cell numbers (Figure 5A), indicating that SDC1 expression plays an important role in proliferation of undifferentiated cells. Then, we investigated the effects of SDC1 stable overexpression. For that purpose, hMADS cells were nucleofected with a mouse SDC1-expressing vector (mSDC1-hMADS cells) or with the control empty vector (control hMADS cells) and we compared the number of cells in both cultures. As shown in Figure 5B, overexpression of SDC1 significantly increased the number of undifferentiated cells. This effect was still detected when cells were exposed to FGF2 or to the PD173074 compound. As shown in Figure 3, SDC1 expression is down-regulated upon induction of adipocyte differentiation. To investigate if SDC1 downregulation is necessary for commitment towards the adipocyte lineage, we compared the adipocyte differentiation of mSDC1- and control hMADS cells. Maintenance of SDC1 expression during differentiation led to inhibition of the formation of lipid-containing cells and to lowering of the triglyceride-synthesizing glycerol-3-phosphate dehydrogenase activity (Figure 5C).
35
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Fig. 2. Effect of SDC1 silencing on WDLPS cell proliferation. 95T1000 and 94T778 cells were plated (15 000 cells/well) and proliferation was assayed after transfection with siRNA targeting SDC1 (SDC1 si1 and SDC1 si2) compared with transfection of a scramble siRNA (si-scrambled). Results are the average of counting of 3 culture wells (12-well plates) after 5 days. *P
Syndecan-1 regulates adipogenesis: new insights in dedifferentiated liposarcoma tumorigenesis Laure-Emmanuelle Zaragosi1,6, Bérengère Dadone2,3,4, JeanFrançois Michiels2,3, Marion Marty5, Florence Pedeutour3,4, Christian Dani1,†, Laurence Bianchini3,*,† 1
Institute of Biology Valrose, UMR7277 CNRS/UMR1091 INSERM/University of Nice-Sophia Antipolis, 06108 Nice, France, 2Department of Pathology, Nice University Hospital, 06202 Nice, France, 3Institute for Research on Cancer and Aging of Nice, CNRS UMR 7284/INSERM U1081, University of NiceSophia Antipolis, 06107 Nice, France, 4Laboratory of Solid Tumor Genetics, Nice University Hospital, 06107 Nice, France and 5Department of Pathology, Bordeaux University Hospital, 33076 Bordeaux, France 6 Present address: CNRS, Institut de Pharmacologie Moléculaire et Cellulaire, UMR7275, University of Nice- Sophia Antipolis, 06560 Sophia Antipolis, France
Syndecan-1 (SDC1/CD138) is one of the main cell surface proteoglycans and is involved in crucial biological processes. Only a few studies have analyzed the role of SDC1 in mesenchymal tumor pathogenesis. In particular, its involvement in adipose tissue tumors has never been investigated. Dedifferentiated liposarcoma, one of the most frequent types of malignant adipose tumors, has a high potential of recurrence and metastastic evolution. Classical chemotherapy is inefficient in metastatic dedifferentiated liposarcoma and novel biological markers are needed for improving its treatment. In this study, we have analyzed the expression of SDC1 in well-differentiated/dedifferentiated liposarcomas and showed that SDC1 is highly overexpressed in dedifferentiated liposarcoma compared with normal adipose tissue and lipomas. Silencing of SDC1 in liposarcoma cells impaired cell viability and proliferation. Using the human multipotent adiposederived stem cell model of human adipogenesis, we showed that SDC1 promotes proliferation of undifferentiated adipocyte progenitors and inhibits their adipogenic differentiation. Altogether, our results support the hypothesis that SDC1 might be involved in liposarcomagenesis. It might play a prominent role in the dedifferentiation process occurring when well-differentiated liposarcoma progress to dedifferentiated liposarcoma. Targeting SDC1 in these tumors might provide a novel therapeutic strategy.
Introduction Syndecan-1 (SDC1) is a cell surface heparan sulphate-bearing proteoglycan, also named CD138, that plays an important role in regulating the initiation and progression of some cancers, including breast cancer and myeloma (1–3). High serum levels of SDC1 in patients affected by myeloma correlate with poor prognosis. Studies in animal models indicate that SDC1 is a potent stimulator of myeloma tumor growth and metastasis and functional analysis confirmed the importance of SDC1 in myeloma pathobiology (4). SDC1 binds growth factors, such as fibroblast growth factors (FGFs), and can potentiate their intracellular signalling pathways, probably mediating the effects of SDC1 in myeloma. Altogether, these data suggest that SDC1 could be a marker of some cancers. Abbreviations: DDLPS, dedifferentiated liposarcomas; FACS, fluorescenceactivated cell sorting; hMADS, human multipotent adipose-derived stem; NSAT, normal subcutaneous adipose tissue; SDC1, Syndecan-1; FGFs, fibroblast growth factors; WDLPS, well-differentiated liposarcomas. †
These authors contributed equally to this work.
Materials and methods WDLPS and DDLPS specimens Eighteen WDLPS and 17 DDLPS were retrieved from the database of the Laboratory of Solid Tumor Genetics (Nice, France). In all cases, the diagnosis of WDLPS or DDLPS was established according to the World Health Organization Classification of Tumors (5). Control samples Normal subcutaneous adipose tissue (NSAT) samples from five non-obese patients who underwent surgery for non-malignant disease were used as controls. WDLPS cell lines 93T449, 94T778 and 95T1000 are WDLPS cell lines that have been established in our laboratory in 1993, 1994 and 1995, respectively. 93T449 and 94T778 were established from a primary WDLPS and its recurrence, respectively. 95T1000 was established from a WDLPS recurrence. These cell lines were extensively described previously by our laboratory (6,7). The three cell lines have been passaged for more than 50 times and regular testing showed no changes in typical WDLPS/DDLPS characteristics: (i) presence of surpernumerary ring or giant rod marker chromosomes (tested by karyotyping), (ii) strong amplification of MDM2 (tested by fluorescence in situ hybridization). Array-CGH analyses are also performed on a regular basis for a more accurate characterization of genomic stability of these cell lines. Cells were last tested a week before all experiments were performed. Isolation and culture of hMADS cells hMADS cells are primary cells that were obtained from the stroma of human adipose tissue in 2003 and were extensively described previously by our laboratory (8). Adipose tissue was collected, with the informed consent of the parents and approbation of the Centre Hospitalier Universitaire de Nice Review Board, as scraps from surgical specimens. The cell populations that have been studied in this work were isolated from the pubic region fat pad of a 5-year old (hMADS2) and a 4-month old (hMADS3) male donors, respectively. After 10 passages of each population, cells were tested in our laboratory by karyotyping and no chromosomal abnormalities were detected. A liquid-nitrogen frozen stock of hMADS2 and hMADS3 cells was established. Upon thawing, cells were maintained for a maximum of 15 passages before all experiments were performed. As these cells are primary cells that are never passaged more than a total of 25 times before being used in experiments, no re-authentication was performed. Moreover, these cells are cultivated in a separate biological safety cabinet to avoid any cross contamination. Proliferation medium for routine
© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
32
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
*To whom correspondence should be addressed. Laboratoire de Génétique des Tumeurs Solides, Faculté de Médecine, 28 avenue de Valombrose, 06107 Nice, Cedex 02, France. Tel: +33 4 93 37 70 12; Fax: +33 4 93 37 70 07; Email: [email protected]
Disrupting the activation or decreasing expression of SDC1 might be a valid therapeutic approach for SDC1 expressing cancers. The expression and the functional role, if any, of SDC1 in tumors of soft tissues have not been reported yet. Adipose tissue tumors are the most frequent soft tissue tumors. Three types of adipose tissue tumors are distinguished by the current international classification upon their degree of malignancy: (i) benign: mainly lipomas; (ii) intermediate: well-differentiated liposarcomas (WDLPS) also called ‘atypical lipomatous tumors’ when peripherally located; (iii) malignant: dedifferentiated liposarcomas (DDLPS), myxoid liposarcoma and pleomorphic liposarcoma. WDLPS are composed mostly of mature fat whereas DDLPS contain both a WDLPS component and a non-lipogenic sarcoma component which in most cases is high grade. More than 40% of DDLPS will relapse locally. A significant proportion of patients will remain with a non-resectable disease that will metastasize in 15–20% of DDLPS cases with a mortality of 30% at 5-year followup. Standard chemotherapy is poorly efficient and alternative options are so far limited. Identification of new therapeutic targets is therefore urgent. We have analyzed the expression and function of SDC1 in WDLPS/DDLPS primary tumors as well as in WDLPS cell lines. In addition, we have used the human multipotent adipose-derived stem (hMADS) cell model of human adipogenesis to investigate the potential role of SDC1 in adipocyte differentiation. Our results provide evidence that high levels of SDC1 may block adipogenic differentiation of liposarcoma cells and maintain the undifferentiated phenotype.
SDC1 in adipocyte progenitors and in liposarcoma
maintenance of hMADS cells is composed of DMEM (low glucose) containing 10% fetal calf serum, 10 mM HEPES, 100 U/ml penicillin and streptomycin and supplemented with FGF2 (PreproTech) as previously reported (9). After reaching 80% confluence, adherent cells were dissociated in 0.25% trypsin EDTA and seeded at 4500 cells/cm2. Cultures were maintained at 37°C in a humidified gassed incubator, 5% CO2 in air. FGFR inhibitor PD173074 was from Selleckchem. hMADS cell differentiation Adipocyte differentiation was performed as described previously (10). Basically, confluent cells were cultured in DMEM/Ham’s F12 media supplemented with transferrin (10 µg/ml), insulin (0.86 µM), triiodothyronine (0.2 nM), dexamethasone (1 µM), isobutyl-methylxanthine (100 µM) and rosiglitazone (500 nM). Three days later, the medium was changed (dexamethasone and isobutylmethylxanthine were omitted). Neutral lipid accumulation was assessed by Oil red O staining (11). Glycerol-3-phosphate dehydrogenase activity was performed in triplicate wells using the method described previously (11). Glycerol-3-phosphate dehydrogenase is an enzyme that is required for the formation of triglycerides.
siRNA and SDC1-expressing vector transfection hMADS and WDLPS cells were transfected with small interfering RNA (siRNA) duplexes using HiPerfect reagent (Qiagen) during the exponential growth cell phase. siRNAs against SDC1 were from Ambion SDC1 si1 sense: GGACUUCACCUUUGAAACCtt, SDC1 si1 antisense: GGUUUCAAAGGUGAAGUCCtt; SDC1 si2 sense: GGAGGAAUUCUAUGCCUGAtt, SDC1 si2 antisense: UGUUUCUUUCAU UGCAUUUtt and siRNA control duplex was from Eurogentec. Cells were transfected with 8 nM siRNA in medium composed of 60% Dulbecco's modified Eagle's medium low glucose, 40% MCDB-201, insulin (10 µg/ml), transferrin (5 µg/ml), selenium (50 ng/ml), dexamethasone (10–9 M), ascorbic sodium acid (50 µg/ml), 2.5 ng/ml FGF2, and supplemented with 0.5% fetal calf serum. For mSDC1 expressing vector (gift of R. Sanderson, University of Alabama at Birmingham, USA), transfection of hMADS cells was performed by nucleofection as we described previously (12) and stable clones were selected using neomycin. FACS analysis Fluorescence-activated cell sorting (FACS) analyses were performed on living hMADS cells after trypsin dissociation and a single phosphate-buffered saline wash as described previously (12). Syndecan-1 antibody was from Santa Cruz Biotechnology: Syndecan-1 (DL-101): sc-12765 and was used according to the manufacturer. Immunohistochemistry Immunohistochemical staining was performed on representative tissue sections from each case, (6 WDLPS cases and 12 DDLPS cases) using a 48-link autostainer (Dako). Sections were heated for 20 min at 97°C in buffer solution (pH 6.0) (PT-link Dako device) and incubated for 20 min with mouse monoclonal antibody to CD138/syndecan-1 (Dako clone MI 15 1:100 dilution) followed by 20-min incubation with the secondary antibody linked to a peroxidase. 3′,3′-diaminobenzidine tetrahydrochloride was used as the chromogen. Slides were counterstained with hematoxylin. The staining score was evaluated according to the intensity of the labelling (+: weak staining, ++: moderate staining and +++: strong staining) and the percentage of positive cells. Plasmocytes served as positive internal controls. When plasmocytes were absent, the technique was validated using an external control. Real time quantitative RT-PCR analysis qRT-PCR analysis was done in WDLPS and DDLPS cases for which frozen material was available (16 WDLPS cases and 7 DDLPS cases). Total RNA was extracted using either the RNeasy lipid tissue mini kit (Qiagen) for tumor samples of primary WDLPS/DDLPS, lipomas, NSAT and hMADS cells or TrizolTM (6) (Invitrogen) for WDLPS cell lines 93T449, 94T778, 95T1000 and short-term culture of normal adipose tissue. The quality of the isolated RNA was evaluated using the 2100 Bioanalyzer and the RNA 6000 Nano kit (Agilent Technologies). The RNA samples were treated by DNA-free™ (Applied Biosystems). One microgram of total RNA was reverse-transcribed into cDNA using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems). Each qRT-PCR experiment was performed three times in duplicate with the ABI PRISM 7500 Detection System and FAM dyes (Applied Biosystems) (WDLPS/DDLPS experiments) or SYBR Green (Applied Biosystems) (hMADS cells experiments) according to the manufacturer’s
Results To analyze the role of SDC1 in adipocytic tumors, we first investigated SDC1 expression in WDLPS and DDLPS. Expression of SDC1 in WDLPS/DDLPS We used qRT-PCR to analyze the expression levels of SDC1 in WDLPS, DDLPS and lipomas. When compared with NSAT, we detected an overexpression of SDC1 in all the 16 WDLPS cases (Figure 1A). This overexpression was moderate in all the WDLPS cases except one which showed a high overexpression level; mean level for WDLPS compared with NSAT was: 8.96 ± 2.8, range 3.75– 50. Strikingly, SDC1 overexpression was found to be much higher in DDLPS than in WDLPS: mean level compared with NSAT: 183.8 ± 90, range 19–730 (Figure 1A). To check whether SDC1 overexpression was characteristic of malignant adipose tumors, we analyzed SDC1 expression in a series of 38 lipomas. We did not detect any significant change of SDC1 expression in lipomas compared with NSAT, with a mean level of 2.22 ± 3.10, range 0.3–14 (Supplementary Figure 1, available at Carcinogenesis Online). Higher SDC1 expression in DDLPS compared with WDLPS was confirmed by microarray re-analysis from GEO datasets, when comparing gene expression of 47 WDLPS and 36 DDLPS cases (Supplementary Figure 2A, available at Carcinogenesis Online). Interestingly, the same type of analysis showed that DDLPS also expressed higher SDC1 levels compared with two other liposarcoma subtypes: pleomorphic and myxoid liposarcomas; with all three liposarcomas expressing higher SDC1 levels than normal fat (Supplementary Figure 2B, available at Carcinogenesis Online). To determine whether SDC1 overexpression in WDLPS/DDLPS resulted from genomic amplification or rearrangement of the SDC1 gene, we performed FISH analysis using, respectively, a BAC probe covering the SDC1 gene (RP11-202B22) and SDC1 ‘break-apart’ BAC probes (SDC1 gene 5′ region: RP11-327N17 and SDC1 gene 3′ region: RP11-152M21 on a series of WDLPS (19 cases) and DDLPS cases (8 cases) including some of the cases we analyzed by qRT-PCR. None of the cases showed any amplification or rearrangement of the SDC1 gene indicating that SDC1 overexpression was not consecutive to amplification or gross structural rearrangement of the gene (data not shown). We performed expression analysis of the SDC1 protein using immunohistochemistry in 6 WDLPS cases and 12 DDLPS cases. As illustrated in Figure 1B, all WDLPS cases were negative for SDC1 staining. In contrast, most of the DDLPS cases (10 of 13; 77%) exhibited cytoplasmic positive staining for SDC1: weak staining (1+) was observed in 46% (6 of 13) of the cases while the staining was moderate (2+) to strong (3+) in 30% (4 of 13) of the cases. Role of SDC1 in proliferation of WDLPS-derived cell lines We analyzed SDC1 expression in WDLPS-derived cell lines. SDC1 expression was higher in all WDLPS-derived cell lines than in NSAT (Figure 1C), suggesting that SDC1 expression is upregulated in adipocytic tumor cells.
33
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Cell proliferation assays Cells were plated at 2.8 × 104 cells per well for hMADS cells and 1.5 × 104 per well for WDLPS cells. After the appropriate time, cells were trypsinized and counted with a Coulter counter.
protocol. RPLP0 and TBP were used as endogenous controls for normalization. NSAT and short-term cultures of normal adipose tissue were used as the reference samples for experiments on WDLPS/DDLPS tumors and WDLPS cell lines, respectively. The reaction mix consisted of 10 µl of TaqMan master mix 2×, 1 µl of TaqMan gene expression mix and 5 µl of 1/10 cDNA in a final volume of 20 µl. The following TaqMan gene expression assays (Applied Biosystems) were used: Hs99999902_m1 (RPLP0) and Hs00896423_m1 (Syndecan-1). For SYBR Green experiments, final reaction volume was 25 µl, including specific primers (0.4 µM), 5 ng of reverse transcribed RNA and 12.5 ml SYBR green mastermix (Applied Biosystems). Primer sequences were for SDC1 PFd: 5′-TTTGCCCCCTGAAGATCAAGA-3′/ PRv: 5′-GAGCTGCGTGTCCTCCAAGT-3′ and for TBP PFd: 5′-CACGAACCACGGCACTGATT-3′/PRv: 5′-TTTTCTTGCTGCCAGTC TGGAC-3′. PCR conditions were as described previously (13). The comparative Ct (threshold cycle) method was used to achieve relative quantification of gene expression.
L.-E. Zaragosi et al.
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Fig. 1. Analysis of SDC1 expression in WDLPS and DDLPS primary tumors and cell lines. (A) Boxplot of qRT-PCR analysis measuring SDC1 expression in WDLPS (16 cases) and DDLPS (7 cases). Gene expression was normalized to levels of RPLP0 mRNA and expressed as fold changes relative to normal subcutaneous adipose tissue. (B) Analysis of SDC1 protein by immunohistochemistry in one WDLPS case and two DDLPS cases. Negative staining for SDC1 in a WDLPS case and positive staining for SDC1 in two DDLPS cases showing, respectively, weak positive staining (+) and strongly positive staining (+++). (C) QRT-PCR analysis measuring SDC1 expression in cultured cells derived from the WDLPS cell lines 93T449, 94T778 and 95T1000 and in cells from short term cultures of normal subcutaneous adipose tissue (NSAT). Gene expression was quantified by qRT-PCR, normalized to levels of TBP mRNA and expressed as fold changes relative to NSAT cells.
To study the function of SDC1 in WDLPS cells, we analyzed the effects of its knockdown by RNA interference in 94T778 and 95T1000 cells. We found that viable cell numbers were significantly reduced by transfection of a siRNA targeting SDC1 compared with a scrambled siRNA, as shown in Figure 2. Expression of SDC1 in human adipocyte progenitors To further understand the role of SDC1 in WDLPS/DDLPS tumorigenesis, we analyzed the putative involvement of SDC1 in human
34
adipogenesis using the hMADS cell model. These adipocyte progenitors have been isolated from the stromavascular fraction of infant adipose tissues (8). They exhibit the characteristics of mesenchymal stem cells, i.e. the capacity to self-renew while maintaining a normal diploid karyotype and the potential to undergo differentiation into various cell types including adipocytes (14,15). We had previously analyzed the transcriptome of hMADS cells to identify factors that control human adipogenesis and we had screened for genes that were differentially expressed in adipocytes compared with their progenitors. Analysis of
SDC1 in adipocyte progenitors and in liposarcoma
these data indicated that SDC1 indeed belonged to the subset of genes which expression was down-regulated during adipogenic differentiation (16). qRT-PCR analysis validated microarray results showing that SDC1 was down-regulated early during adipogenesis (Figure 3A), as its expression was already inhibited 24 h after treatment of hMADS cells with the adipogenic cocktail. This result was confirmed by studying expression of SDC1 protein at the cell surface by FACS analysis. As shown in Figure 3B, the number of SDC1-positive hMADS cells dramatically decreased upon exposure of the cells to the adipogenic medium. The percentage of SDC1-positive cells decreased from 47 ± 8% in undifferentiated hMADS to 18% at day 1 of differentiation; 14% at day 5 and 3.7% at day 10. We have previously reported that FGF2 plays an autocrine and critical role in proliferation and self-renewal of hMADS cells (9). Interestingly, we noticed that inhibition of SDC1 and of FGF2 gene expression were concomitant during adipocyte differentiation (Supplementary Figure 3, available at Carcinogenesis Online). This observation prompted us to investigate the contribution of FGF pathway to the control of SDC1 expression. Flow cytometry analysis revealed that SDC1 expression was upregulated in undifferentiated hMADS cells upon addition of exogenous FGF2. As shown in Figure 4A, FGF2 greatly induced expression of SDC1 gene. FACS analysis confirmed that the number of SDC1positive cells increased upon addition of FGF2 (Figure 4B). FGF2 treatment increased SDC1-positive cell fraction from 19 ± 1.8% to 52 ± 3%. Conversely, in the absence of exogenous FGF2, addition of a specific inhibitor of FGF receptor, the PD173074 compound, inhibited SDC1 expression (Figure 4B). About 19% of the cells were SDC1positive in control conditions versus 13 ± 1.4% in PD173074-treated cells. Altogether, these data strongly suggest that SDC1 expression is
under the control of the FGF pathway and that SDC1 could play a role in proliferation and differentiation of hMADS cells. SDC1 plays a critical role in proliferation and differentiation of hMADS cells We addressed the role of SDC1 in hMADS cell proliferation using gene silencing and overexpression approaches. hMADS cells were transiently transfected with SDC1-siRNA duplexes (see Supplementary Figure 4, available at Carcinogenesis Online, for the efficiency of SDC1 silencing). The consequences of SDC1 depletion on the number of undifferentiated cells were analyzed. In contrast to the treatment of hMADS cells with the scrambled siRNA, SDC1 knockdown significantly reduced hMADS cell numbers (Figure 5A), indicating that SDC1 expression plays an important role in proliferation of undifferentiated cells. Then, we investigated the effects of SDC1 stable overexpression. For that purpose, hMADS cells were nucleofected with a mouse SDC1-expressing vector (mSDC1-hMADS cells) or with the control empty vector (control hMADS cells) and we compared the number of cells in both cultures. As shown in Figure 5B, overexpression of SDC1 significantly increased the number of undifferentiated cells. This effect was still detected when cells were exposed to FGF2 or to the PD173074 compound. As shown in Figure 3, SDC1 expression is down-regulated upon induction of adipocyte differentiation. To investigate if SDC1 downregulation is necessary for commitment towards the adipocyte lineage, we compared the adipocyte differentiation of mSDC1- and control hMADS cells. Maintenance of SDC1 expression during differentiation led to inhibition of the formation of lipid-containing cells and to lowering of the triglyceride-synthesizing glycerol-3-phosphate dehydrogenase activity (Figure 5C).
35
Downloaded from http://carcin.oxfordjournals.org/ at Michigan State University on February 20, 2015
Fig. 2. Effect of SDC1 silencing on WDLPS cell proliferation. 95T1000 and 94T778 cells were plated (15 000 cells/well) and proliferation was assayed after transfection with siRNA targeting SDC1 (SDC1 si1 and SDC1 si2) compared with transfection of a scramble siRNA (si-scrambled). Results are the average of counting of 3 culture wells (12-well plates) after 5 days. *P
