THE STEM CELL NICHE

Adiponectinemia Controls Pro-Angiogenic Cell Therapy

PHILIPPE EREN, STEPHANE CAMUS, GIANFRANCO MATRONE, TENI G. EBRAHIMIAN, DELPHINE FRANCOIS, ALAIN TEDGUI, BASTIEN SILVESTRE, OLIVIER P. BLANC-BRUDE JEAN SE

` Institut National de la Sante et de la Recherche Medicale U.689, Cardiovascular Research Center Lariboisiere, ` Hopital Lariboisiere, et Universite Paris 7, Paris, France
Key Words. Adipose tissue Progenitor cells Angiogenesis Endothelium Apoptosis

ABSTRACT
Angiogenic cell therapy with the transplantation of endothelial progenitor cells (EPC) or bone marrow mononuclear cells (BM-MNC) receives considerable attention as an approach to revascularize ischemic tissues. Adiponectin is a circulating hormone produced by the apM1 gene in adipocytes. Adiponectin modulates lipid metabolism and obesity, and it was recently found to promote physiological angiogenesis in response to ischemia. Patients with multiple cardiovascular disease risk factors or myocardial infarction may benet from progenitor cell therapy, but they display depressed adiponectinemia. We hypothesized that adiponectin stimulation of transplanted cells is critical for their pro-angiogenic function. We aimed to establish whether adiponectinemia in the cell donor or in the cell recipient determines the success of pro-angiogenic cell therapy. In vitro, we found that conditioned media derived from wild-type adipocytes (adipo-CM) or puried adiponectin strongly enhanced BM-MNC survival and proliferation and stimulated EPC differentiation, whereas adipoCM from apM12/2 adipocytes was one-half less effective. On the other hand, wild-type and apM12/2 BM-MNC displayed similar resistance to apoptosis and proliferation rates. In vivo, wild-type, and apM12/2 BM-MNC induced similar angiogenic reactions in wild-type ischemic hindlimbs. In contrast, wild-type BM-MNC had much diminished effects in apM12/2 ischemic hindlimbs. We concluded that adiponectin enhances BM-MNC survival and proliferation, and adiponectinemia in the cell therapy recipient is essential for the pro-angiogenic benets of cell therapy. These observations imply that progenitor cell transplantation might only induce angiogenesis in patients with high adiponectinemia. STEM CELLS 2009;27:27122721

Disclosure of potential conicts of interest is found at the end of this article. Adipocyte-secreted cytokines (adipokines) that circulate in the blood, such as adiponectin and leptin, have recently been found to regulate some aspects of EPC biology. Adiponectin, also known as 30 kDa adipocyte-derived complement-related protein (Acrp30), is the apM1 (adipose tissue most abundant protein-1) gene product. Adiponectin is best known for its role in fat tissue metabolism and insulin resistance [9]. Hypoadiponectinemia is thought to mediate the development of the metabolic syndrome, obesity, insulin resistance, and diabetes, and may contribute to cardiovascular disorders [8, 10]. In vitro, adiponectin promotes endothelial survival [11, 12], and Walsh and colleagues have shown that adiponectin expression in mice with ischemic hearts or limbs promotes physiological angiogenesis [13]. Conversely, apM1/ mice display increased tissue damage after severe hindlimb ischemia [14]. More recently, adiponectin was shown to enhance hematopoetic stem cell numbers in the bone marrow [15], as well as

INTRODUCTION
New hopes for the treatment of tissue ischemia and therapeutic neovascularization stemmed from autologous cell therapy with bone marrow mononuclear cells (BM-MNC) and endothelial progenitor cells (EPC) in particular. Pro-angiogenic EPC have been isolated from several tissues including adult bone marrow, peripheral blood and white adipose tissue, and umbilical cord blood [1, 2]. These EPC were successfully used to limit ischemic tissue damage and revascularize ischemic limbs and cardiac muscle in experimental animals [3, 4]. However, clinical trials of bone marrow cell transplantation only displayed very limited success in patients with acute myocardial infarction [5 7], raising questions regarding the mechanisms that control the efcacy of cell therapy in humans [8].

Author contributions: All of the authors participated in the study. In particular: O.B.B.: conception and design, nancial support, collection and/or assembly of data, data analysis and interpretation, manuscript writing, nal approval of manuscript; S.C. and P.E.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing; S.C. and P.E.: contributed equally to this work; J.S.S.: conception and design, nancial support, collection and/or assembly of data, data analysis and interpretation, nal approval of manuscript; A.T.: nancial support, administrative support, nal approval of manuscript; G.M., D.F., and T.G.E: collection and/or assembly of data. Sources of support to acknowledge: ANR P005432, INSERM, CNRS, MENRT, and the European Vascular Genomics Network (LSHMCT-2003-503254). ` Correspondence: Olivier P. Blanc-Brude, Ph.D., Cardiovascular Research Center Lariboisiere, Institut National de la Sante et de la ` Recherche Medicale U.689, Hopital Lariboisiere, 41 Bd de la chapelle, F-75010 Paris, France. Telephone: 33-1-5321-6698; Fax: 33-1-4281-3128; e-mail: [email protected] Received March 10, 2009; accepted for publication August 28, 2009; rst published online in STEM CELLS EXPRESS Month 00, 2009; available online without subscription thorugh the open access option. C V AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1002/stem.219

STEM CELLS 2009;27:27122721 www.StemCells.com

Eren, Camus, Matrone et al.

2713

EPC mobilization from bone marrow towards circulating blood upon ischemic stress [16], suggesting a link between adipokines and EPC biology. Moreover, leptin, another adipokine that regulates energy intake and expenditure including appetite and metabolism [17, 18], has also been suggested to enhance EPC vascular tube formation in vitro [19] and recruitment to vascular lesions in mice [20]. These studies established adipokines as a molecular bridge between cardiovascular risk factors, adipose tissue and EPC biology. Cardiovascular disease risk factors [21], such as dyslipidemia [22, 23], obesity [24], diabetes [2527], arterial hypertension [28], and smoking [2932] are associated with depressed adiponectinemia, On the other hand, clinical studies demonstrated that coronary arterial disease patients display one-half less circulating EPC than healthy volunteers [33]. Circulating EPC numbers inversely correlate with the Framingham score for cardiovascular risk factors, as well as diabetes, obesity, arterial hypertension, and smoking [8]. Furthermore, circulating EPC derived from coronary arterial disease [33], diabetes [34, 35], or smoking patients [36] show reduced migration, proliferation, and differentiation capacity in vitro, as well as impaired angiogenic and vasculogenic properties upon transplantation [37, 38]. Since CVD risk factors, atherosclerosis, and myocardial infarction are all associated with both defective EPC function and hypo-adiponectinemia [3942], the assumption emerged that low adiponectin may explain the defective pro-angiogenic functions in EPC [8]. However, the impact of adiponectinemia has never been assessed in the context of cell transplantation. For instance, one may ask whether cell therapy is enhanced or compromised when EPC are isolated from a donor with high or low adiponectinemia, for delivery into a recipient with low adiponectinemia. To begin to answer these questions, we investigated wild-type and apM1/ bone marrow cell function in vitro and in proangiogenic cell therapy in the murine model of ischemic hindlimb. Unexpectedly, we found that adiponectinemia in the progenitor cell recipient with an ischemic hindlimb, rather than adiponectinemia in the progenitor cell donor, impacts pro-angiogenic cell therapy.

forced the differentiation of conuent pre-adipocytes in DMEM F-12, 10% NCS, 33 nM dexamethasone, 5 lg/ml insulin, 1 lg/ml transferrin, 2 ng/ml T3/tri-iodothyronine for 10 days, producing 75% Oil Red-O-positive adipocytes with large lipid-laden cytoplasmic vesicles (Fig. 2A). For bone marrow mononuclear cells (BM-MNC), mouse femurs and tibiae were dissected and their extremities were cut. The bone marrow was ushed out, dissociated, and centrifuged onto histopaque gradients (600 g, 25 minutes). The disc of mononuclear cells was collected and washed in RPMI-1640. Fresh BM-MNC were stained with primary rabbit antibodies to mouseadiponectin receptors adipoR1, adipoR2 (Interchim, France, http://www.interchim.com), and T-cadherin (Tebu-Bio, France, http://www.tebu-bio.com), followed by APC-conjugated donkey anti-rabbit secondary antibody (Jackson Labs, France) and uorescence-assisted cell sorting (FACS) analysis. Fresh BM-MNC typically showed about 5% c-Kit and Sca-1 double positive cells by FACS (not shown). BM-MNC were differentiated into EPC onto gelatin-vitronectin coated plastic (1.25 106 cells/500 ll/ well in 12 well plates) in EBM, a minimal medium favorable to endothelial proliferation with lower growth factor concentrations [3, 4]. Optimal endothelial differentiation was assessed in growth factor-rich EBM2, by staining for both Dil-AcLDL uptake and BS-1 lectin binding. Data shown is mean percentage of double positive cells in over four independent experiments. Light microscopy observation of cell spreading, tight cobblestone patterns, and contact inhibition of proliferation helped evidence endothelial characteristics. Immunocytochemical staining was performed with anti-mouse CD45 (clone 30-F11) and anti-mouse CD11b (clone M1/70) antibodies (BD Biosciences, Germany, http://www. bdbiosciences.com) to evidence leukocyte rather than EPC differentiation.

Conditioned Media
Adipo-CM was produced by adipocytes cultured in fresh DMEMF12 with 10% NCS for 7 days. Cell fragments were eliminated by centrifugation (800 g, 15 minutes). Mouse adiponectin and leptin (Euromedex, France, http://www.euromedex.com), tumornecrosis factor-a (TNFa), inammatory interleukin-6 (IL6) (BD Biosciences, France), and VEGF165 (R&D Systems, France) concentrations were determined by ELISA according to the manufacturers instructions.

MATERIALS
Materials

AND

METHODS

Evaluation of Apoptosis and Proliferation

To determine cell cycle progression, cultured BM-MNC were collected prior to full conuency by gentle ushing followed by short trypsinization. To evaluate DNA content, BM-MNC were xed overnight in ice-cold 70% ethanolH2O and incubated in 0.05% Triton X100, 100 lg/ml of RNase A, and 50 lg/ml of propidium iodide for 1 hour in the dark before uorescence-assisted cell sorting (FACS). DNA content proles were divided as subG1 (apoptosis), G0/G1 (resting), and G2/M (proliferation). Data shown is the mean percentage of cells in each phase, in 36 independent experiments. Cultured BM-MNC were also pretreated with conditioned media or recombinant adipokines for 24 hours to 5 days and challenged with staurosporine (100 nM), etoposide (200 lM), or C2ceramide (100 lM) for 8 hours. Nuclei of 4% PFA-xed cells were stained with DAPI for examination of nuclear morphology by epiuorescence microscopy (Zeiss, France). The percentage of apoptotic nuclei showing condensed chromatin and fragmented nuclear bodies (karyorrhexis, pyknosis) was determined and averaged in ve random elds per treatment in over three independent experiments.

Recombinant mouse leptin, staurosporine, etoposide, C2-ceramide, bovine skin gelatin, type I rat-tail collagen, vitronectin, BS-1 lectin, histopaque-1083, DAPI, RNase A, and propidium iodide were from Sigma-Aldrich (St. Louis, http://www.sigmaaldrich.com), Dil-conjugated acetylated low-density lipoprotein (Dil-AcLDL) was from Harbor Bioproducts (Tebu, France, http://www.tebu-bio.com), recombinant mouse adiponectin (Acrp30) was from BioVision (via Clinisciences, France, http://www.clinisciences.com), and human vascular endothelial growth factor165 (VEGF) from R&D Systems (Minneapolis, http://www.rndsystems.com). Endothelial cell basal medium (EBM; 1 ng/ml FGF2) and EBM2 media (10 ng/ml FGF2, 0.5 ng/ml VEGF, and 20 ng/ml IGF1) were from Promocell (http:// www. promocell.com). DMEM-F12, M199, and RPMI-1640 were from Gibco-Life Sciences (Grand Island, NY, http://www.invitrogen. com), and neonatal calf serum (NCS) was from Dominique Dutscher (France, http://www.dutscher.fr).

Mouse Cell Culture

Pre-adipocytes were isolated from fat deposits dissected from 6 week old male mice, as described [43]. Inguinal white adipose tissues were digested with collagenase. Dissociated pre-adipocytes were passed through a nylon mesh and separated from adipocytes by centrifugation, before culture in DMEM-F12, 10% NCS. We

In Vivo Evaluation of Post-Ischemic Revascularization

We used our model of mouse hindlimb ischemia after ligation of the right femoral artery [44] with groups of 1012 male C57bl6J mice, of 12 weeks of age. C57bl6J wild-type mice were from

www.StemCells.com

2714

Adiponectinemia Controls Pro-Angiogenic Cell Therapy

Figure 1. Effects of adiponectin and leptin on BM-MNC proliferation and apoptosis. BM-MNC were cultured on gelatin-vitronectin in endothelial cell basal medium (EBM) only (control), vascular endothelial growth factor (VEGF)-rich EBM2 with bulletkit, or in EBM with recombinant mouse leptin (100 ng/ml; hatched bars) or adiponectin (10 lg/ml; full bars) for up to 5 days. (A) Representative phase contrast images of BMMNC cultures (magnication 200; insets: 400 magnication). (B) Quantication of cell numbers per high power eld sem. Inset shows dose-response to adiponectin in % increase after 5 days, from 0.01 to 10 lg/ml. (C) DNA content proles of BM-MNC stained with propidium iodide and analyzed by uorescence-assisted cell sorting after 3 days. Sub-G1 (apoptosis) or G2/M (proliferation) phases of the cell cycle are shown under bars on each side of G0/G1 peaks. Data expressed in percent cells sem; representative DNA content prole insets. (D) After 5 days in culture, apoptosis was induced with staurosporine or C2-ceramide for 8 hours. DAPI-stained apoptotic nuclei were counted, and data are shown as percentages of apoptotic cells sem. * indicates p < .05 versus control. (E) Representative uorescence images of DAPI-stained BM-MNC cultured in EBM (control) or adiponectin and challenged with staurosporine. Arrows point toward pyknotic or fragmented apoptotic nuclei. Abbreviations: BM-MNC, bone marrow mononuclear cells.

Charles River (http://www.criver.com). Adiponectin-decient mice (apM1/) [45] matched in sex, age, body weight, and genetic background were kindly provided by Dr. Tohru Funahashi from the Osaka University School of Medicine, Japan. Adiponectin deciency was conrmed by ELISA (not shown). Glycemia was evaluated prior to surgery, after 10 hours of fasting. No difference was observed between wild-type and apM1/ mice (144 and 162 mg/dL; p > .5). In some experiments, 100 ll of control medium or adipo-CM was administered to wild-type mice by intramuscular injection in two sites of the ischemic leg (gastrocnemius and anterior tibial), 5 hours after ligature, and repeated every 4 days (not shown). In other experiments, BM-MNC were cultured for 5 days with 20%

control DMEM-F12 or 20% adipo-CM, collected, and washed with postburn serum (PBS). Five hours after ligature, 2 105 cells were injected intravenously to the mice, as cell therapy. In a third experiment, fresh BM-MNC from wild-type or apM1/ bone marrow were injected to wild-type or apM1/ recipients. After two weeks, vessel density was evaluated as described [44] by (1) high-denition barium-contrast microangiography (1 g/ml systemic) and image acquisition with a digital x-ray transducer and computerized quantication of vessel density (percentage of image pixels occupied by vessels); (2) assessment of capillary density by immunostaining with anti-bronectin antibody (rabbit polyclonal, dilution 1/50, Chemicon International (Temecula, CA, http://www.chemicon.com)) and morphometric

Eren, Camus, Matrone et al.

2715

Figure 2. Effects of adipocyte-conditioned media on BM-MNC survival and proliferation. (A) Phase contrast photomicrograph of in vitro-differentiated adipocytes with lipid vesicles, used to condition culture medium (200 magnication). (B) ELISA determination of mouse adiponectin, vascular endothelial growth factor (VEGF)165, IL-6, and TNFa content of wild-type (wt) and apM1/ wild-type adipocytes (adipo-CM). * indicate p < .05 versus wt adipo-CM; 810 culture media of each type were tested. BM-MNC were cultured in EBM (control), VEGF-rich EBM2 with bulletkit, or endothelial cell basal medium (EBM) supplemented with 150% wt adipo-CM. (C) Cell numbers per high power eld sem after 5 days of culture. (D) DNA content proles of BM-MNC treated with 50% wt adipo-CM for 3 days, stained with propidium iodide, and analyzed by uorescence-assisted cell sorting. Sub-G1 (apoptosis) or G2/M (proliferation) phases of the cell cycle are shown under bars on each side of G0/G1 peaks. Data expressed in percent cells sem. (E) Representative phase contrast images of BM-MNC after 5 days of culture (magnication 400). (F) Cell numbers per high power eld sem after 5 days of culture in 20% wt or apM1/ adipo-CM. * indicates p < .05 versus control. (#) indicates p < .05 versus wt adipo-CM. Abbreviations: BM-MNC, bone marrow mononuclear cells; TNF, tumor-necrosis factor; VEGF, vascular endothelial growth factor.

quantication (Histolab, Microvisions, France); and (3) laserDoppler imaging of cutaneous perfusion.

Adiponectin Detection by Western Blotting

Kinetics of adiponectin expression after induction of ischemia in mice was assessed at 0, 2, 7, 14, and 21 days after ligature (six mice per point). Since adiponectin circulates in multimerized forms of low (LMW; trimers, about 90 KDa), medium (MMW; hexamers, about 180 KDa), and high molecular weight (HMW; oligomers, over 300 KDa), we chose nondenaturing gradient Western blotting [46] to determine total adiponectin expression and multimerization by calculating HMW/MMW and HMW/LMW ratios on 50 lg (muscle) or 10 lg (plasma) of protein loaded per lane, using a

monoclonal rat anti-adiponectin antibody (R&D Systems; clone 251321). Data were expressed relative to control mouse expression levels and normalized for GAPDH expression in muscles (AbCys S.A.; clone VMA374). Multimerization proles of recombinant murine adiponectin were also assessed.

Statistical Analysis
Data were expressed as mean s.e.m. Statistical evaluation was conducted with unpaired Student t-test. For animal studies, oneway analysis of variance (ANOVA) was used. Post hoc Bonferronis t-test comparisons were performed to identify which group differences accounted for the overall signicance with ANOVA.

www.StemCells.com

2716

Adiponectinemia Controls Pro-Angiogenic Cell Therapy

Figure 3. Effects of adipokines on bone marrow mononuclear cells (BM-MNC) survival. BM-MNC were cultured on gelatinvitronectin for 5 days in endothelial cell basal medium (EBM) only (gray bars), or 20% wt (full bars) or apM1/ (hatched bars) wild-type adipocytes (adipo-CM). Apoptosis was induced with staurosporine or C2-ceramide for 8 hours, before xation and staining with DAPI. (A) Percentage of apoptotic nuclei. (B) Representative uorescence images of DAPI-stained BMMNC nuclei after staurosporine challenge. Arrows point towards pyknotic or fragmented apoptotic nuclei. (C) BM-MNC cultured in wt adipo-CM for 3 days and analyzed by Western blot to assess expression of anti-apoptotic proteins XIAP, survivin, and cIAP2, compared to b-actin. (D) Quantitative densitometric image analysis, expressed in relative arbitrary units, n 4. * indicates p < .05 versus control; (#) indicates p < .05 versus wt adipo-CM. Abbreviations: XIAP, X-linked inhibitor of apoptosis protein.

Experiments were repeated at least three times, and statistical signicance was achieved when p < .05.

RESULTS
Effects of Adiponectin and Leptin on BM-MNC
We chose to focus on adipocyte-secreted adiponectin and leptin to evaluate whether major adipokines modulate BM-MNC proliferation and survival. We cultured BM-MNC with puried mouse recombinant adiponectin from 0.01 to 10 lg/ml. After 5 days in culture, adiponectin signicantly increased BM-MNC numbers from 10 ng/ml, with more than a doubling at 10 lg/ml (Fig. 1A,1B; p < .05). Adiponectin also increased proliferation (G2/M fraction) by 60% after 3 days and reduced spontaneous apoptosis (Sub-G1 fraction) by 40% (Fig. 1C). There were no comparable effects of recombinant mouse leptin at 100 ng/ml (Fig. 1A-1C). We could conrm that the majority (7080%) of freshly isolated BM-MNC express the adiponectin receptors adipoR1, adipoR2, and Tcadherin on their surface (supporting information Fig. I).

Adipocyte Secretions Modulate BM-MNC Growth and Survival

In addition to stimulating proliferation of BM-MNC, recombinant mouse adiponectin conferred BM-MNC increased resistance against apoptosis triggered by staurosporine or ceramide, as measured by DAPI-stained pyknotic or fragmented nuclei quantication (Fig. 1D,1E). We next aimed to assess the relative contribution of adiponectin to the combined effects of total adipocyte secretions. We isolated pre-adipocytes from subcutaneous white fat deposits, forced adipocyte differentiation in vitro, and produced over 90% cells with extensive lipid vesicles (Fig. 2A). We collected adipocyte-conditioned media (adipo-CM) after 7 days in culture. ELISA measurements (Fig. 2B) revealed that adipo-CM contained about

50 ng/ml of adiponectin (p < .05) and 130 pg/ml of leptin (not shown). Adipo-CM generated from apM1/ adipocytes showed the expected absence of adiponectin but similar levels of angiogenic VEGF165 (about 630 ng/ml), inammatory interleukin-6 (IL6; about 4 ng/ml), and tumor-necrosis factora (TNFa; 200 pg/ml). In addition, we quantied cells with large intracellular lipid vesicles after 15 days as a marker of terminal differentiation of pre-adipocytes into adipocytes. We observed similar levels of adipocyte differentiation (about 30%) with wild-type and apM1/ pre-adipocytes under our conditions (supporting information Fig. II). We assessed BMMNC proliferation and resistance to apoptosis in culture. After 3 days, adipo-CM stimulated a 120% increase in the BMMNC G2/M fraction versus control (Fig. 2C), with a simultaneous decrease by about 75% in spontaneous apoptosis by FACS. The dramatic proliferation of BM-MNC in adipo-CM was dose-dependent (Fig. 2D), resulting in a 12-fold increase in total BM-MNC numbers versus control EBM after 5 days, and a 10-fold increase versus VEGF-rich EBM2 bulletkit (50%, v/v). Drastic phenotype changes occurred to BM-MNC cultured with adipo-CM, which formed sheets of fully spread, contiguous, cobblestone-like cells similar to endothelium within 710 days (Fig. 2E). This contrasted with the isolated round BM-MNC observed in controls, or with VEGF-rich EBM2 bulletkit (not shown). In contrast, BM-MNC cultured in apM1/ adipo-CM failed to reach full conuency and spread within 7 days (Fig. 2E). The stimulation of BMMNC proliferation by apM1/ adipo-CM was also inferior to wild-type adipo-CM by 35% (p < .05; Fig. 2F). After 5 days in culture in wild-type or apM1/ adipo-CM, we challenged BM-MNC with staurosporine or C2-ceramide and quantied DAPI-stained pyknotic and fragmented nuclei (Fig. 3A,3B). Wild-type adipo-CM rendered BM-MNC highly resistant to death, showing 80% reduction in apoptosis (Fig. 3B). However, apM1/ adipo-CM was half less effective in blocking staurosporine- or ceramide-induced apoptosis (60% reduction, p < .05). Wild-type adipo-CM triggered the specic

Eren, Camus, Matrone et al.

2717

Figure 4. Effects of adipokines on bone marrow mononuclear cells (BM-MNC) differentiation. BM-MNC were cultured on gelatin-vitronectin for up to 5 days in endothelial cell basal medium (EBM) alone (control), vascular endothelial growth factor (VEGF)-rich EBM2 with bulletkit, or EBM supplemented with recombinant mouse adiponectin or leptin, or 50% wt or apM1/ wild-type adipocytes (adipo-CM). Endothelial progenitor cells (EPC) were triple-stained with BS-1 lectin (green), dil-conjugated acetylated low-density lipoprotein (Dil-AcLDL) (red), and DAPI (blue nuclei). (A) Graph shows percentages of BS-1 lectin- and Dil-AcLDL-stained EPC in each culture, sem. * indicates p < .05 versus control. (B) Representative images of BS-1 lectin, Dil-AcLDL, DAPI stainings (merge), and phase contrast micrographs. (C) Phase contrast micrographs of BMC immunostained with antiCD45, anti-CD11b antibodies, or relevant isotype-matched IgG (control) after 5 days in culture.

expression of the endogenous caspase inhibitor X-linked inhibitor of apoptosis protein (XIAP) but not cIAP2 or survivin (Fig. 3C,3D). In contrast, apM1/ adipo-CM was one-half less cytoprotective against staurosporine- or ceramide-induced apoptosis (about 60% reduction for each, p < .05; Fig. 3A).

Adipocyte Secretions Modulate BM-MNC Differentiation

We evaluated the proportion of adherent BM-MNC double positive for Dil-AcLDL uptake and BS1-lectin binding after culture in adipo-CM or recombinant adipokines for 37 days, a combination of features coherent with EPC differentiation (Fig. 4A,4B). EPC differentiation remained null at 3 days of culture and was enhanced by about 30% after 7 days in recombinant adiponectin, or VEGF-rich EBM2 bulletkit (p < .05), whereas recombinant leptin had no signicant effect. Wild-type adipo-CM approximately doubled EPC differentiation and so did apM1/ adipo-CM (p < .05). Some BM-MNC in control medium (2030%) remained positive for the leukocyte marker CD45, but BM-MNC became entirely www.StemCells.com

negative for CD11b after culture in wild-type adipo-CM, as shown by immunocytochemistry (Fig. 4C). To further describe the pro-angiogenic potential of total combined adipokines secreted in vitro, we administered adipoCM (100 ll) to ischemic calves by intramuscular injection after femoral artery ligation. Adipo-CM stimulated neovascularization by about 30% more than PBS (p < .05), as measured by foot cutaneous perfusion and angiographic scores (Fig. 5A,5B). We also cultured BM-MNC in medium supplemented with or without adipo-CM (50%) for 3 days, washed media off, harvested and injected the cells intravenously to mice after femoral artery ligation. BM-MNC cultured with adipo-CM increased ischemic hindlimb neovascularization by 30% more than BMMNC grown in EBM only (p < .05; Fig. 5C,5D).

Adiponectin Expression During Mouse Hindlimb Ischemia

To evaluate the role of adiponectin in pro-angiogenic cell therapy, we assessed its expression and multimerization in our murine model of hindlimb ischemia. Using native Western

2718

Adiponectinemia Controls Pro-Angiogenic Cell Therapy

apM1/ mice. After femoral artery ligation, we injected PBS as control, or wild-type, or apM1/ BM-MNC to emulate pro-angiogenic cell therapy. Wild-type BM-MNC strongly enhanced neovascularization in wild-type mice, versus PBS only (50% increase), as measured by foot cutaneous perfusion (Fig. 6A), angiographic score (Fig. 6B), and capillary density (Fig. 6C). To our surprise, apM1/ and wild-type BM-MNC produced similar stimulations of neovascularization in wild-type (wt) mice. Only the induction of capillary density appeared partly compromised in apM1/ versus wt BM-MNC (p < .05; Fig. 6C). In striking contrast, wild-type BM-MNC failed to stimulate full neovascularization in apM1/ mice, with no signicant enhancement of hindlimb blood ow or densication of the vascular tree in the ischemic area.

Pro-Angiogenic Function in Adiponectin-Decient BM-MNC

To begin to explain the differences in neovascularization above, we evaluated whether BM-MNC derived from apM1/ animals differed from wild-type cells in their cell cycle status or their capacity to proliferate, resist apoptotic challenges, or differentiate into EPC in vitro. Freshly isolated wild-type and apM1/ BM-MNC were stained with Propidium Iodide (PI) and analyzed by FACS. Fresh wild-type and apM1/ BM-MNC showed similar proliferative and apoptotic indices (supporting information Fig. IVA). After 7 days of culture in VEGF-rich EBM2, wild-type and apM1/ BM-MNC showed similar rates of cell accumulation, endothelial differentiation, and equal resistance to staurosporine- or etoposide-induced apoptosis (supporting information Fig. IVB-IVD).

Figure 5. Mouse ischemic hindlimb revascularization with adipoCM. Mouse (m ! 7) hindlimb revascularization was measured 14 days after femoral artery ligation followed by administration of adipo-CM or BM-MNC. In a rst experiment (A, B), wild-type mice received intramuscular injections of 100 ll of postburn serum (control) or adipocyteconditioned medium (adipo-CM). (A) Quantication of foot cutaneous blood ow by laser-Doppler perfusion imaging. (B) Microangiography of hindlimb vasculature. In a second experiment (C, D), wild-type BMMNC were cultured in vitro in the presence of control medium (control) or in 10% adipocyte-conditioned medium (adipo-CM), for 5 days. BMMNC were then administered to wild-type mice with ischemic limbs. (C) Quantication of foot cutaneous blood ow via laser-Doppler perfusion imaging. (D) Microangiographic scoring of hindlimb vasculatures. Values shown are mean values sem. * denotes p < .05 versus control. Abbreviations: BM-MNC, bone marrow mononuclear cells; adipo-CM, wild-type adipocytes.

DISCUSSION
We analyzed the impact of adipocyte secretions, and adiponectin in particular, on pro-angiogenic cell therapy with bone marrow-derived progenitor cells.

Blot techniques, we compared multimerization (distribution into LMW, MMW, and HMW species) of recombinant murine adiponectin to that of adipo-CM or endogenous adiponectin contained in plasma and muscles from wild-type and apM1/ mice. Recombinant mouse adiponectin (supporting information Fig. IIIA), mouse adipo-CM (supporting information Fig. IIIB), and plasma (supporting information Fig. IIIC) all displayed predominantly HMW and MMW species. Adiponectin could be fully reduced to monomers under denaturing conditions (supporting information Fig. IIIA). In ischemic calf muscles, we found a transient increase in the proportion of HMW versus MMW or LMW species between days 4 and 10 (25% and 50% at day 10 after ischemia, respectively; p < .05; supporting information Fig. IIID, IIIE), revealing increased adiponectin multimerization during calf muscle neovascularization. Multimerization returned to basal levels after 21 days, when neovascularization was complete. In contrast, total adiponectin (denatured monomers) increased in a nonsignicant manner (p > .10) at day 10 of the ischemic period (supporting information Fig. IIIF, IIIG).

Adiponectin Enhances BM-MNC Proliferation, Survival, and Therapeutic Activity

In vitro, we used recombinant adipokines and produced adipoCM as a tool to study the combined effects of adipocytederived cytokines. BM-MNC cultured in adipo-CM reached conuency rapidly and formed a homogenous sheet of endothelium-like cells, which never occurred in control conditions, even in the presence of a growth factor cocktail including VEGF. Adipo-CM was also a strikingly potent anti-apoptotic stimulus for BM-MNC in resting conditions and against multiple apoptotic challenges, associated with a three-fold increase in anti-apoptotic XIAP expression. Since other survival proteins such as cIAP2 and survivin were not enhanced by adipo-CM, increased XIAP may mediate the enhanced resistance to apoptosis. Moreover, direct intramuscular injection of adipo-CM enhanced ischemic limb neovascularization, and the culture of BM-MNC in adipo-CM signicantly enhanced their pro-angiogenic efcacy in cell therapy (Fig. 5). These data suggest that the adipocyte secretome as a whole is highly favorable to proangiogenic cell therapy with BM-MNC. We detected signicant concentrations of adiponectin in our adipo-CM (60 ng/ml), a concentration reached or largely exceeded in mouse (about 7.5 lg/ml [13]) and human (120 lg/ml [47, 48]) plasma, with an inverse correlation to body weight. Recombinant adiponectin (10 lg/ml) alone was also able to stimulate both BM-MNC proliferation and survival. However, the proliferative effects of adiponectin remained

Ischemic Neovascularization and Cell Therapy in Adiponectin-Decient Mice

To determine the role of adiponectin in pro-angiogenic cell therapy, we induced hindlimb ischemia in wild-type or

Eren, Camus, Matrone et al.

2719

Figure 6. Transgenic apM1/ mouse ischemic hindlimb revascularization. Wildtype (wt) or apM1/ mice (m ! 7) underwent femoral artery ligation and were administered PBS as control or fresh BMMNC isolated from wt or apM1/ mice. Mouse hindlimb revascularization was measured by three independent methods, 14 days after ligation and cell therapy. (A) Quantication of foot cutaneous blood ow via laser-Doppler perfusion imaging, with representative color-coded images (normal perfusion red, reduced green, and low blue). (B) Microangiography of hindlimb vasculature, with representative bariumcontrast images. (C) Numeration of anti-bronectin-labeled capillaries; representative images of ischemic calf sections by epiuorescence microscopy. Values shown are mean values sem; * denotes p < .05 versus wt mice with PBS. () denotes p < .05 between mice injected with wt or apM1/ BM-MNC. (#) denotes p < .05 between wt and apM1/ mice injected with wt BM-MNC. Abbreviations: BMMNC, bone marrow mononuclear cells; PBS, postburn serum.

inferior to those of whole adipo-CM, suggesting that other adipokines also contribute to this increase. In addition, adipoCM and adiponectin increased endothelial differentiation more modestly over 7 days but did not modify inammatory differentiation. The raise in EPC numbers occurred after BMMNC adhesion was complete, consitent with increased BMMNC differentiation rather than selective apoptosis or late adhesion of EPC. Immunodetection by FACS demonstrated cell surface expression of the adiponectin receptors AdipoR1 and AdipoR2 in over 70% of our BM-MNC (supporting information Fig. I), conrming a previous report of strong AdipoR1 expression in bone marrow [49]. Interestingly, about 80% cells also proved positive for T-Cadherin, a pro-angiogenic HMW-adiponectin receptor [9] that mediates adiponectin association with angiogenic endothelium in tumors [50]. In conclusion, adiponectin was highly favorable to BM-MNC survival and pro-angiogenic activity at near-physiological concentrations and may act through its three known receptors.

Effects of Adiponectin Relative to Other Adipokines

We produced adipo-CM from wild-type and apM1/ adipocytes to help dene the effects of adiponectin relative to and combined with other adipose tissue secretions. ELISA quantication conrmed total absence of adiponectin in apM1/ adipo-CM. Since adiponectin is known to repress IL6 and TNFa expression [51], assessing the secretion of these cytokines as well as VEGF165 in our system appeared important. www.StemCells.com

Surprisingly, none of these molecules were affected by adiponectin deciency, and neither was adipocyte differentiation in our conditions. Proliferation and the cytoprotective effects triggered by apM1/ adipo-CM was all reduced by about 50% compared to wild-type adipo-CM. This assigned a predominant role for adiponectin in the stimulation of BM-MNC survival and growth, among other adipokines. In contrast, apM1/ adipo-CM was about as efcient as wild-type adipo-CM in stimulating endothelial differentiation, suggesting that other adipokines may be responsible for this effect. It is likely that a host of pro-angiogenic adipokines are present in adipo-CM and trigger the responses observed with apM1/ adipo-CM. Adiponectin may operate as an enhancer, in synergy with these other adipokines, to mediate optimal stimulations. In contrast, leptin levels in adipo-CM remained below 100 pg/ml (not shown) and well under circulating levels (5200 ng/ml in mouse [52] and human plasma, correlated to body mass index [53]). Recombinant leptin also failed to modulate BM-MNC proliferation or endothelial differentiation in our model and was ruled out as a signicant player in our conditions, despite known leptin receptor expression in bone marrow progenitor cells [54].

Multimerized Adiponectin in Ischemic Muscles

Total adiponectin levels, assessed as denatured adiponectin monomers (supporting information Fig. III) remained constant in calf muscle over the course of ischemia and

2720

Adiponectinemia Controls Pro-Angiogenic Cell Therapy

revascularization. However, adiponectin multimerization was transiently and locally upregulated in ischemic muscle, doubling 48 hours after ischemia, coinciding with pro-angiogenic and inammatory cytokine expression and vascular leakage [55, 56]. Multimerization of recombinant adiponectin, or adiponectin in adipo-CM, was consistent with that observed in plasma and muscle (supporting information Fig. III). Since high (HMW) but not low molecular weight (LMW) species of adiponectin are thought to upregulate endothelial survival [12, 13] and function [57], adiponectin patterns in our model were compatible with angiogenesis.

which remained nonsignicant. It is possible that adiponectin deciency in the recipient organism results in depressed metabolic rates specically in the transplanted cells, blocking their pro-angiogenic function. Further work is required to clarify this point.

CONCLUSION
We showed that adipose-tissue secretions and adiponectin in particular play a critical role in stimulating BM-MNC survival, proliferation, and pro-angiogenic function. We also found that adiponectin deciency in the cell therapy recipient entirely abrogates the pro-angiogenic efcacy of transplanted BM-MNC, whereas transplanted apM1/ BM-MNC remain fully functional. We conclude that hypo-adiponectinemia alone does not signicantly impair the function of BM-MNC. Instead, target tissue adiponectin in the recipient animal controls the therapeutic activity of transplanted BM-MNC and prevails over the cell donor microenvironment. In light of these results, it is possible that hypo-adiponectinemia, as seen in obese, smoking, and multiple cardiovascular risk factor patients, explains in part the relative failure of pro-angiogenic cell therapy trials with BM-MNC. Correcting adiponectinemia in such patients could be benecial to cell therapy.

Adiponectin Deciency Compromises Cell Therapy

Our apM1/ mice displayed no signicant reduction in neovascularization without cell therapy. Shibata and colleagues had previously investigated the issue of physiological neovascularization and reported reduced ischemic hindlimb neovascularization in apM1/ mice in a stringent ischemia model with double ligation and the excision of a segment of femoral artery [14]. The severity of this model may explain the discrepancy with our results obtained in a single ligation. More recently, this group suggested that adiponectin may contribute to maintaining endogenous EPC numbers in the bone marrow and physiological EPC passage into the blood stream upon ischemia [16]. To go beyond this work, we show that wild-type BM-MNC administration enhanced neo-vascularization in wild-type ischemic hindlimbs, but it failed in apM1/ ischemic mice. Adiponectin is known to affect whole-body metabolic efciency through insulin-sensitizing effects in liver and skeletal muscle [9]. Adiponectin increases oxidative reactions and reduces plasma fatty acid and glucose levels [58]. Hence, adiponectin deciency may affect vascular cell and BM-MNC metabolism. The normal spontaneous neo-revascularization in apM1/ mice does support the idea that endothelial cells are targeted by adiponectin deciency in our model. Furthermore, wild-type and apM1/ BM-MNC displayed nearly similar therapeutic pro-angiogenic activity in ischemic wildtype mice (Fig. 6), as well as equivalent rates of adhesion, proliferation, and sensitivity to apoptosis in vitro (supporting information Fig. IV). We only observed a trend toward reduced pro-angiogenic effects of apM1/ BM-MNC,

ACKNOWLEDGMENTS
The authors would like to acknowledge the expert technical advice of Dr. Louis Casteilla (Toulouse, France) for adipocyte differentiation protocols and Dr Norikazu Maeda for the apM1/ mice. Sources of support include a grant from the Agence Nationale pour la Recherche to OBB (ANR-Physio2007 P005432), recurrent funding from the CNRS, the INSERM, and the European Vascular Genomics Network, an initiative of the european commissions sixth framework program (LSHM-CT2003-503254). P.E. and S.C. beneted from doctoral fellowships from the French government MENRT.
row cells in acute myocardial infarction: Results from the autologous stem cell transplantation in acute myocardial infarction (ASTAMI) randomized controlled trial. Am Heart J 2007;154(4):710.e18. Werner N, Nickenig G. Inuence of cardiovascular risk factors on endothelial progenitor cells: Limitations for therapy? Arterioscler Thromb Vasc Biol 2006;26(2):257266. Whitehead JP, Richards AA, Hickman IJ et al. AdiponectinA key adipokine in the metabolic syndrome. Diabetes Obes Metab 2006; 8(3):264280. Hopkins TA, Ouchi N, Shibata R, Walsh K. Adiponectin actions in the cardiovascular system. Cardiovasc Res 2007;74(1):1118. Lin LY, Lin CY, Su TC et al. Angiotensin II-induced apoptosis in human endothelial cells is inhibited by adiponectin through restoration of the association between endothelial nitric oxide synthase and heat shock protein 90. FEBS Lett 2004;574(13):106110. Kobayashi H, Ouchi N, Kihara S et al. Selective suppression of endothelial cell apoptosis by the high molecular weight form of adiponectin. Circ Res 2004;94(4):e2731. Ouchi N, Kobayashi H, Kihara S et al. Adiponectin stimulates angiogenesis by promoting cross-talk between AMP-activated protein kinase and Akt signaling in endothelial cells. J Biol Chem 2004;279(2): 13041309. Shibata R, Ouchi N, Kihara S et al. Adiponectin stimulates angiogenesis in response to tissue ischemia through stimulation of amp-activated protein kinase signaling. J Biol Chem 2004;279(27):2867028674. DiMascio L, Voermans C, Uqoezwa M et al. Identication of adiponectin as a novel hemopoietic stem cell growth factor. J Immunol 2007;178(6):35113520.

REFERENCES
8 1 2 3 4 5 Urbich C, Dimmeler S. Endothelial progenitor cells: Characterization and role in vascular biology. Circ Res 2004;95(4):343353. Jia L, Takahashi M, Yoshioka T, Morimoto H, Ise H, Ikeda U. Therapeutic potential of endothelial progenitor cells for cardiovascular diseases. Curr Vasc Pharmacol 2006;4(1):5965. Kalka C, Masuda H, Takahashi T et al. Transplantation of ex vivo expanded endothelial progenitor cells for therapeutic neovascularization. Proc Natl Acad Sci U S A 2000;97(7):34223427. Kawamoto A, Gwon HC, Iwaguro H et al. Therapeutic potential of ex vivo expanded endothelial progenitor cells for myocardial ischemia. Circulation 2001;103(5):634637. Britten MB, Abolmaali ND, Assmus B et al. Infarct remodeling after intracoronary progenitor cell treatment in patients with acute myocardial infarction (TOPCARE-AMI): mechanistic insights from serial contrast-enhanced magnetic resonance imaging. Circulation 2003; 108(18):22122218. Erbs S, Linke A, Schachinger V et al. Restoration of microvascular function in the infarct-related artery by intracoronary transplantation of bone marrow progenitor cells in patients with acute myocardial infarction: The Doppler substudy of the reinfusion of enriched progenitor cells and infarct remodeling in acute myocardial infarction (REPAIR-AMI) Trial. Circulation 2007;116(4):366374. Lunde K, Solheim S, Aakhus S et al. Exercise capacity and quality of life after intracoronary injection of autologous mononuclear bone mar9 10 11

12 13

14 15

Eren, Camus, Matrone et al.

2721

16 Shibata R, Skurk C, Ouchi N et al. Adiponectin promotes endothelial progenitor cell number and function. FEBS Lett 2008;582(11): 16071612. 17 Skilton MR, Celermajer DS. The effects of obesity-related peptides on the vasculature. Curr Vasc Pharmacol 2006;4(1):7985. 18 Coll AP, Farooqi IS, ORahilly S. The hormonal control of food intake. Cell 2007;129(2):251262. 19 Wolk R, Deb A, Caplice NM et al. Leptin receptor and functional effects of leptin in human endothelial progenitor cells. Atherosclerosis 2005;183(1):131139. 20 Schroeter MR, Leifheit M, Sudholt P et al. Leptin enhances the recruitment of endothelial progenitor cells into neointimal lesions after vascular injury by promoting integrin-mediated adhesion. Circ Res 2008;103(5):536544. 21 Patel DA, Srinivasan SR, Xu JH et al. Adiponectin and its correlates of cardiovascular risk in young adults: The Bogalusa heart study. Metabolism 2006;55(11):15511557. 22 Matsushita K, Yatsuya H, Tamakoshi K et al. Comparison of circulating adiponectin and proinammatory markers regarding their association with metabolic syndrome in Japanese men. Arterioscler Thromb Vasc Biol 2006;26(4):871876. 23 Fernandez-Real JM, Vendrell J, Ricart W. Circulating adiponectin and plasma fatty acid prole. Clin Chem 2005;51(3):603609. 24 Ouchi N, Ohishi M, Kihara S et al. Association of hypoadiponectinemia with impaired vasoreactivity. Hypertension 2003;42(3):231234. 25 Sakuta H, Suzuki T, Yasuda H, et al. Adiponectin levels and cardiovascular risk factors in Japanese men with type 2 diabetes. Endocr J 2005;52(2):241244. 26 Qi L, Li T, Rimm E et al. The 276 polymorphism of the APM1 gene, plasma adiponectin concentration, and cardiovascular risk in diabetic men. Diabetes 2005;54(5):16071610. 27 Spranger J, Kroke A, Mohlig M et al. Adiponectin and protection against type 2 diabetes mellitus. Lancet 2003;361(9353):226228. 28 Patel JV, Lim HS, Hughes EA, et al. Adiponectin and hypertension: A putative link between adipocyte function and atherosclerotic risk? J Hum Hypertens 2007;21(1):14. 29 Takefuji S, Yatsuya H, Tamakoshi K et al. Smoking status and adiponectin in healthy Japanese men and women. Prev Med 2007;45(6): 471475. 30 Imatoh T, Miyazaki M, Kadowaki K et al. Interaction of low serum adiponectin levels and smoking on coronary stenosis in Japanese men. Int J Cardiol 2006;110(2):251255. 31 Iwashima Y, Katsuya T, Ishikawa K et al. Association of hypoadiponectinemia with smoking habit in men. Hypertension 2005;45(6): 10941100. 32 Thamer C, Stefan N, Stumvoll M et al. Reduced adiponectin serum levels in smokers. Atherosclerosis 2005;179(2):421422. 33 Vasa M, Fichtlscherer S, Adler K et al. Increase in circulating endothelial progenitor cells by statin therapy in patients with stable coronary artery disease. Circulation 2001;103(24):28852890. 34 Tepper OM, Galiano RD, Capla JM et al. Human endothelial progenitor cells from type II diabetics exhibit impaired proliferation, adhesion, and incorporation into vascular structures. Circulation 2002; 106(22):27812786. 35 Loomans CJ, de Koning EJ, Staal FJ et al. Endothelial progenitor cell dysfunction: a novel concept in the pathogenesis of vascular complications of type 1 diabetes. Diabetes 2004;53(1):195199. 36 Kondo T, Hayashi M, Takeshita K et al. Smoking cessation rapidly increases circulating progenitor cells in peripheral blood in chronic smokers. Arterioscler Thromb Vasc Biol 2004;24(8):14421447. 37 Miranville A, Heeschen C, Sengenes C et al. Improvement of postnatal neovascularization by human adipose tissue-derived stem cells. Circulation 2004;110(3):349355. 38 Tamarat R, Silvestre JS, Le Ricousse-Roussanne S et al. Impairment in ischemia-induced neovascularization in diabetes: Bone marrow

39

40 41 42 43 44 45 46

47 48 49 50 51 52 53 54 55

56

57 58

mononuclear cell dysfunction and therapeutic potential of placenta growth factor treatment. Am J Pathol 2004;164(2):457466. Lim S, Koo BK, Cho SW et al. Association of adiponectin and resistin with cardiovascular events in Korean patients with type 2 diabetes: The Korean atherosclerosis study (KAS) A 42-month prospective study. Atherosclerosis 2008;196(1):398404. Nilsson PM, Engstrom G, Hedblad B et al. Plasma adiponectin levels in relation to carotid intima media thickness and markers of insulin resistance. Arterioscler Thromb Vasc Biol 2006;26(12):27582762. Kumada M, Kihara S, Sumitsuji S et al. Association of hypoadiponectinemia with coronary artery disease in men. Arterioscler Thromb Vasc Biol 2003;23(1):8589. Pischon T, Girman CJ, Hotamisligil GS et al. Plasma adiponectin levels and risk of myocardial infarction in men. Jama 2004;291(14): 17301737. Bjorntorp P, Karlsson M, Pertoft H et al. Isolation and characterization of cells from rat adipose tissue developing into adipocytes. J Lipid Res 1978;19(3):316324. You D, Waeckel L, Ebrahimian TG et al. Increase in vascular permeability and vasodilation are critical for proangiogenic effects of stem cell therapy. Circulation 2006;114(4):328338. Maeda N, Shimomura I, Kishida K et al. Diet-induced insulin resistance in mice lacking adiponectin/ACRP30. Nat Med 2002;8(7): 731737. Waki H, Yamauchi T, Kamon J et al. Impaired multimerization of human adiponectin mutants associated with diabetes. Molecular structure and multimer formation of adiponectin. J Biol Chem 2003; 278(41):4035240363. Staiger H, Tschritter O, Machann J et al. Relationship of serum adiponectin and leptin concentrations with body fat distribution in humans. Obes Res 2003;11(3):368372. Vendrell J, Broch M, Vilarrasa N et al. Resistin, adiponectin, ghrelin, leptin, and proinammatory cytokines: relationships in obesity. Obes Res 2004;12(6):962971. Su AI, Wiltshire T, Batalov S et al. A gene atlas of the mouse and human protein-encoding transcriptomes. Proc Natl Acad Sci U S A 2004;101(16):60626067. Hebbard LW, Garlatti M, Young LJ et al. T-cadherin supports angiogenesis and adiponectin association with the vasculature in a mouse mammary tumor model. Cancer Res 2008;68(5):14071416. Wulster-Radcliffe MC, Ajuwon KM, Wang J et al. Adiponectin differentially regulates cytokines in porcine macrophages. Biochem Biophys Res Commun 2004;316(3):924929. Maffei M, Halaas J, Ravussin E et al. Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nat Med 1995;1(11):11551161. Considine RV, Sinha MK, Heiman ML et al. Serum immunoreactiveleptin concentrations in normal-weight and obese humans. N Engl J Med 1996;334(5):292295. Umemoto Y, Tsuji K, Yang FC et al. Leptin stimulates the proliferation of murine myelocytic and primitive hematopoietic progenitor cells. Blood 1997;90(9):34383443. Coufnhal T, Silver M, Kearney M et al. Impaired collateral vessel development associated with reduced expression of vascular endothelial growth factor in ApoE/ mice. Circulation 1999;99(24): 31883198. Lee CW, Stabile E, Kinnaird T et al. Temporal patterns of gene expression after acute hindlimb ischemia in mice: Insights into the genomic program for collateral vessel development. J Am Coll Cardiol 2004;43(3):474482. Torigoe M, Matsui H, Ogawa Y et al. Impact of the high-molecularweight form of adiponectin on endothelial function in healthy young men. Clin Endocrinol (Oxf) 2007;67(2):276281. Kadowaki T, Yamauchi T. Adiponectin and adiponectin receptors. Endocr Rev 2005;26(3):439451.

See www.StemCells.com for supporting information available online.

www.StemCells.com