Background— Adipose tissue development and remodeling are closely associated with the growth of vascular network. We hypothesized that adipose tissue may contain progenitor cells with angiogenic potential and that therapy based on adipose tissue-derived progenitor cells administration may constitute a promising cell therapy in patients with ischemic disease. Methods and Results— In mice, cultured stromal-vascular fraction (SVF) cells from adipose tissue have a great proangiogenic potential, comparable to that of bone marrow mononuclear cells in the mouse ischemic hindlimb model. Similarly, cultured human SVF cells differentiate into endothelial cells, incorporate into vessels, and promote both postischemic neovascularization in nude mice and vessel-like structure formation in Matrigel plug. In vitro, these cells represent a homogeneous population of CD34- and CD13-positive cells, which can spontaneously express the endothelial cell markers CD31 and von Willebrand factor when cultured in semisolid medium. Interestingly, dedifferentiated mature human adipocytes have the potential to rapidly acquire the endothelial phenotype in vitro and to promote neovascularization in ischemic tissue and vessel-like structure formation in Matrigel plug, suggesting that cells of endothelial and adipocyte phenotypes may have a common precursor. Conclusions— This study demonstrates, for the first time, that adipocytes and endothelial cells have a common progenitor. Such adipose lineage cells participate in vascular-like structure formation in Matrigel plug and enhance the neovascularization reaction in ischemic tissue. These results also highlight the concept that adipose lineage cells represent a suitable new cell source for therapeutic angiogenesis in ischemic disease.

Summary Like mesenchymal stem cells from bone marrow (BM‐MSCs), adipose tissue‐derived adult stem cells (ADAS cells) can differentiate into several lineages and present therapeutical potential for repairing damaged tissues. The use of allogenic stem cells can enlarge their therapeutical interest, provided that the grafted cells could be tolerated. We investigate here, for the first time, the immunosuppressive properties of ADAS cells compared with the well‐characterized immunosuppressive properties of BM‐MSCs. ADAS cells did not provoke in vitro alloreactivity of incompatible lymphocytes and, moreover, suppressed mixed lymphocyte reaction (MLR) and lymphocyte proliferative response to mitogens. The impairment of inhibition when ADAS cells and BM‐MSCs were separated from lymphocytes by a permeable membrane suggests that cell contact is required for a full inhibitory effect. Hepatocyte growth factor is secreted by both stem cells but, similar to interleukin‐10 and transforming growth factor‐β (TGF‐β), the levels of which were undetectable in supernatants of MLR inhibited by ADAS cells or BM‐MSCs, it did not seem implicated in the stem cell suppressive effect. These findings support that ADAS cells share immunosuppressive properties with BM‐MSCs. Therefore, ADAS cell‐based reconstructive therapy could employ allogenic cells and because of their immunosuppressive properties, ADAS cells could be an alternative source to BM‐MSCs to treat allogenic conflicts.

According to the state of mitochondrial respiration, the respiratory chain generates superoxide anions converted into hydrogen peroxide. Two uncoupling proteins (UCP) able to modulate the coupling between the respiratory chain and ATP synthesis are now identified and could be involved in mitochondrial H2O2 generation. UCP1 is specific to brown adipose tissue (BAT) whereas UCP2 is expressed in numerous tissues, particularly in monocytes/macrophages. Preincubation of BAT mitochondrial fractions with GDP, an inhibitor of UCP1, induced a rise in mitochondrial membrane potential (assessed by rhodamine 123 uptake) and H2O2 production. An uncoupling agent reversed this effect. Liver mitochondria exhibited a similar phenotype. GDP was also able to raise membrane potential and H2O2 production of the mitochondria from nonparenchymal cells expressing UCP2, but was completely ineffective on mitochondria from hepatocytes deprived of UCP2. The GDP effect was also observed with mitochondrial fractions of the spleen or thymus, which highly expressed UCP2. Altogether, these results strongly suggest that UCP2 is sensitive to GDP and that the UCPs, particularly UCP2, are able to modulate H2O2 mitochondrial generation. This supports a role for UCP2 in cellular (patho-) physiological processes involving free radicals generated by mitochondria, such as oxidative damage, inflammation, or apoptosis.