Here, we report the isolation of a human multipotent adipose-derived stem (hMADS) cell population from adipose tissue of young donors. hMADS cells display normal karyotype; have active telomerase; proliferate >200 population doublings; and differentiate into adipocytes, osteoblasts, and myoblasts. Flow cytometry analysis indicates that hMADS cells are CD44+, CD49b+, CD105+, CD90+, CD13+, Stro-1−, CD34−, CD15−, CD117−, Flk-1−, gly-A−, CD133−, HLA-DR−, and HLA-Ilow. Transplantation of hMADS cells into the mdx mouse, an animal model of Duchenne muscular dystrophy, results in substantial expression of human dystrophin in the injected tibialis anterior and the adjacent gastrocnemius muscle. Long-term engraftment of hMADS cells takes place in nonimmunocompromised animals. Based on the small amounts of an easily available tissue source, their strong capacity for expansion ex vivo, their multipotent differentiation, and their immune-privileged behavior, our results suggest that hMADS cells will be an important tool for muscle cell–mediated therapy.

The differentiation of multipotent cells into undesirable lineages is a significant risk factor when performing cell therapy. In muscular diseases, myofiber loss can be associated with progressive fat accumulation that is one of the primary factors leading to decline of muscular strength. Therefore, to avoid any contribution of injected multipotent cells to fat deposition, we have searched for a highly myogenic but nonadipogenic muscle-derived cell population. We show that the myogenic marker CD56, which is the gold standard for myoblast-based therapy, was unable to separate muscle cells into myogenic and adipogenic fractions. Conversely, using the stem cell marker CD34, we were able to sort two distinct populations, CD34(+) and CD34(-), which have been thoroughly characterized in vitro and in vivo using an immunodeficient Rag2(-/-)gamma(c) (-/-) mouse model of muscle regeneration with or without adipose deposition. Our results demonstrate that both populations have equivalent capacities for in vitro amplification. The CD34(+) cells and CD34(-) cells exhibit equivalent myogenic potential, but only the CD34(-) population fails to differentiate into adipocytes in vitro and in vivo after transplantation into regenerative fat muscle. These data indicate that the muscle-derived cells constitute a heterogeneous population of cells with various differentiation potentials. The simple CD34 sorting allows isolation of myogenic cells with no adipogenic potential and therefore could be of high interest for cell therapy when fat is accumulated in diseased muscle.

Our results support an association between congenital cryptorchidism and fetal exposure to PCBs and possibly DDE. Higher concentrations in milk could be a marker of higher exposure or for an impaired detoxification pattern in genetically predisposed individuals.