We originally identified preadipocyte factor-1 (Pref-1) as an inhibitor of adipogenesis by the fact that constitutive expression of full-length Pref-1A inhibits differentiation of 3T3-L1 cells into adipocytes. Subsequently, we found that the membrane form of Pref-1 is proteolytically processed at two sites in the extracellular domain, resulting in the larger (50kDa) and smaller (25kDa) soluble forms. A specific form(s) of Pref-1, which is active in inhibiting adipocyte differentiation, has not been elucidated. Here, various artificial constructs and alternative-splicing variants of Pref-1 were stably transfected into 3T3-L1 cells, or conditioned media from COS cells transfected with the various forms were added into differentiating 3T3-L1 cells. Judging by Oil Red O staining for lipid accumulation and expression of adipocyte markers, we determined that, unlike the full-length Pref-1A and the constructed large soluble form, the artificial membrane form of Pref-1 lacking the processing site proximal to the membrane was not effective in inhibiting adipogenesis. Furthermore, conditioned media from COS cells transfected with the construct containing only the first three epidermal growth factor repeats, corresponding to the small soluble form, was not effective in inhibiting adipocyte differentiation. Of the four alternative-splicing products, Pref-1A and Pref-1B, which generate both large and small soluble forms, inhibited adipogenesis, whereas Pref-1C and Pref-1D, which lack the processing site proximal to the membrane and therefore generate only the smaller soluble form, did not show any effect. We conclude that only the large soluble form, and not the transmembrane or the small soluble form, of Pref-1 is biologically active and that alternative splicing therefore determines Pref-1 function in adipocyte differentiation.

This content is only available as a PDF.
You do not currently have access to this content.