We demonstrated that local expression of a dominant-negative type II TGF-beta receptor prevents live fibrogenesis and dysfunction in dimethylnitrosamine-treated rats. Using the same model, we have now tested whether a soluble TGF-beta receptor expressed in skeletal muscle can effectively suppress TGF-beta signaling in a remote organ (the liver). We constructed an adenovirus expressing an entire ectodomain of human TGF-beta type II receptor fused to the Fc portion of human IgG (AdTbeta-ExR). This soluble receptor secreted from AdTbeta-ExR-infected cells bound TGF-beta and blocked TGF-beta-signaling in vitro. After intramuscular injection of AdTbeta-ExR in rats, the soluble receptor protein was detectable in the blood for at least 3 weeks. When such rats were treated with dimethylnitrosamine, liver fibrosis was markedly attenuated without apparent systemic or local side effects. The hepatic hydroxyproline content was reduced to a level indistinguishable from that achieved by local expression of the dominant-negative TGF-beta receptor. Since a qualitatively and quantitatively similar suppression was achieved by the two methods, it may be concluded that the new strategy can achieve a complete inhibition of TGF-beta signaling under pathophysiological conditions in vivo. This strategy should facilitate clarification of the role of TGF-beta in vivo in various organs where direct gene transfer seems to be difficult.