The pleiotropic cytokine interleukin 6 (IL-6) plays a role in the pathogenesis of various diseases, such as multiple myeloma, autoimmune and inflammatory diseases and osteoporosis. Therefore, specific inhibitors of IL-6 may have clinical applications. We previously succeeded in developing receptor antagonists of IL-6 that antagonized wild-type IL-6 activity on the human Epstein-Barr virus (EBV)-transformed B cell line CESS and the human hepatoma cell line HepG2. However, these proteins still had agonistic activity on the human myeloma cell line XG-1. We here report the construction of a novel mutant protein of IL-6 in which two different mutations are combined that individually disrupt the association of the IL-6/IL-6 receptor (R) alpha complex with the signaltransducing "beta" chain, gp130, but leave the binding of IL-6 to IL-6R alpha intact. The resulting mutant protein (with substitutions of residues Gln160 to Glu, Thr163 to Pro, and replacement of human residues Lys42-Ala57 with the corresponding residues of mouse IL-6) was inactive on XG-1 cells and weakly antagonized wild-type IL-6 activity on these cells. By introducing two additional substitutions (Phe171Leu, Ser177Arg), the affinity of the mutant protein for IL-6R alpha was increased fivefold, rendering it capable of completely inhibiting wild-type IL-6 activity on XG-1 cells. Moreover, this mutant also antagonized the activity of IL-6, but not that of leukemia inhibitory factor, oncostatin M, or GM-CSF on the human erythroleukemia cell line TF-1, demonstrating its specificity for IL-6. These data demonstrate the feasibility of developing specific IL-6R antagonists. The availability of such antagonists may offer an approach to specifically inhibit IL-6 activity in vivo.