Mutation of three cationic surface residues of human cyclophilin A (hCyPA), R69, K125, and R148, to both anionic and neutral residues left its intrinsic peptidyl-prolyl isomerase (PPIase) activity and cyclosporin A (CsA) binding unaffected, but altered its ability to inhibit the serine phosphatase activity of calcineurin (CN). R69E was 13-fold less effective (Ki = 3400 nM) than wild-type hCyPA (Ki = 270 nM) in presenting CsA for calcineurin phosphatase inhibition, while R148E was 17-fold more effective (Ki < or = 16 nM), and human CyPB was 13-fold better (Ki < or = 21 nM), establishing that a composite drug/protein surface is being recognized. The phosphoserine phosphatase reaction catalyzed by CN using unlabeled phosphoserine RII19 peptide was coupled to a continuous spectrophotometric assay to measure inorganic phosphate production using the enzyme purine ribonucleoside phosphorylase and the substrate N7-methyl-2-thioguanosine [Webb, M. R. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 4884-4887]. With this assay, we have determined that human cyclophilin A complexed with the immunosuppressive drug cyclosporin A is a noncompetitive inhibitor of calcineurin phosphatase activity. This mutational analysis identified hCyPA residues that interact with CN, and comparison to similar data on FKBP allowed us to begin to map out the CN recognition surface. The p-nitrophenylphosphatase activity of CN was stimulated ca. 3-fold by CyP.CsA, presumably reflecting altered active site geometry and selective access of this small substrate.