The mechanism whereby the universal apoptogen and serine/threonine phosphatase inhibitor okadaic acid (OA) kills cells, is still unclear. To create a novel tool for probing of OA action, fibroblasts were selected for OA-resistance after infection with a retroviral Jurkat T-cell cDNA expression library. Twenty-one clones were selected. Two of these (OAR1, OAR2) were studied in detail. OAR1 and 2 had each a retrovirally introduced short cDNA, corresponding to a human gene (oar1 and oar2, respectively) with unknown function. Reintroduction of oar1 or oar2 cDNA into wild-type cells reproduced the OA-resistant phenotype. OAR1 and 2 were cross-resistant to other phosphatase inhibitors (calyculin A, cantharidin), but not to staurosporine or microinjected Cytochrome c, thus, indicating a disturbance in a limited number of death pathways, upstream or independent of apaf-1/caspases-3/9. The action of OA involved caspase-dependent and caspase-independent components. Both components were less efficient in OAR1 and 2, than in wild-type cells. Subtle differences existed between OA-induced phosphoprotein patterns in wild-type cells, OAR1, and OAR2, indicating that a narrow selection of protein phosphorylation events had been targeted. We propose that the clones have defects in a hitherto non-elucidated signal pathway linking OA-induced protein phosphorylation to initiation of a death execution pathway provided with a caspase-dependent amplification loop. The novel OA-resistant cell clones will be used to elucidate the significance for apoptosis of oar1 and 2, their link to altered protein phosphorylation, and the potential link of the latter to initiation of apoptosis.