To elucidate the role of phosphorylation of p53 we used the baculovirus expression system to obtain high yields of protein eventually in distinct phosphorylation states. Initially, we obtained only marginal phosphorylation, despite high levels of expression. Two-dimensional phosphopeptide maps exhibited the same pattern as known from rat cells although some sites were underrepresented. Coexpression of simian virus 40 (SV40) large T antigen or cyclin-dependent kinases, cdc2 or cdk2, had only marginal effects on the phosphorylation state of p53. However, when we employed the phosphatase inhibitor okadaic acid, overall phosphorylation of p53 was drastically enhanced in a dose-dependent manner and resembled that of p53 from SV40-transformed rat cells. This hyperphosphorylation resulted in enhanced binding of a consensus oligonucleotide as revealed by electrophoretic mobility shift assays. To assess the role of individual phosphorylation sites, we generated a set of mutants at putative or identified sites. All mutants retained the ability to bind wild-type conformation-specific antibody Pab1620, to complex with SV40 large T antigen, and to bind to the consensus oligonucleotide. Moreover, most mutants exhibited enhanced DNA binding upon okadaic acid treatment, except for a mutant at the cdk site which failed to do so. These data show that: (a) insect cells contain all the protein kinases necessary for phosphorylation of a mammalian protein, p53; (b) in insect cells the ratio of kinase/phosphatase activities differs from that in mammalian cells so that underphosphorylation of recombinant proteins in this system may result from high phosphatase activities rather than saturation of kinases with recombinant substrate; (c) the system can be manipulated to obtain subpopulations of recombinant protein in a desired phosphorylation state, and (d) phosphorylation may regulate the DNA-binding activity of p53.