Recent studies have found a higher frequency of the PTEN tumor-suppressor gene alterations in invasive bladder carcinoma than in superficial disease, suggesting that PTEN is important in this process. A role of PTEN in bladder cancer invasion is further suggested by the fact that PTEN is a regulator of cell motility, a necessary component of tumor invasion. However, it is unknown whether PTEN is mechanistically involved in 'in vivo' tumor invasion or merely an epiphenomenon and, if the former is true, whether this process is dependent on its protein or lipid phosphatase activities. To address these issues, we stably transfected several commonly used human bladder cancer cell lines with known invasive phenotypes with either wild-type PTEN constructs or those deficient in the lipid phosphatase (G129E) or both protein and lipid phosphatase (G129R) activities. Here we show that chemotaxis was inhibited by both the wild-type and G129E mutant of PTEN but not by G129R-transfected cells. Using a novel organotypic in vitro invasion assay, we evaluated the impact of wild-type and mutant PTEN transgene expression on the invasive ability of T24T, a human bladder cancer cell line with a functionally impaired PTEN. Results indicate that the G129E mutant blocks invasion as efficiently as wild-type PTEN transfection. In contrast to the wild-type gene, this mutant has no effect on cell clonogenicity in agar. To further establish the role of PTEN in tumor invasion, we evaluated vector- and PTEN-transfected T24T cells in an orthotopic in vivo assay that faithfully reproduces human disease. Microscopic examination of murine bladders at the completion of this experiment parallels the results obtained with the organotypic assay. Our results are the first demonstration: (1) that the inhibitory effects of PTEN on cell motility translate into suppression of in vivo invasion; (2) that PTEN can inhibit tumor invasion even in the absence of its lipid phosphatase activity; (3) how organotypic in vitro approaches can be used as surrogates of in vivo invasion allowing rapid dissection of molecular processes leading to this phenotype while reducing the number of animals used in research.