We have recently found that sequences of alpha-tubulin and gamma- actin cDNA clones derived from the human leukemic cell line HL-60 contain evidence of multiple somatic mutations. These changes, occurrings in highly conserved regions, suggest a causual relationship to the neoplastic growth of HL-60 cells. The alpha- tubulin changes include drastic alterations of a GTP-binding site homology. Out of the very small number of cancer-cell cytoskeletal proteins examined by ourselves and others, three have shown multiple mutations. Taken together, these and other results have prompted us to propose the existence of a new class of oncogenes which involves the cytokeletal system. Cytoskeletal oncogenes would appear to have roles different from the nuclear-acting (e.g. myc, fos) or the growth-factor/receptor/protein-kinase (e.g., vis, erb-B, ras) classes of oncogenes. To further define the role of cystokeletal mutations occurring in neoplastic cells we will first isolate genomic clones of the mutant gamma-actin and alpha-tubulin genes from HL-60 cells. We will then directly test the oncogene potential of these genes in model systems for multistage tumorigenesis. Because of the assistance of expert collaborators we will be able to test the oncogene potential of these mutant genes in myeloid, lymphoid and epithelial tumorigenesis in addition to the more conventional fibroblast systems. These systems will permit us to detect collaboration between mutant cytoskeletal genes and others such as ras and myc. We may detect such collaborations as increase in anchorage independence, decreases in latent period or effects on the invasiveness of resulting tumors. We will also use new rapid screening procedures to screen a large number of tumor cell lines or fresh tumors for evidence of cytoskeletal mutations in other neoplasms. RNase A detecting candidate mutants. The new polymerase Chain Reaction technique provides a rapid way to clone any such candidates in order to precisely determine the nature of the mutations involved. These experiments will test the suggestion that a high percentage of cancer cells may contain cytoskeletal mutations and investigate the way in which cytoskeletal mutations may collaborate with other oncogenes in cancer initiation or progression.