Viral genes capable of inducing vascular tumors in the skin of transgenic mice are the tat gene of HIV-1 and polyoma virus' middle T antigen gene. Instead of vascular tumors, the tat gene of HTLV-I causes thymic atrophy and mesenchymal tumors in transgenic mice. No proof exists that any of these genes contribute to the induction of KS but HIV-1 tat is a strong suspect. The gene product K-FGF of the oncogene K-fgf/hst (int) uses bFGF receptors, is homologous with bFGF and acts as a mitogen for fibroblasts, endothelial cells and melanocytes. The overexpression of the K-fgf gene in KS is not proven unequivocally; some doubts exist suggesting the activation of this gene during the laboratory procedure of transfection with KS cell heavy DNA. Growth factor(s) not well identified (IL-6?) are released from HTLV-I- or II, or HIV-1- or 2-infected T4 lymphocytes and in particular from HIV-1-infected macrophages. This growth factor(s) promote(s) the continuous proliferation of endothelial cells and KS cells. AIDS-KS cells release other growth factors identical with or closely related to basic FGF, a major inducer of angioneogenesis. In addition, acidic FGF, IL-1 alpha and -beta, GM-CSF, PDGF-B and TGF-beta are released from AIDS-KS cells. The release of GM-CSF is induced by IL-1. GM-CSF promotes granulocytic, monocytic and endothelial cell proliferation. TGF-beta is known to suppress lymphocyte-mediated cytotoxicity and may act as a local immunosuppressive factor together with interferon inactivators. We theorize that when TGF-beta production ceases, TNF-beta (lymphotoxin) production switches on leading to programmed cell death (apoptosis) of KS cells resulting in regression of these lesions. The newly discovered angiogenesis factors VEGF/VPF may emerge as protooncogene-oncogene products analogous to PDGF and c-sis activation. AIDS-KS heavy DNA transfects NIH3T3 cells. NIH3T3 cells carrying this gene induced angiosarcomas when implanted in mice. An as yet unidentified large virus (mycoplasma?) was derived from these cells during passages in culture. No causative relationship between this agent and Kaposi sarcoma has as yet been established. Even though IFN-alpha exerts antiretroviral effects in AIDS, we propose that the therapeutic effect of IFN-alpha in AIDS-KS is based on antiangiogenesis activity by suppressing protooncogenes-oncogenes of the FGF family.(ABSTRACT TRUNCATED AT 400 WORDS)