To study the pathogenesis of Burkitt lymphoma, we introduced activated c-myc genes into human EBV-infected lymphoblastoid cells derived from in vitro infection of normal cord blood or directly from infected peripheral blood from AIDS patients. In both cell types the constitutive expression of exogenous c-myc caused negative regulation of endogenous c-myc expression, changes in growth properties typical of transformed cells, and acquisition of tumorigenicity in immunodeficient mice. In all myc-transfected populations the degree of malignancy directly correlated with the level of c-myc mRNA. EBV infection and c-myc activation are thus sufficient for the tumorigenie conversion of human B cells in vitro, strongly supporting the hypothesis that these same two pathogenetic steps may be involved in the in vivo development of Burkitt lymphoma.

Structural and functional alterations of proto-oncogene loci, including mutations, amplifications, chromosomal translocations, and proviral insertions, are consistently found in human and animal tumors and are strongly implicated in tumorigenesis (for review see Bishop, 1987). It is also clear that each of these individual alterations is likely to represent only one step in the tumorigenic process, since the activation of more than one oncogene is generally necessary both for malignant transformation in vitro and for tumorigenesis in vivo (for review see Weinberg, 1982). In this respect, while the repertoire of oncogene activations found in different types of tumors is constantly increasing, very little is known of the number and type of genetic alterations necessary for the development of any single tumor; in fact, the pathogenesis of any naturally occurring tumor has yet to be elucidated at the genetic level. Different types of tumors may involve the activation of identical as well as different, probably tissue-specific, oncogenes, and in some cases one of the genetic alterations may be provided by a tumor virus. The recapitulation of these pathogenetic steps is critical for understanding both carcinogenesis in different tissues and the specific role of different oncogenes in normal and neoplastic ceils. An excellent model system for these types of investigations is Burkitt lymphoma (BL)(for review see Magrath, 1983, 1986) a B cell tumor in which a rather homogeneous morphological phenotype includes various epide-