Abstract
Transgenic mice carrying oncogenes present a useful model with which to assess the tissue‐specific action of oncogenes. These mice are usually predisposed to a specific type of neoplastic growth. The tumors that arise are usually monoclonal in origin and become only apparent after a variable latency period, suggesting that additional events are required for tumor formation. Identification of these additional events is highly relevant: it might give access to the genes that can synergize with a preselected oncogene in tumorigenesis and could facilitate the identification of the biochemical pathways in which these genes act. Retroviruses can be instrumental in identifying cooperating oncogenes. Proto‐oncogene activation or tumor suppressor gene inactivation by insertional mutagenesis is an important mechanism by which the non‐acute transforming retroviruses can induce tumors in several species. Owing to the sequence tag provided by the provirus, the relevant proto‐oncogene can be directly identified by cloning of the DNA flanking the proviral insertion site. We have exploited this potential of retroviruses by infecting Eμ‐pim‐1 and Eμ‐myc transgenic mice, which are predisposed to lymphomagenesis, with Moloney murine leukemia virus (MuLV). A strong acceleration of tumor induction ensued upon infection of these mice with MuLV. More importantly, it allowed us to identify a number of additional common insertion sites marking both previously known as well as new (putative) oncogenes. In a significant portion of the tumors more than one oncogene was found to be activated, indicating that within this system the synergistic effect of at least three genes can be established. Furthermore, the distribution of proviruses over these common insertion sites suggests that this methodology enables us to assign each of these genes to distinct complementation groups in tumorigenesis. To get insight into the function of these oncogenes, we have inactivated one of them, the pim‐1 oncogene, via homologous recombination in ES cells. We have shown that by using a pim‐1 targeting vector in which transcription and translation of the resistance marker was made dependent on the acquisition of both a promoter and an in‐frame translational initiation codon, the gene could be inactivated with high efficiency. Subsequently, we inactivated the second allele by using a different selectable marker. The absence of the pim‐1 gene product does not appear to have any detrimental effect on the growth of ES cells and hematopoietic cells derived from these ES cells in vitro, despite the fact that pim‐1 is highly expressed in wild‐type ES cells. More detailed analysis of pim‐1‐less cells and mice will be necessary to gain insight into the relevance of the highly conserved pim‐1 proto‐oncogene in development and maintenance of mammalian organisms.