To understand the mechanism of Axl signaling, we have initiated studies to delineate downstream components in interleukin-3-dependent 32D cells by using a chimeric receptor containing the recombinant epidermal growth factor (EGF) receptor extracellular and transmembrane domains and the Axl kinase domain (EAK [for EGF receptor-Axl kinase]). We have previously shown that upon exogenous EGF stimulation, 32D-EAK cells are capable of proliferation in the absence of interleukin-3. With this system, we determined that EAK-induced cell survival and mitogenesis are dependent upon the Ras/extracellular-signal-regulated protein kinase (ERK) cascade. Although the phosphatidylinositol-3 kinase pathway is activated upon EAK signaling, it appears to be dispensable for the biological actions of the Axl kinase. Furthermore, we demonstrated that different threshold levels of Ras/ERK activation are needed to induce a block to apoptosis or proliferation in 32D cells. Recently, we have identified an Axl ligand, GAS6. Surprisingly, GAS6-stimulated 32D-Axl cells exhibited no blockage to apoptosis or mitogenic response which is correlated with the absence of Ras/ERK activation. Taken together, these data suggest that different extracellular domains dramatically alter the intracellular response of the Axl kinase. Furthermore, our data suggest that the GAS6-Axl interaction does not induce mitogenesis and that its exact role remains to be determined.

The Axl receptor tyrosine kinase is a transforming oncogene in NIH3T3 cells. In order to de®ne structural requirements of the Axl receptor necessary for transformation we passaged recombinant retroviruses carrying the axl cDNA in NIH3T3 cells, generating randomly mutated axl variants. Using this strategy, we have isolated three axl viral strains (1B1, SV8, and FFa4) that show augmented 3T3 cell transforming capacity associated with elevated p140 Axl . Upon sequencing, the 1B1 and SV8 proviruses possessed only silent mutations, making p140 Axl overexpression the most likely explanation for their increased transformation activity. However, the characterization of FFa4 revealed a deletion of sequences encoding the carboxy-terminal 45 amino acids leading to the generation of a chimeric transcript comprised of a truncated Axl receptor with a segment of the 3' UTR region. Mutational analysis revealed that the transforming activity of FFa4 was speci®c to the formation of the chimeric receptor rather than to the carboxyl-terminal truncation. Intriguingly, none of the viral strains were able to transform the murine cell lines NR-6 and 32D despite equivalent expression of surface p140 Axl protein. Further analysis showed that Axl's transforming potential is dependent on the host cell type, the presence of a putative pp190 as a facilitator for transformation, and the level of p140 Axl expression. Taken together, these results underscore the complexity of Axl biology which is dependent on receptor stoichiometry and the cellular background.