Reevaluation of the roles of protein S and gas6 as ligands for the receptor tyrosine kinase Rse/Tyro 3

Recently, Stitt et al. (1995) reported that protein S (PS), but not Gas6, is a potent ligand for the receptor tyrosine kinase known as Rse, Tyro 3, Brt, Sky, and Tif (hereafter referred to as Rse/Tyro 3). PS is an abundant serum protein previously characterized as an essential anticoagu-lant. Gas6, which was identified as a gene whose expression is increased by growth arrest, shares 43% amino acid identity and overall domain organization with PS (Manfio-letti et al., 1993). Stitt et al. (1995) based their conclusions on experiments describing interspecies interactions of bovine Gas6 with murine Rse/Tyro 3 (mRse/Tyro 3) and on interactions of bovine and human PS (hPS) with mRse/ Tyro 3. Consistent with the results of Stitt et al. (1995), we identified bovine PS as a ligand for human Rse/Tyro 3 (hRse/Tyro 3), and we also found that hPS can act as a ligand for mRse/Tyro 3. However, when we analyzed the more relevant intraspecies interactions, we obtained different results. We found that human Gas6 (hGas6), but not hPS, acted as a potent ligand for hRse/Tyro 3 (Figure 1). The hRse/Tyro 3 we have studied is very likely the true homolog of mRse/Tyro 3 since they share 90% amino acid identity and a similar expression pattern (Mark et al., 1994; Lai et al., 1994). To characterize the Rse/Tyro 3 ligand, we constructed soluble receptor proteins containing the extraceltular domain of either hRse/Tyro 3 or mRseFryro 3 fused to the Fc portion of human immunoglobulin G1 (hRse-lgG and mRse-lgG). A similar fusion protein (termed Tyro 3-Fc) was utilized by Stitt et al. (1995) to characterize the binding of bovine Gas6 and hPS to mRse/Tyro 3. We first determined whether hRse-lgG or mRse-lgG differed in its ability to bind to hPS or hGas6 containing an epitope tag that allows for side-by-side comparison of the binding properties of the two proteins. Either h Rse-lgG or mRse-lgG was incubated with conditioned medium containing putative ligands. Complexes were captured with protein A (specific for the IgG fusion protein) and visualized with an antibody specific for the epitope-tagged putative ligand. While hGas6 was bound by hRse-lgG, it was not efficiently bound by mRse-lgG (Figure 2). The reciprocal result was obtained in analysis of binding of hPS to hRse-lgG and mRse-lgG; hPS was bound by mRse-lgG but not by hRse-lgG. These results are consistent with an apparent difference in affinity of hPS for human as opposed …