During preparation of this manuscript

During preparation of this manuscript, some studies investigating effects of Sunitinib on Axl activity in renal cancer have been published [44], [45], [46], [47]. However, there are several important considerations to take into account interpreting results of these studies. First, Sunitinib doses that are claimed to inhibit or stimulate Axl activity and caffeic acid phenethyl ester are very high and extensively off target in comparison to IC50 values to true Sunitinib targets. Also, such high Sunitinib concentrations are physiologically irrelevant resulting in target doses coupled to intolerable side effects in patients and is not approved by the FDA. [26], [48] Second, a commercial Gas6 that is not fully active due to deficient y-carboxylation [4] is used. Third, Axl-activation is presented using an antibody intended to identify Axl-pTyr702 that does not illustrate Axl-phosphorylation in total. Altogether, these issues make earlier reports on Sunitinib effects on Axl activity difficult to interpret. We have prepared a high quality Gas6 that has been thoroughly characterized and we have performed appropriate Axl IP and IB for entire Axl-phosphorylation. Hence, convinced that we are correctly analyzing Gas6-Axl phosphorylation, we do not find that Sunitinib inhibits Gas6-mediated Axl activation, as has been suggested in some studies [44]. Nor does Sunitinib alone stimulate Axl levels or activation as suggested in others [45], [46], [47]. Qu et al. [47] nicely studied Axl involvement in Sunitinib-related CCRCC resistance using a functional antibody recognizing Axl-pTyr779, but unfortunately the concentration of Sunitinib used was too high to state relevance in real life. We use physiologically relevant Sunitinib doses comparable to target dose in circulation of patients being treated with Sunitinib [26], [48]. However, locally within the tumor, Sunitinib concentrations potentially may vary from concentration in circulation, and therefore we also assayed higher and lower Sunitinib concentrations ranging 10-fold in both directions with similar results. Interestingly, we found that Gas6+Sunitinib enhanced CCRCC-mediated OPN-secretion. OPN is a multi-tumorigenic protein driving tumor-angiogenesis through other mechanisms than those targeted by Sunitinib. Also, OPN drives generation of CAFs from normal fibroblasts. [49] CAFs secrete tumor-promoting factors beneficial for tumor growth, survival and metastasis and polarization of infiltrating macrophages to become tumor-supporting [29]. Importantly, increased OPN levels in circulation of CCRCC patients are associated with higher metastatic burden [38]. Additionally, we found Gas6+Sunitinib to increase MMP7-formation and to enhance OPN-cleavage. Circulatory MMP7 levels in CCRCC patients is, like OPN, associated with decreased survival and enhanced metastasis [50], and increased MMP7-mediated OPN-fragmentation is associated with enhanced invasiveness of cancer cells [32]. When investigating random migration that mimics EMT-induced migration, we found that Gas6+Sunitinib affects the migratory phenotype of CCRCC cells with an early inhibition followed by a late stimulation of migration, potentially due to phenotypic changes of CCRCC cells to become more motile. Indeed, in cells grown in low serum where Gas6 induces migration on its own, we found Gas6 to enhance expression of SLUG, an EMT-inducer known to increase CCRCC metastasis [30]. To our knowledge cells cannot undergo transformation and migrate simultaneously, fitting an early inhibition of migration during cell transformation, followed by a “catch-up” of cells that are more motile and a later gained increased migration compared to normal cells. Others have reported a Gas6-stimulatory migration [14], but we found that Gas6 alone did not stimulate CCRCC cell migration in full serum, while doing so in lower serum concentrations. The divergent effects of Gas6 in various serum concentrations might explain this discrepancy in between studies. Another noteworthy and important factor is that in their migration experiment Rankin et al. [14] used the commercial Gas6 lacking full y-carboxylation. Hence, their effect seen on migration cannot be concluded to be due to proper Gas6-dependent Axl activation.