To study the effect of DDR inhibition by
To study the effect of DDR inhibition by imatinib, nilotinib and dasatinib in a cellular environment we generated DDR1 and DDR2 expressing HEK293 cell lines by stable transfection. Of the five DDR1 splice variants we focused on the 1b variant whose over-expression has been associated with pulmonary disorders (Matsuyama et al., 2005a, Matsuyama et al., 2005b) although in preliminary experiments DDR1a was also found to be active in the assay (data not shown) whereas the c, d and e variants lack kinase activity. Using receptor auto-phosphorylation as the read-out we determined the IC50 values for all 3 compounds. We found the ranking order of the three compounds to be the same as in the biochemical assay; with imatinib being the least and dasatinib the most potent inhibitor. Interestingly, when comparing IC50 values obtained in the two different assays, both imatinib and nilotinib were found to be approximately 10-fold more potent in the cellular assay than in the kinase assay. The observed differences can be due to the fact that in the biochemical assay the constitutive activity of the intracellular domains was assayed for, whereas in the cellular context the fully functional, membrane spanning, collagen activated receptors were tested, which were likely to have different conformational states. As described above DDR1 expression and activity has been associated with various cancers and pulmonary and fibrotic disorders. While DDR1 was primarily identified as an epithelial receptor with a potential role in various carcinomas, intriguingly a unique role for DDR1 has also been hypothesized in monocytic 5-(N,N-dimethyl)-Amiloride hydrochloride in idiopathic pulmonary fibrosis (Matsuyama et al., 2005a). It was found that CD14+ monocytes/macrophages isolated from bronchoalveolar lavage fluid of idiopathic pulmonary fibrosis patients had elevated levels of DDR1 expression, particularly of the 1b isoform. Furthermore, stimulation of these cells with collagen via DDR1 leads to the generation of a number of pro-inflammatory chemokines believed to play a role in the inflammatory component of the disease (Agostini and Gurrieri, 2006). In order to evaluate the inhibitory potential of imatinib, nilotinib and dasatinib in a disease relevant cellular assay we generated DDR1b over expressing THP-1 monocytic cell lines using lentiviral delivery of the expression constructs. Imatinib and dasatinib both inhibited MCP-1 release and similarly to the biochemical and the receptor auto-phosphorylation assays, dasatinib was found to be more potent although both compounds had significantly higher IC50 values than in the other two assays. Intriguingly, however, nilotinib did not inhibit MCP-1 production even at the highest concentration tested. To better understand this unexpected result we evaluated the effect of nilotinib on receptor auto-phosphorylation in DDR1-THP-1 cells and found that nilotinib appeared to be inactive in this assay, whereas imatinib was shown to inhibit receptor auto-phosphorylation. This is in agreement with the lack of potency of nilotinib in the MCP-1 release assay, however, the underlying reason is presently unclear. It is conceivable that a lack of intracellular concentrations of nilotinib in THP-1 cells plays a role in this observation, although in systems evaluated thus far nilotinib readily penetrates cells by passive diffusion and is not a substrate for efflux pumps and therefore further studies are needed to fully understand this finding. Future experiments include the testing of compounds with similar chemical structures to imatinib and nilotinib in the MCP-1 release assay to determine if there is a structural basis for this observation. In summary, we have characterized in detail the inhibitory effect of imatinib, nilotinib and dasatinib on the collagen receptor tyrosine kinases DDR1 and DDR2. Our findings may contribute to the better understanding of the therapeutic potential of these compounds and highlight the potential of these and related kinase inhibitors for the treatment of diseases associated with pathological DDR1 or DDR2 activity. For example, both DDR1 and DDR2 have been associated with fibrotic disorders and imatinib has been shown to be effective in animal models of fibrosis (Daniels et al., 2004, Aono et al., 2005). However, while imatinib is a kinase inhibitor with a relatively narrow selectivity profile further investigations are needed to delineate whether its observed anti-fibrotic effect is due to its activity against a particular kinase, such as PDGF receptor or DDR1 or 2. Alternatively, the potential of imatinib to attenuate the fibrosis in in vivo models may be due to its activity against more than one kinase involved in the fibrotic process.