Structural and biochemical study of kindlin-3 in integrin activation

Kindlin-3, a member of the FERM family proteins, functions as an adaptor between cell membrane and cytoskeleton proteins. It is mostly expressed in hematopoietic and endothelial cells. Leukocyte adhesion deficiency syndrome (LAD-III) in human has been linked to the expression of mutant Kindlin-3 proteins, which is defective in integrin activation. Kindlin-3 has been implicated in mediating cell-ECM (extracellular matrix) inside-out and outside-in signaling by co-activating integrin β tails with Talins. Interestingly, Talin and Kindlin-3 independently interact with different regions of integrin β tail. However, it is still unknown how do they cooperate with each other in order to activate integrin. Despite its importance in cell signaling, the structure of fulllength Kindlin-3 has not been characterized. Furthermore, the mechanism of how it interacts with its various binding partners in order to conduct cell signaling remains largely elusive. To address these issues, we aimed to biochemically characterize full-length Kindlin-3, as well as determine the structure of its apo form and in complex with interacting partners. For large-scale purification of Kindlin-3, both bacterial and insect cell based expression systems were employed. Curiously, while Kindlin-3 expressed in E. coli exhibited monomer and hexamer forms in solution, Kindlin-3 expressed in insect cells revealed monomer and trimer forms. We succeeded in crystallizing the monomer form of Kindlin-3 from E. coli expression system that diffract to 3.5Å. We determined Kindlin-3 structure using SAD (Single-wavelength anomalous dispersion) method, revealing a trimer with monomers connected via PH and F3 subdomains. Surprisingly, in our structure, the integrin β tail binding sites of F3 subdomains are occupied by the PH domains from the adjacent molecules. Therefore, we propose that there might be an auto-inhibition mechanism, by which, the F3 subdomains can be available or hidden for integrin depending on the polymerization state of Kindlin-3. This would provide means for regulation of integrin activation. In addition, auto-inhibition of Kindlin-3 could be mediated by post-translational modifications we described in PH domain. The large molecular size of the Kindlin-3 hexamer allowed its structure characterization using negative staining and cryo-EM. While a triangular structure made up of six monomers could be visualized, detailed structure characterization is unfortunately not possible at 9.84Å. However, the crystals of the hexamer form of Kindlin-3 without the flexible loop were obtained and diffract to 4Å. Hopefully shedding more light on Kindlin-3 structure. In addition, we have also purified the human full-length Kindlin-2 from insect cell expression system. In contrast to Kindlin-3, the full-length Kindlin-2 is in monomer and tetramer form in solution. Tetramer form was used for negative staining microscopy hinting at the conformation of Kindlin-2, but further structural and biochemical studies are still ongoing. Mass spectrometry determined numerous posttranslational modifications in Kindlin-2, several of which have never been determined before. Also, the functions of these post-translational modifications remain to be characterized. Taken together, our structure sheds new light on the mechanism of integrin activation and cell signaling regulation by Kindlin-3. Furthermore, structural characterization of Kindlin-3 can be useful for LADIII therapeutic methods study and researched about other Kindlins.​Doctor of Philosophy (SBS