Preview

PDF, English Download (3MB) | Terms of use

Abstract

Integrins are metazoan heterodimeric receptors for cell adhesion and interaction with extracellular matrix. Nowadays 18 integrin alpha chains and 8 beta chains are known in human cells, forming 24 different heterodimers specific for wide array of ligands, including collagens, laminins, and fibronectin. Dynamic assembly and disassembly of ligand-bound integrins as well as rearrangement of unengaged integrins on a cell surface depends on the intracellular trafficking. It is crucial for the processes related to cell adhesion, spreading, and migration, such as neutrophil migration or metastasis development in cancer. Number of studies addressed endocytosis and recycling of integrins, yet, the comprehensive model of integrin trafficking is still missing as the pathways of integrin retrograde trafficking can differ depending on the integrin activation status, binding to the ligand, growth factors presence. An additional level of complexity dealt in the integrin trafficking studies is a certain promiscuity in integrin α and β chain interactions, therefore the knowledge on the specificity of a trafficking pathway for a certain heterodimer is often lacking. This study focused on internalization regulators of integrin α2β1, which is a ubiquitously expressed receptor of collagens, mainly collagen I, one of the most abundant extracellular matrix proteins, which is known as a metastasis suppressor in human cancer. Studies on α2β1 endocytosis are sparse and mostly focused so far on the trafficking of α2β1 clustered by addition of virus or ligand. To identify and characterize novel regulators of internalization of non-clustered integrin α2β1, we established a quantitative fluorescent-microscopy-based integrin α2 internalization assay. We demonstrated that bulk internalization of endogenously expressed non-clustered integrin α2 in HeLa cells is dependent on both clathrin and caveolin1, and occurs via Rab5-, Rab4- and Rab11-positive endosomes. Next, we used RNAi to screen 386 cytoskeleton and cytoskeleton-associated genes for their involvement in integrin α2 internalization. We identified 122 primary screening hits as potential regulators of integrin α2 trafficking, number of which were previously known as regulators of endocytosis, focal adhesion formation, and cell migration, and as direct interactors of integrins. In the validation assays we could reproduce the results of the primary screening for 43% of expressed hits in the two groups of hits selected for validation (14 molecules that were targeted by multiple siRNAs in the primary screening, and 12 kinesin molecular motors). Three validated kinesin hits have never previously been associated with endocytic trafficking events. To get an insight in the function of kinesins in integrin internalization, we further characterized the role of KIF15 in the trafficking of integrin α2. It was identified as a strong inhibitor of the early steps of integrin α2 endocytosis, but not transferrin or EGF. We demonstrated that knock-down of KIF15 lead in re-location of the integrin-specific clathrin adaptor Dab2 from the plasma membrane, which is likely to be the reason for the block of integrin α2 internalization. Taken together, this study establishes the methods to address the internalization of non-clustered integrin α2β1, identifies the number of novel regulators of integrin trafficking, and provides insight into the role of KIF15 molecular motor in the integrin α2 internalization.