Cell-matrix interactions : master regulators of cancer cell fate?

Sammanfattning: The development and homeostasis of a multicellular organism require fundamental biological processes like cell proliferation, cell differentiation, cell migration, and controlled cell death. The extracellular matrix (ECM) guides many of these functions, via cell-matrix interactions that function as mechanical and biochemical signaling hubs. Changes in the ECM composition or organization may impact cellular behavior, both in health and disease. In this thesis, I have explored the effects of cell-extracellular matrix interactions on cellular processes, with a special focus on elucidating the molecular underpinnings of how extracellular matrix stiffening regulates breast cancer cell phenotypes. In study I, we identified and characterized a new class of integrin-containing adhesion complex that we named “reticular adhesions” (RAs). They were formed by integrin αVβ5 in the absence of classical adhesion components like talin-1, vinculin, and F-actin. Unlike classical adhesions, they persisted throughout cell division during which they provided ECM anchoring necessary for efficient division and spatial memory transmission between cell generations. The characterization of reticular adhesions thus provided a solution to the long-standing question of mitotic cell-ECM attachment. Studies II, III, and IV, all investigated the effect of ECM stiffness on breast cancer cells. The ECM stiffness increases with breast cancer progression and the stiffening is known to drive breast cancer cell proliferation and invasion. However, the molecular details of this phenomenon are not yet fully understood. In study II, we confirmed a stiffnessinduced phenotypic switch in the high-grade breast carcinoma cell line, MCF10CA1a, with a ductal carcinoma in situ (DCIS) phenotype on a stiffness mimicking normal breast tissue stiffness, and an invasive ductal carcinoma (IDC) phenotype on a slightly higher stiffness, resembling breast tumor stiffness. Transcriptomic profiling of these two cellular states revealed only minor differences. Still, the stiffness-driven shift in mRNA resembled the changes differing IDC lesions from co-occurring DCIS lesions in patients, suggesting that stiffness may contribute to this transition and that hampering the mechanosignaling could prevent the progression of pre-invasive to invasive breast cancer. In study III, we used the same model as in study II, and quantitative mass spectrometry to compare the proteome of the two stiffness-dependent cellular states. The differences were much larger at the protein level, implying a previously underappreciated post-transcriptional regulation of many genes as a result of mechanical signaling. Among the stiffnessregulated genes, we found an enrichment of mevalonate pathway enzymes and confirmed the importance of this metabolic pathway for the stiffness-induced malignant phenotype. One of these enzymes, Hydroxymethylglutaryl-CoA Synthase (HMGCS1), was upregulated in human breast tumor tissue compared to normal breast tissue, and the level of expression was correlated to the collagen organization, suggesting a stiffness-dependent regulation also in patients. Further, the synthesis rate of HMGCS1 depended on integrin and Rac1 signaling and the expression of a constitutively active Rac1 mutant could mimic matrix stiffening and promote HMGCS1 protein levels as well as a malignant phenotype on low stiffness. In study IV, we explored yet another level of regulation in our model when we used peptide chip arrays to profile the kinase activity in the two cellular states. The combination of the kinome profiling with a small siRNA-based screen allowed us to define the inhibitor of nuclear factor kappa-B kinase subunit epsilon, IKBKE, as a mechanosensitive kinase important for the maintenance of the stiffness-induced phenotype. Thus, this thesis provides novel molecular insight into the regulation of cell-matrix interactions in cellular fate, especially on how mechanical properties of the ECM can induce breast cancer stage switching.