Functional links between Snail-1 and Cx43 account for the recruitment of Cx43-positive cells into the invasive front of prostate cancer

D Ryszawy, M Sarna, M Rak, K Szpak… - …, 2014 - academic.oup.com
D Ryszawy, M Sarna, M Rak, K Szpak, S Kędracka-Krok, M Michalik, M Siedlar…
Carcinogenesis, 2014academic.oup.com
Suppressive function of connexin (Cx) 43 in carcinogenesis was recently contested by
reports that showed a multifaceted function of Cx43 in cancer progression. These studies did
not attempt to model the dynamics of intratumoral heterogeneity involved in the metastatic
cascade. An unorthodox look at the phenotypic heterogeneity of prostate cancer cells in vitro
enabled us to identify links between Cx43 functions and Snail-1-regulated functional
speciation of invasive cells. Incomplete Snail-1-dependent phenotypic shifts accounted for …
Abstract
Suppressive function of connexin(Cx)43 in carcinogenesis was recently contested by reports that showed a multifaceted function of Cx43 in cancer progression. These studies did not attempt to model the dynamics of intratumoral heterogeneity involved in the metastatic cascade. An unorthodox look at the phenotypic heterogeneity of prostate cancer cells in vitro enabled us to identify links between Cx43 functions and Snail-1-regulated functional speciation of invasive cells. Incomplete Snail-1-dependent phenotypic shifts accounted for the formation of phenotypically stable subclones of AT-2 cells. These subclones showed diverse predilection for invasive behavior. High Snail-1 and Cx43 levels accompanied high motility and nanomechanical elasticity of the fibroblastoid AT-2_Fi2 subclone, which determined its considerable invasiveness. Transforming growth factor-β and ectopic Snail-1 overexpression induced invasiveness and Cx43 expression in epithelioid AT-2 subclones and DU-145 cells. Functional links between Snail-1 function and Cx43 expression were confirmed by Cx43 downregulation and phenotypic shifts in AT-2_Fi2, DU-145 and MAT-LyLu cells upon Snail-1 silencing. Corresponding morphological changes and Snail-1 downregulation were seen upon Cx43 silencing in AT-2_Fi2 cells. This indicates that feedback loops between both proteins regulate cell invasive behavior. We demonstrate that Cx43 may differentially predispose prostate cancer cells for invasion in a coupling-dependent and coupling-independent manner. When extrapolated to in vivo conditions, these data show the complexity of Cx43 functions during the metastatic cascade of prostate cancer. They may explain how Cx43 confers a selective advantage during cooperative invasion of clonally evolving, invasive prostate cancer cell subpopulations.
Oxford University Press