Abstract 4377: AKT positively regulates Rho-GTP by attenuating the GAP activity of the DLC1 tumor suppressor: a mechanistic study with translational implications

Abstract

Abstract Many human cancers have high AKT activity and constitutive up-regulation of Rho-GTP, which is a major factor in the neoplastic process, but AKT is not known to regulate Rho-GTP. In this study, we show that AKT positively regulates Rho-GTP by directly phosphorylating tumor suppressor DLC1 and attenuating the Rho-GAP activity of DLC1, which catalyzes the conversion of active Rho-GTP to inactive Rho-GDP. Interestingly, AKT was found to increase Rho-GTP and its downstream activities in DLC1-positive cancer lines whether the DLC1 was endogenous or transfected, but not in DLC1-negative lines. Similarly, when epithelial cell lines were stimulated with EGF, they activated AKT, which increased Rho-GTP in a DLC1-dependent manner. Three Serines (S298, S329, S567) in DLC1 have AKT consensus motifs and were phosphorylated by AKT in vitro and in vivo. Their phosphorylation attenuated the Rho-GAP and tumor suppressor activities of DLC1 - including decreased cell migration, focal adhesion turnover, anchorage-independent growth, and tumor growth in mice - as did the combined phosphomimetic mutant DLC1-3D. By contrast, the combined Serine to Alanine mutant DLC1-3A was even more active than wild type DLC1 (DLC1-WT). Remarkably, an AKT inhibitor stimulated the tumor suppressor and Rho-GAP activities of DLC1-WT, but did not influence DLC1-3A or DLC1-3D. The N-terminal half of DLC1 (amino acids 1-600) bound the DLC1 Rho-GAP domain (amino acids 609-878), as determined by in vivo complex formation and by microscale thermophoresis, which measures protein interactions in close-to-native conditions. There was increased binding and decreased Rho-GAP activity when the three Serines in the N-terminus were phosphorylated or carried the 3D mutations. Conversely, there was decreased binding and increased Rho-GAP activity when the Serines were not phosphorylated or carried the 3A mutations. In a xenograft mouse model, AKT inhibitor treatment of DLC1-positive palpable tumors reduced their size, their Rho-GTP level, and their Rho-dependent signaling, but did not affect these three parameters in isogenic DLC1-negative tumors. Similarly, in the MMTV-PyMT breast cancer model, which had high AKT activity, high Rho-GTP, and expressed DLC1, an AKT inhibitor reduced the size of palpable tumors, and reduced Rho-GTP and its downstream signaling. AKT inhibition reduced DLC1 phosphorylation in both tumor models. We conclude: AKT can increase Rho-GTP by phosphorylating three N-terminal Serines in DLC1, which attenuates its Rho-GAP and tumor suppressor functions; the N-terminus of DLC1 is an auto-inhibitory domain that reversibly binds the Rho-GAP domain; AKT attenuates DLC1 functions by phosphorylating the Serines in the N-terminus, which increases its binding the Rho-GAP domain; and AKT inhibition has greater anti-tumor activity in DLC1-positive tumors than in DLC1-negative tumors. Citation Format: Brajendra K. Tripathi, Tiera Grant, Philipp Mertins, Xiaolan Qian, Dunrui Wang, Alex G. Papageorge, Steven A. Carr, Douglas R. Lowy. AKT positively regulates Rho-GTP by attenuating the GAP activity of the DLC1 tumor suppressor: a mechanistic study with translational implications. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4377.