RESUMO
Persistent HGF/Met signaling drives tumor growth and dissemination. Proteoglycans within the tumor microenvironment might control HGF availability and signaling by affecting its accessibility to Met (HGF receptor), likely defining whether acute or sustained HGF/Met signaling cues take place. Given that betaglycan (BG, also known as type III TGFß receptor or TGFBR3), a multi-faceted proteoglycan TGFß co-receptor, can be found within the tumor microenvironment, we addressed its hypothetical role in oncogenic HGF signaling. We found that HGF/Met promotes lung cancer and endothelial cells migration via PI3K and mTOR. This effect was enhanced by recombinant soluble betaglycan (solBG) via a mechanism attributable to its glycosaminoglycan chains, as a mutant without them did not modulate HGF effects. Moreover, soluble betaglycan extended the effect of HGF-induced phosphorylation of Met, Akt, and Erk, and membrane recruitment of the RhoGEF P-Rex1. Data-mining analysis of lung cancer patient datasets revealed a significant correlation between high MET receptor, HGF, and PREX1 expression and reduced patient survival. Soluble betaglycan showed biochemical interaction with HGF and, together, they increased tumor growth in immunocompetent mice. In conclusion, the oncogenic properties of the HGF/Met pathway are enhanced and sustained by GAG-containing soluble betaglycan.
RESUMO
Systematic analysis of tumor transcriptomes, combined with deep genome sequencing and detailed clinical assessment of hundreds of patients, constitutes a powerful strategy aimed to identify potential biomarkers and therapeutic targets to guide personalized treatments. Oncogenic signaling cascades are integrated by multidomain effector proteins such as P-Rex1, a guanine nucleotide exchange factor for the Rac GTPase (RacGEF), known to promote metastatic dissemination of cancer cells. We hypothesized that patients with high P-Rex1 expression and reduced survival might be characterized by a particular set of signaling proteins co-expressed with this effector of cell migration as a central component of a putative signaling hub indicative of poor prognosis. High P-Rex1 expression correlated with reduced survival of TCGA Lower Grade Glioma (LGG) patients. Thus, guided by PREX1 expression, we searched for signaling partners of this RacGEF by applying a systematic unbiased in silico data mining strategy. We identified 30 putative signaling partners that also correlated with reduced patient survival. These included GPCRs such as CXCR3, GPR82, FZD6, as well as MAP3K1, MAP2K3, NEK8, DYRK3 and RPS6KA3 kinases, and PTPN2 and PTPN22 phosphatases, among other transcripts of signaling proteins and phospho-substrates. This PREX1 signaling hub signature correlated with increased risk of shorter survival of LGG patients from independent datasets and coincided with immune and endothelial transcriptomic signatures, indicating that myeloid infiltration and tumor angiogenesis might contribute to worsen brain tumor pathology. In conclusion, P-Rex1 and its putative signaling partners in LGG are indicative of a signaling landscape of the tumor microenvironment that correlates with poor prognosis and might guide the characterization of signaling targets leading the eventual development of immunotherapeutic strategies.
RESUMO
Gßγ marks the inner side of the plasma membrane where chemotactic GPCRs activate Rac to lead the assembly of actin filaments that push the cell to move forward. Upon dissociation from heterotrimeric Gi, Gßγ recruits and activates P-Rex1, a Rac guanine nucleotide exchange factor (RacGEF). This cytosolic chemotactic effector is kept inactive by intramolecular interactions. The mechanism by which Gßγ stimulates P-Rex1 has been debated. We hypothesized that Gßγ activates P-Rex1 by a two-step mechanism based on independent interaction interfaces to recruit and unroll this RacGEF. Using pulldown assays, we found that Gßγ binds P-Rex1-DH/PH as well as PDZ-PDZ domains. These domains and the DEP-DEP tandem interact among them and dissociate upon binding with Gßγ, arguing for a stimulatory allosteric effect. In addition, P-Rex1 catalytic activity is inhibited by its C-terminal domain. To discern P-Rex1 recruitment from activation, we studied Q-Rhox, a synthetic RhoGEF having the PDZ-RhoGEF catalytic DH/PH module, insensitive to Gßγ, swapped into P-Rex1. Gßγ recruited Q-Rhox to the plasma membrane, indicating that Gßγ/PDZ-PDZ interaction interface plays a role on P-Rex1 recruitment. In conclusion, we reconcile previous findings and propose a mechanistic model of P-Rex1 activation; accordingly, Gßγ recruits P-Rex1 via the Gßγ/PDZ-PDZ interface followed by a second contact involving the Gßγ/DH/PH interface to unleash P-Rex1 RacGEF activity at the plasma membrane.