Design principles of pleiotropic G‐protein signaling through guanine nucleotide exchange modulators (GEMs)

Abstract

Listen

Translate article

Heterotrimeric (henceforth trimeric) G‐proteins are molecular switches that control signal transduction. Dysregulation of the G‐protein pathway can lead to aberrant signal transduction and herald oncogenesis. Canonical G‐protein signaling, which is triggered exclusively by G protein‐coupled receptors (GPCRs), is spatially and temporally restricted, i.e. triggered exclusively at the plasma membrane (PM) via a process that completes within a few hundred milliseconds. Recently, a new paradigm has emerged, in which tyrosine‐based signals initiated by growth factor receptor tyrosine kinases (RTKs; e.g., EGFR) are relayed via non‐canonical activation of trimeric G proteins; such signaling has unique spatiotemporal features and is orchestrated by a non‐receptor guanine exchange modulator (GEM), GIV (a.k.a Girdin). GIV‐GEM acts pleiotropically as a guanine exchange factor (GEF) for Gαi and a guanine dissociation inhibitor (GDI) for Gαs; such modulation of two opposing Gα‐proteins integrates downstream cAMP signals depending on the abundance of functional copies in cells, fueling sinister properties such as invasiveness, stemness, survival, chemoresistance and angiogenesis. Here we developed the first network‐based compartmental model for signal transduction between RTKs and G proteins; two pathways of paramount importance in biology. The primary goal of this model was to identify the mechanistic regulation of cAMP, PKA, and CREB in different cellular compartments over time and to predict how network crosstalk affects cAMP dynamics. The model was constrained by experimental data and was robust to changes in kinetic parameters but sensitive to changes in initial concentrations of signaling species. Our model not only captured the unique temporal aspects of RTK→GIV→Gαi/s→cAMP signaling, but also predicted previously unforeseen impacts of EGF signaling [receptor endocytosis, copy number variations, activating mutations, etc.] and PDE activity on GIV‐dependent cAMP modulation. Predictions of the model appear to hold true in various disease states, most prominently in cancers.