Context-Dependent Roles of Rnd3 in Cancer: Revisiting a Functional Paradox

Abstract

Highlights What are the main findings? Rnd3 exerts a conserved molecular action across tumors: inhibition of RhoA/ROCK1-dependent contractility. A single Rnd3-driven inhibition of RhoA/ROCK1 bifurcates into opposite tumor outcomes via distinct downstream effectors. What are the implications of the main findings? Context-specific accessory effectors, not tumor lineage alone, redirect this conserved action. Predicting whether Rnd3 acts as a suppressor or facilitator requires measuring the pre-existing cellular and mechanical state, rather than relying on tumor lineage or Rnd3 levels alone. Abstract Rnd3 is an atypical member of the Rho GTPase family whose activity is mainly regulated by expression, localization and protein stability rather than canonical GDP/GTP cycling. In cancer, Rnd3 has been described both as a tumor suppressor and as a tumor-promoting factor, creating an apparent functional paradox. We propose that this paradox is resolved by a mechanistic invariant: Rnd3 exerts a conserved inhibition of RhoA/ROCK1-dependent actomyosin contractility, whose phenotypic output is redirected by context-specific accessory effectors rather than reversed. In this review, we revisit this paradox by integrating evidence from mechanistic studies, tumor models and patient-associated datasets. We propose that Rnd3 should not be interpreted through a binary oncogene/tumor-suppressor framework, but rather as a context-dependent regulator of tumor cell state. In many tumor settings, Rnd3 repression or loss of Rnd3 function favors proliferation, apoptosis resistance and therapy resistance through pathways involving Notch, NF-κB, EGFR/ERK, EZH2-dependent chromatin regulation, m6A-mediated RNA control, microRNAs and chaperone-mediated autophagy. However, in selected contexts, including RTK-driven glioblastoma, hepatocellular carcinoma, non-small-cell lung cancer, melanoma and gastric cancer, Rnd3 may support tumor fitness, migration or invasive plasticity. We therefore propose a functional stratification model in which Rnd3 output depends on the biological process, tumor lineage, pathway activity and mechanical state of the cell.