A total reconstitution of receptor-mediated Ras activation on supported-lipid-membrane microarrays showed that the bimodal, switch-like Ras activation response is not produced by classic deterministic (kinetic) bistability but by the extreme processivity of the Ras activator SOS, which activates hundreds of Ras molecules during a single membrane-binding event, so that stochastic bursts at low copy number split the activation distribution into two modes. The source work left three linked questions open: whether this processivity-driven stochastic bimodality mechanism operates in other signaling proteins with comparable regulatory complexity (PLC-γ, PI3K, N-WASP); what precise biophysical mechanisms let LAT scaffold condensation and Ras clustering engage or disengage the processive SOS activity; and whether the simulated resistance of the bimodal response to pharmacological inhibition holds in cellular or in vivo contexts. Follow-up work has clarified pieces of each arm — the autoinhibition-release origin of SOS positive feedback, theoretical models of LAT condensates, and reconstitutions of PLCγ1 gating and Ras–PI3Kγ feedback — but none of the three questions is resolved.

Background

Ras is a small GTPase, a molecular switch that toggles between an active GTP-bound and an inactive GDP-bound state and sits at the center of the Ras–MAPK growth-control pathway. Ras is switched on by SOS, a guanine nucleotide exchange factor (GEF) that catalyzes GDP release from Ras, and switched off by GTPase-activating proteins such as p120 RasGAP. Two features of SOS are central here. First, SOS is highly processive: once engaged on the membrane, a single SOS molecule can activate hundreds of Ras molecules during one membrane-binding event. Second, SOS is regulated by an allosteric Ras-binding site and by an autoinhibition mechanism that is released slowly after SOS is recruited to membrane scaffolds. Switch-like, bimodal Ras activity (a two-peaked distribution of activation states across otherwise identical reactions) has long been observed in Ras–MAPK signaling and is conventionally attributed to deterministic bistability — two stable steady states generated by nonlinear positive feedback in the rate equations, for example allosteric activation of SOS by RasGTP. A distinct possibility is stochastic bistability: a two-peaked distribution arising from stochastic fluctuations at low molecular copy number in a system whose deterministic rate equations have a single steady state. Lee, Kim, Alvarez, Ren, DeGrandchamp, Lew, and Groves (Science Advances, 2024; first posted as a 2023 bioRxiv preprint) reconstituted the complete receptor-mediated Ras activation–deactivation reaction — phosphorylated LAT scaffold, recruitment of SOS by the adaptor protein Grb2, SOS-catalyzed exchange, and GAP-mediated deactivation — on supported lipid bilayers (artificial membranes on a solid support) patterned into micron-scale corrals that confine the reaction to physiological molecular copy numbers. They found a bimodal Ras activation response that (i) is not deterministic bistability, because the deterministic rate equations with measured kinetics have a unique steady state and the bimodality disappears as reaction size grows; (ii) requires SOS processivity, because constructs with weak processivity show no bimodality even with positive feedback; and (iii) is controlled by the condensation state of LAT (Linker for Activation of T cells, the T-cell receptor scaffold protein that forms condensed phase-separated clusters) and by Ras clustering and local Ras density. The condensation of multivalent scaffold proteins such as LAT into dense clusters (biomolecular condensates formed by liquid–liquid phase separation) is a general feature of receptor-proximal signaling. Stochastic simulations using measured rate constants (the Gillespie algorithm, a standard exact method for simulating stochastic chemical kinetics) reproduce the processivity-driven bimodality, and simulations including a drug (fig. S10 of that paper) show that at inhibitor concentrations where deterministic kinetics predict successful suppression, small reaction systems still stochastically activate Ras — a predicted resistance to pharmacological inhibition. Related single-molecule studies from the same group showed that the positive feedback in full-length SOS arises from the autoinhibition-release step rather than allosteric activation (Ren et al., Biophysical Journal, 2024), and several signaling enzymes besides SOS share this architecture of a membrane substrate with a cytosolic enzyme in a competitive activator/deactivator cycle — notably phospholipase C-γ (PLC-γ, which cleaves the membrane lipid PIP2 to generate the second messengers IP3 and DAG), phosphoinositide 3-kinases (PI3K, which phosphorylate phosphatidylinositol lipids), and N-WASP (a regulator of actin polymerization). The reconstitution results raise the unsolved question of how general this mechanism is and what its functional consequences are.

Problem Statement

The overarching objective is to establish whether processivity-driven stochastic bimodality — two-peaked, switch-like activation produced by the extreme processivity of an activating enzyme at low molecular copy number, without any deterministic bistability in the rate equations — is a general mechanism of membrane signaling, and what its functional consequences are, as posed by the SOS–Ras reconstitution of Lee et al. (Science Advances, 2024). Three linked questions define the open core. (1) Generality: does an analogous processivity-driven stochastic bimodality arise in other signaling proteins with comparable regulatory complexity, specifically PLC-γ, PI3K, and N-WASP (or, conversely, is a different bimodality mechanism realized in those systems, and if so which)? Here the three named proteins delimit the class of interest rather than acting as conjunctive requirements; a demonstration (or refutation) for any one of them is an independently refutable contribution to this arm, while the overarching generality claim is settled only by the pattern across the class. (2) Mechanism: what are the precise biophysical mechanisms by which LAT condensation and Ras clustering engage or disengage processive SOS activity — testing the hypothesized chain from condensate-driven multivalent engagement of SOS, through prolonged membrane dwell time and autoinhibition release, to processive Ras activation, and establishing whether these control points operate in living cells rather than only in reconstituted membranes. (3) Pharmacological consequence: is the simulated resistance of the processivity-driven bimodal response to pharmacological inhibition (persistence of stochastic Ras-activation bursts at inhibitor concentrations that suppress the deterministic response) realized experimentally and, if so, with what implications for SOS- and Ras-targeted therapeutics? Each arm depends on the source's own criterion of stochastic versus deterministic bimodality, not merely on the observation of a switch-like response.

References

  1. Albert A. Lee, Neil H. Kim, Steven Alvarez, He Ren, Joseph B. DeGrandchamp, L. J. Nugent Lew, Jay T. Groves, 'Bimodality in Ras signaling originates from processivity of the Ras activator SOS without deterministic bistability', Science Advances 10(25), eadi0707 (2024). DOI: 10.1126/sciadv.adi0707, https://doi.org/10.1126/sciadv.adi0707 (published version)
  2. Albert A. Lee, Neil H. Kim, Steven Alvarez, He Ren, Joseph B. DeGrandchamp, L. J. Nugent Lew, Jay T. Groves, 'Bimodality in Ras signaling originates from processivity of the Ras activator SOS without classic kinetic bistability', bioRxiv (2023). DOI: 10.1101/2023.07.17.549263, https://doi.org/10.1101/2023.07.17.549263 (preprint version)
  3. He Ren, Albert A. Lee, L. J. Nugent Lew, Joseph B. DeGrandchamp, Jay T. Groves, 'Positive feedback in Ras activation by full-length SOS arises from autoinhibition release mechanism', Biophysical Journal (2024). DOI: 10.1016/j.bpj.2024.07.014, https://doi.org/10.1016/j.bpj.2024.07.014
  4. Jay T. Groves, 'Single-molecule biophysics in signaling: Functionality from stochastic effects', Current Opinion in Cell Biology (2025). DOI: 10.1016/j.ceb.2025.102597, https://doi.org/10.1016/j.ceb.2025.102597
  5. Shumpei Morita, Jay T. Groves, 'Parallel reactions on a single T cell receptor offer a robust kinetic proofreading mechanism', Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2514057122, https://doi.org/10.1073/pnas.2514057122
  6. Yannick A. D. Omar, Simou Sun, Mehran Kardar, Jay T. Groves, Arup K. Chakraborty, 'A unifying model of LAT condensates in reconstitution experiments', bioRxiv (2025). DOI: 10.1101/2025.10.31.685772, https://doi.org/10.1101/2025.10.31.685772
  7. He Ren, H.-R. Lee, Yannick A. D. Omar, Joseph B. DeGrandchamp, C.-L. Hung, T. J. Eisen, D. Stamou, J. Kuriyan, Arup K. Chakraborty, Jay T. Groves, 'LAT condensation gates PLCγ1 activation via bimodal LAT phosphorylation', bioRxiv (2026). DOI: 10.64898/2026.04.29.721478, https://doi.org/10.64898/2026.04.29.721478
  8. Sophia Doerr, Andrés Olavarrieta-Colasurdo, Scott D. Hansen, 'Reconstitution of Ras-PI3Kγ membrane communication and feedback using light-induced signaling inputs', Nature Communications (2026). DOI: 10.1038/s41467-026-75652-y, https://doi.org/10.1038/s41467-026-75652-y
  9. Katherine A. Yonosh, James R. Faeder, 'Mathematical modeling predicts quantization of Ras activation in membrane corrals', Research Square preprint (2025). DOI: 10.21203/rs.3.rs-7889340/v1, https://doi.org/10.21203/rs.3.rs-7889340/v1