The guanine-nucleotide-exchange factor SOS (Son of Sevenless) is the principal conduit from receptor tyrosine kinases to Ras, and positive feedback in SOS-mediated nucleotide exchange is the textbook explanation for switch-like (bimodal, hysteretic) Ras activation in cells. Two decades of structural and cellular work attributed this feedback to selective allosteric activation of SOS by Ras•GTP at a distal regulatory site. Recent single-molecule reconstitution of full-length SOS has dismantled central premises of that account: the mean catalytic rate of an individual SOS molecule is independent of the nucleotide state at the allosteric site, a fully reconstituted Ras activation–deactivation cycle produces bimodal Ras output from SOS processivity without deterministic bistability, and kinetic analysis of receptor-recruited full-length SOS locates the ensemble positive feedback in Ras•GTP-accelerated release of autoinhibition. Whether any of the candidate mechanisms — allosteric catalytic enhancement, autoinhibition release, membrane-intermediate-state stabilization, or processivity gated by scaffold condensation — is the operative source of feedback and of graded-to-digital conversion in living cells remains undetermined. The problem is to establish, by direct measurement and causally separable perturbations in cells, which mechanism the SOS pathway actually uses, and thereby how graded receptor inputs are converted into graded versus switch-like Ras outputs.

Background

Ras proteins are small GTPases anchored at the plasma membrane that act as molecular switches, cycling between an inactive GDP-bound and an active GTP-bound state; in the active state Ras recruits effectors such as Raf to launch the MAPK (mitogen-activated protein kinase) cascade controlling proliferation, differentiation, and cell-fate decisions. Loading of GTP onto Ras is catalyzed by guanine-nucleotide-exchange factors (GEFs). The principal GEF downstream of receptor tyrosine kinases is SOS, recruited to activated receptors (EGFR, the epidermal growth factor receptor, or the LAT (linker for activation of T cells) scaffold of the T-cell receptor) through the adaptor protein Grb2, which binds the C-terminal proline-rich (PR) domain of SOS.

SOS is a multi-domain enzyme whose catalytic core — the REM (Ras exchange motif) and Cdc25 (catalytic) domains — is held in an autoinhibited conformation by the N-terminal histone-fold (HF), Dbl-homology (DH), and pleckstrin-homology (PH) domains together with the PR domain. Uniquely among RasGEFs, SOS carries two Ras binding sites: the catalytic site on the Cdc25 domain, where Ras•GDP is opened for nucleotide exchange, and a distal allosteric site between the REM and Cdc25 domains. Crystal structures showed Ras•GTP bound at this allosteric site and demonstrated that such binding structurally activates SOS, while the autoinhibited conformation occludes the site (Margarit et al. 2003; Sondermann et al. 2004). Because the reaction product (Ras•GTP) would thereby activate its own producer, allosteric activation by Ras•GTP became the canonical, textbook mechanism of positive feedback in Ras signaling, and it is embedded in standard quantitative models of the pathway.

In cells this feedback was assigned a decisive functional role. Disrupting the allosteric pocket blunts growth-factor-induced Ras/ERK activation dynamics (Boykevisch et al. 2006). Ras activation in lymphoid cells is digital — cells are 'on' or 'off' — and exhibits hysteresis, meaning a lower stimulus suffices to maintain than to initiate activation, conferring short-term memory of prior antigen encounter; integrating modeling and experiments, this switch-like behavior was attributed to SOS feedback, with the second lymphocyte GEF, RasGRP (Ras guanine-nucleotide-releasing protein), supplying an analog, graded component (Das et al. 2009). Bimodal SOS-mediated Ras activation also underlies models of the sharp boundary separating positive and negative selection of thymocytes (Prasad et al. 2009).

Single-molecule reconstitution of SOS on supported lipid membranes has now dismantled key premises of this picture. Measuring individual SOS molecules, the expected enhancement of the mean catalytic rate by Ras•GTP at the allosteric site was found to be absent, although Ras•GTP modulated stochastic fluctuations into highly active, long-lived functional states (Iversen et al. 2014). In native full-length SOS, the PR domain obstructs engagement of allosteric Ras (Lee et al. 2017). Full-length SOS undergoes a slow, multi-step autoinhibition-release process on the membrane that delays the onset of catalysis by tens of seconds, enabling kinetic proofreading — a time-delay mechanism for ligand discrimination — gated by phase-separated LAT/EGFR scaffold condensates, dense molecular clusters formed when the scaffold proteins demix into a separate phase (Huang et al. 2019), and SOS autoactivation kinetics have been characterized in solution (Hoang et al. 2021). Two 2024 reconstitution studies then addressed the feedback question directly: in a total reconstitution of the Ras activation–deactivation cycle (SOS opposed by the GTPase-activating protein p120-RasGAP), the bimodal Ras response arises from the pronounced processivity of SOS — one membrane-bound SOS molecule activates hundreds of Ras molecules — without deterministic bistability, meaning without coexistence of two stable steady-state Ras-output levels for the same input, controlled by the condensation state of the LAT scaffold (Lee et al. 2024); and kinetic analysis of receptor-recruited full-length SOS showed that its membrane recruitment is Ras-nucleotide-state independent and that the ensemble-level positive feedback stems from Ras•GTP accelerating the autoinhibition-release step rather than enhancing the catalytic rate (Ren et al. 2024).

These candidate mechanisms are not minor variants of one another: they place feedback at different kinetic steps and make different predictions about whether switch-like Ras output requires bistability at all. In living cells they have never been discriminated. Single-molecule tracking in EGF-stimulated cells found that allosteric-site mutants of SOS reduce a membrane-bound intermediate dwell state and nearly abolish Raf translocation, supporting operation of the SOS–Ras•GTP loop in cells, but with overexpressed tagged SOS and without identifying the kinetic step at which feedback acts (Nakamura et al. 2016). In B cells, endogenous Ras/ERK biosensors showed that digital ERK responses persist even when feedback-mediated Ras activation is impaired, implying additional amplification sources beyond SOS feedback (Mclaurin and Weiner 2019). Which mechanism the SOS pathway actually uses in cells — and how it thereby selects graded versus switch-like Ras responses — remains open.

Problem Statement

Determine, in living cells, the operative molecular mechanism of positive feedback in SOS-mediated Ras activation, and thereby the mechanism by which the SOS pathway selects graded versus switch-like (bimodal) Ras responses to receptor stimulation. The question has three linked parts, none of which settles it alone: (i) which of the four candidate mechanisms carry the feedback in the cellular context, established by direct measurement of SOS state and turnover together with Ras output in the same cells (where hysteresis means that a lower stimulus suffices to maintain than to initiate activation) and by perturbations whose effects on the candidate feedback steps are separable, deciding for each candidate whether it is operative or ruled out rather than identifying a single alternative: occupancy of the allosteric Ras site with consequent catalytic-rate enhancement; Ras•GTP-accelerated release of SOS autoinhibition; stabilization of membrane-bound intermediate states of SOS; and processive Ras turnover gated by scaffold-condensation (LAT/EGFR) organization; a mechanism supported only by in vitro reconstitution or modeling does not establish this part in cells. (ii) the mechanism by which the SOS pathway selects graded versus switch-like output; and (iii) whether switch-like Ras output in cells requires deterministic bistability (coexistence of two stable steady-state output levels for the same input) at all. The answer must be consistent with the established in-cell observations that motivated the question — digital, hysteretic Ras activation in lymphoid cells; persistence of digital ERK responses when feedback-mediated Ras activation is impaired in B cells; and the membrane-intermediate-state dynamics of SOS in growth-factor-stimulated cells.

References

  1. Margarit S.M., Sondermann H., Hall B.E., Nagar B., Hoelz A., Pirruccello M., Bar-Sagi D., Kuriyan J. Structural evidence for feedback activation by Ras·GTP of the Ras-specific nucleotide exchange factor SOS. Cell 112:685–695 (2003). DOI 10.1016/S0092-8674(03)00149-1. https://doi.org/10.1016/S0092-8674(03)00149-1
  2. Sondermann H., Soisson S.M., Boykevisch S., Yang S.-S., Bar-Sagi D., Kuriyan J. Structural analysis of autoinhibition in the Ras activator Son of sevenless. Cell 119:393–405 (2004). DOI 10.1016/j.cell.2004.10.005. https://doi.org/10.1016/j.cell.2004.10.005
  3. Boykevisch S., Zhao C., Sondermann H., Philippidou P., Halegoua S., Kuriyan J., Bar-Sagi D. Regulation of Ras signaling dynamics by Sos-mediated positive feedback. Current Biology 16:2173–2179 (2006). DOI 10.1016/j.cub.2006.09.033. https://doi.org/10.1016/j.cub.2006.09.033
  4. Das J., Ho M., Zikherman J., Govern C., Yang M., Weiss A., Chakraborty A.K., Roose J.P. Digital signaling and hysteresis characterize Ras activation in lymphoid cells. Cell 136:337–351 (2009). DOI 10.1016/j.cell.2008.11.051. https://doi.org/10.1016/j.cell.2008.11.051
  5. Prasad A., Zikherman J., Das J., Roose J.P., Weiss A., Chakraborty A.K. Origin of the sharp boundary that discriminates positive and negative selection of thymocytes. Proceedings of the National Academy of Sciences USA 106:528–533 (2009). DOI 10.1073/pnas.0805981105. https://doi.org/10.1073/pnas.0805981105
  6. Iversen L., Tu H.-L., Lin W.-C., Christensen S.M., Abel S.M., Iwig J., Wu H.-J., Gureasko J., Rhodes C., Petit R.S., Hansen S.D., Thill P., Yu C.-H., Stamou D., Chakraborty A.K., Kuriyan J., Groves J.T. Ras activation by SOS: allosteric regulation by altered fluctuation dynamics. Science 345:50–54 (2014). DOI 10.1126/science.1250373. https://doi.org/10.1126/science.1250373
  7. Lee Y.K., Low-Nam S.T., Chung J.K., Hansen S.D., Lam H.Y.M., Alvarez S., Groves J.T. Mechanism of SOS PR-domain autoinhibition revealed by single-molecule assays on native protein from lysate. Nature Communications 8:15061 (2017). DOI 10.1038/ncomms15061. https://doi.org/10.1038/ncomms15061
  8. Huang W.Y.C., Alvarez S., Kondo Y., Lee Y.K., Chung J.K., Lam H.Y.M., Biswas K.H., Kuriyan J., Groves J.T. A molecular assembly phase transition and kinetic proofreading modulate Ras activation by SOS. Science 363:1098–1103 (2019). DOI 10.1126/science.aau5721. https://doi.org/10.1126/science.aau5721
  9. Nakamura Y., Hibino K., Yanagida T., Sako Y. Switching of the positive feedback for RAS activation by a concerted function of SOS membrane association domains. Biophysics and Physicobiology 13:1–11 (2016). DOI 10.2142/biophysico.13.0_1. https://doi.org/10.2142/biophysico.13.0_1
  10. Hoang H.M., Umutesi H.G., Heo J. Allosteric autoactivation of SOS and its kinetic mechanism. Small GTPases 12:44–59 (2021). DOI 10.1080/21541248.2019.1601954. https://doi.org/10.1080/21541248.2019.1601954
  11. Mclaurin J.D., Weiner O.D. Multiple sources of signal amplification within the B-cell Ras/MAPK pathway. Molecular Biology of the Cell 30:1610–1620 (2019). DOI 10.1091/mbc.e18-09-0560. https://doi.org/10.1091/mbc.e18-09-0560
  12. Lee A.A., Kim N.H., Alvarez S., Ren H., DeGrandchamp J.B., Lew L.J.N., Groves J.T. Bimodality in Ras signaling originates from processivity of the Ras activator SOS without deterministic bistability. Science Advances 10:eadi0707 (2024). DOI 10.1126/sciadv.adi0707. https://doi.org/10.1126/sciadv.adi0707
  13. Ren H., Lee A.A., Lew L.J.N., DeGrandchamp J.B., Groves J.T. Positive feedback in Ras activation by full-length SOS arises from autoinhibition release mechanism. Biophysical Journal 123:3295–3303 (2024). DOI 10.1016/j.bpj.2024.07.014. https://doi.org/10.1016/j.bpj.2024.07.014