Skip to main content

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

bioRxiv logoLink to bioRxiv
[Preprint]. 2026 Oct 6:2026.10.05.756774. [Version 1] doi: 10.64898/2026.10.05.756774

Binding kinetics enable tumor-selective MAPK and growth suppression by a Type 1 RAF Inhibitor

Mathieu Desaunay 1,#, Tara L Peters 2,#, Evangelia Matenoglou 3, Beau Baars 1, Bijaya Gaire 1, Ana Orive-Ramos 1, Li Ren 2, Joseph P Lyssikatos 2, Michael R Burkard 2, Dalton Dacus 2, Matthew J Sale 4, Stuart A Aaronson 5, Frank McCormick 4, Evripidis Gavathiotis 3, Stefan D Gross 2,*, Poulikos I Poulikakos 1,*
PMCID: PMC13692113  PMID: 42926471

Abstract

Therapeutic inhibition of the RAS/MAPK cascade, an oncogenic driver in over one-third of human cancers, has been limited by a core trade-off: effective pathway blockade in tumors versus dose-limiting toxicity in normal tissues. Although Type 1.5 RAF inhibitors successfully mitigated this trade-off in BRAF-monomer–driven tumors, an equivalent strategy has been lacking for RAF-dimer–driven cancers, including those with RAS mutations. Here, we characterize ELV-3111, a next-generation, potent, and selective Type 1 RAF inhibitor that overcomes this limitation via a distinct mechanism of action. ELV-3111 induces paradoxical MAPK hyperactivation in normal tissues by binding RAF and promoting the active conformation (priming) followed by rapid dissociation (fast off-rate), a biochemical feature that spares RAS-mutant tumors, where RAF is already primed. In these and other RAF dimer-driven tumor models, ELV-3111 achieves potent and selective MAPK and growth suppression. When combined with a MEK inhibitor, ELV-3111 enables complementary pharmacology: additive inhibition in tumors and compensatory effects in normal tissues. The combination produced durable regressions across RAS-mutant and BRAF-mutant models, including a RAS-mutant model refractory to current therapies, while maintaining favorable tolerability. Thus, we uncover a generalizable mechanism that enables Type 1 RAF inhibitor plus MEK inhibitor combinations to achieve tumor-selective MAPK suppression, with the potential to inform targeted therapy design.

Introduction

Over a third of human tumors are driven by deregulated RAS/MAPK signaling, most commonly associated with mutations in RAS or BRAF(1–3). RAS/MAPK-targeting therapies, including direct RAS inhibitors and small-molecule inhibitors of RAF, MEK and ERK, have been developed as effective treatments for these cancers. Direct mutant-selective KRAS inhibitors represent a major recent advance and have demonstrated meaningful clinical benefit with generally favorable tolerability. However, adaptive and acquired resistance frequently restore MAPK signaling, resulting in disease progression in most patients. More broadly, despite clinical activity, the efficacy of RAS/MAPK-targeting therapies is often limited by a therapeutic ceiling, whereby toxicities emerge at doses that remain insufficient to achieve complete and durable MAPK pathway suppression in tumors.

Protein kinase inhibitors are structurally classified according to the conformation of the DFG motif and αC-helix that they stabilize, including Type 1 (DFG-IN, αC-IN), Type 1.5 (DFG-IN, αC-OUT), and Type 2 (DFG-OUT, αC-IN) inhibitors(4). In RAF kinases, these structural classes have distinct functional consequences. We previously showed that the position of the αC-helix determines inhibitor selectivity toward RAF monomers versus dimers: αC-OUT Type 1.5 inhibitors selectively inhibit RAF monomers, whereas αC-IN Type 1 and Type 2 inhibitors inhibit both RAF monomers and dimers(5, 6).

Current clinically approved RAF inhibitors belong to the Type 1.5 class and selectively inhibit monomeric BRAF(V600X)(7). In cells with dimeric RAF, including normal cells, these RAF inhibitors paradoxically activate RAF and MAPK signaling via an allosteric mechanism described previously by us(5, 8, 9) and others(10). This opposing pharmacology (MAPK inhibition in tumors and paradoxical MAPK activation in normal tissues) enabled combinations with MEK inhibitors (MEKis)(11–13) or EGFR antagonists(14), enhancing efficacy and tolerability by simultaneously suppressing MAPK signaling in tumors while buffering pathway inhibition in normal tissues. Consequently, combined RAF and MEK inhibition has become the standard of care for BRAF(V600X) tumors.

Despite their clinical success, their inactivity in dimeric RAF contexts restricted the therapeutic utility of Type 1.5 RAF inhibitors to BRAF(V600X) tumors. To broaden therapeutic utility, Type 2 RAF inhibitors, which inhibit both monomeric and dimeric RAF, were developed to target a wider spectrum of RAS/MAPK-driven tumors, including RAS-mutant cancers. However, as single agents, Type 2 RAFis have shown limited clinical activity(15, 16), necessitating combinations with MEK inhibitors. In contrast to Type 1.5 RAF inhibitors, however, both agents suppress MAPK signaling in normal tissues, narrowing the therapeutic index. Consequently, dose reductions required to manage toxicity compromise pathway inhibition in tumors, yielding predominantly stable disease rather than durable tumor regressions(17). Type 1 RAF inhibitors, which also inhibit both monomeric and dimeric RAF, were among the earliest RAF inhibitors developed, but none advanced beyond early clinical development because of insufficient potency and selectivity for therapeutic use.

Here, we show that ELV-3111, a novel, potent and selective Type 1 RAF inhibitor, effectively overcomes key limitations of both Type 1.5 and Type 2 RAF inhibitors. ELV-3111 potently inhibits BRAF class I/II/III-, CRAF-, and RAS-mutant cancer models in vitro and in vivo and overcomes multiple mechanisms of resistance to current RAF inhibitors, including BRAF(V600E) splice variants and RAS co-mutations. Unlike Type 2 RAF inhibitors, ELV-3111 induces robust paradoxical MAPK hyperactivation in normal cells through a distinct mechanism involving RAS-dependent RAF priming upon inhibitor binding followed by rapid inhibitor dissociation. This biochemical and pharmacologic profile forms the basis for a therapeutic strategy combining a Type 1 RAF inhibitor with MEK inhibition that produces robust tumor control in RAS- and BRAF-mutant models, including RAS-mutant tumors refractory to current MAPK-targeting therapies.

Thus, we uncover a previously unrecognized mechanism of paradoxical MAPK activation by Type 1 RAF inhibitors that integrates the therapeutic advantages of both Type 1.5 and Type 2 RAF inhibitors, extending the therapeutic principle underlying RAF/MEK inhibition in BRAF(V600X) tumors to RAF dimer-driven cancers, including RAS-mutant tumors.

Results

ELV-3111 is a Type 1 RAF inhibitor with broad activity and potent inhibition of both monomeric and dimeric RAF.

ELV-3111 (Fig. 1A and fig. S1, A to F) was discovered and developed through optimization of a chemical series of BRAF inhibitors, and is a potent, highly selective BRAF inhibitor based on a scaffold that binds the kinase active site in a Type 1 binding (4). A crystal structure of ELV-3111 and the BRAF kinase domain solved to a 2.85 Å resolution (fig. S1, G and H and Table S1) revealed that ELV-3111 binds BRAF in a DFG-IN/aC-helix-IN (Type 1) binding conformation (Fig. 1B). To understand the selectivity of ELV-3111 versus the broader kinome, the compound was profiled in biochemical assays against a panel of 372 protein kinases at a concentration of 1μM. Only 17 of these kinases were inhibited by more than 50% and were therefore considered potential anti-targets (Fig. 1C and Table S2). The IC50 values of ELV-3111 for each of these 17 kinases were determined, and ELV-3111 was found to be 17 to 844-fold more selective for BRAF over any of the potential off-target kinases (Table S2). Some off-targets identified in the kinase screen, such as ABL1 or c-KIT, are known to cause toxicity in humans, so for a subset of these, we also determined cellular IC50s in relevant functional assays. Results demonstrated that in cells, ELV-3111 is at least 40-fold more selective for BRAF than these anti-targets (Table S2).

Fig. 1. ELV-3111 is a Type 1 RAF inhibitor with broad activity and potent inhibition of both monomeric and dimeric RAF.

Fig. 1.

Fig. 1.

(A) ELV-3111 chemical structure. (B) X-ray co-crystal structure of BRAF kinase domain bound to ELV-3111 (orange) or Vemurafenib (blue) at a 2.85Å resolution. (C) ELV-3111 was screened at 1μM for activity against 371 kinases with 100μM ATP using DMSO as control. (D) NanoLuc tagged A/B/CRAF was expressed in HEK293 cells and treated with K10 (ARAF) or K4 (B/CRAF) tracers and a concentration range of the indicated compounds for 2 hours. BRET ratio was measured and represented as curves (mean values ± SEM, n = 3 independent experiments). (E) Profiling data for ELV-3111 and other RAFis calculated from dose responses ranging from 1nM to 10,000nM. An in-cell western procedure was used to measure pERK inhibition after 2 hours of treatment. Proliferation assays were performed in the indicated cell lines after 3 (A375 and OCI-AML-2) or 5 (SK-MEL-30) days of treatment. Serum shift was calculated from A375 pERK, except for Naporafenib which was calculated from SK-MEL-30. pERK IC50s or GI50s were represented as a table (geometric mean, n = 3 independent experiments). (F) BRAF(V600E) monomeric SK-MEL-239, BRAF(V600E) dimeric resistant SK-MEL-239 C3, NRAS(Q61K) SK-MEL-2 and RAF/RAS(WT) Human Fibroblasts, RPE.1, HPDE and MEFs were treated with increasing concentrations of the indicated RAFis for 6 days. Cell viability was measured and represented as relative viability compared to DMSO-treated controls (mean values ± SEM, n = 3 independent experiments). (G) BRAF(V600E) monomeric SK-MEL-239, BRAF(V600E) dimeric resistant SK-MEL-239 C3, NRAS(Q61K) SK-MEL-2 and RAF/RAS(WT) Human Fibroblasts, RPE.1, HPDE and MEFs were treated with increasing concentrations of different types of RAFis for 1 hour. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling. MEK was used as loading control. (H) HeLa cells engineered to express doxycycline-inducible KRAS(Q61K) were treated with doxycycline (500ng/mL) for 48 hours, followed by treatment with increasing concentrations of different types of RAFis for 1 hour. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling. MEK was used as loading control. Exposure times were adjusted separately for −Dox and +Dox conditions to visualize treatment effects; baseline signaling is shown in fig. S2, E and F. (I) Schematic representation of the in-cell BRET target engagement assay. HEK293 cells were co-transfected with BRAF-NanoLuc and NRAS(Q61K) or KRAS(G12C) constructs. Cells were then co-treated with 62.5nM of LY-Tracer and increasing concentrations of different types of RAFis for 2 hours. After treatment, cells were incubated with 20μM Fluorofurimazine, and BRET signal was measured. (J and L) Data are presented as relative BRET ratios normalized to DMSO-treated controls (mean values ± SEM, n = 3 independent experiments). (K and M) BRET ratio IC50s values were calculated and represented as bar graphs (mean values ± SEM, n = 3 independent experiments, ns = not significant (P>0.05), *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, ordinary one-way ANOVA).

To determine the binding profile and functional activity of ELV-3111 across RAF isoforms and RAF-driven cellular contexts, we employed a NanoBRET cell-based tracer displacement assay(18). ELV-3111 was found to potently bind ARAF, BRAF and CRAF, with similar affinity to the Type 2 RAFi Naporafenib(19) and with greater binding of CRAF compared to either the Type 1 RAFi GDC-0879(20) or the Type 1.5 RAFi Encorafenib(13) (Fig. 1D and fig. S2A). ELV-3111, and these same RAFis were then evaluated for their ability to suppress MAPK signaling in both the BRAF(V600E) A375 and NRAS(Q61K) SK-MEL-30 cell lines, in normal cell culture medium as well as in the presence of 100% human serum to model the attenuating effects of plasma protein binding when dosed in patients. ELV-3111 potently suppressed MAPK signaling in both RAF monomer- and dimer-driven contexts, with superior activity compared to all the other RAFis tested (Fig. 1E). The antiproliferative activities of these RAFis were assessed in the A375, SK-MEL-30 and OCI-AML-2 cell lines, the latter of which harbors a CRAF fusion gene. ELV-3111 was the only RAFi tested that could inhibit growth of all three cell lines with similar potency (Fig. 1E), as Naporafenib has been shown to have reduced potency against monomeric BRAF(V600X)(21).

To define the breadth of activity of ELV-3111 in tumor cells, we compared its MAPK pathway and cell growth-suppressing effects to those of other RAFis with various binding modes. Two Type 1 RAFis (ELV-3111 and GDC-0879), one Type 1.5 RAFi (Encorafenib), and two Type 2 RAFis (Naporafenib and LY3009120) were evaluated in cellular models with either monomeric BRAF(V600E) (parental SK-MEL-239), dimeric BRAF(V600E) (clone C3 SK-MEL-239, expressing the splice variant p61BRAF(V600E)), which is known to drive resistance to the approved Type 1.5 RAFis due to p61BRAF(V600E) homodimerization(22). SK-MEL-2 cells, signaling via dimeric RAF downstream of mutant RAS (RAS(MUT)), and primary Human Fibroblasts, Human Pancreatic Duct Epithelial cell line (HPDE), RPE.1 cells, and Mouse Embryonic Fibroblasts (MEFs) were used as in vitro models of RAS(MUT) and RAS(WT)/BRAF(WT) (representing normal RAS signaling), respectively. Consistent with its binding mode, treatment with ELV-3111 potently inhibited cell growth and MAPK signaling in both monomeric and dimeric BRAF(V600E) cells (Fig. 1, F and G and fig. S2B). As expected, the Type 1.5 RAFi Encorafenib exhibited more potent cell growth inhibition in the monomeric model than in the dimeric model (~720-fold difference). The Type 2 RAFi Naporafenib exhibited greater inhibition in the dimeric model (~150-fold difference), while LY3009120 displayed equipotent activity in both models, consistent with previous reports(5, 21, 23) (Fig. 1F and fig. S2B). Further, ELV-3111 suppressed MAPK signaling and cell growth more potently in RAS(MUT) compared to RAS(WT) cells, with RAS(MUT) selectivity ratio of ~5-fold for growth inhibition and ~23-fold for MAPK pathway inhibition (Fig. 1, F and G and fig. S2, C and D). The potency and selectivity of ELV-3111 exceeded those of the Type 1 RAFi GDC-0879, as well as of the Type 2 RAFis Naporafenib and LY3009120. As expected, the Type 1.5 RAFi Encorafenib was inactive in both RAS(MUT) and RAS(WT) contexts, consistent with its inability to inhibit RAF dimers. These observations correlated well with the MAPK inhibition potency of each RAFi in the cell line models used (Fig. 1G and fig. S2D).

Further, we evaluated ELV-3111 together with representative Type 1, Type 2, or Type 1.5 RAFis in an isogenic model with doxycycline-inducible RAS(MUT) expression. As expected, induction of RAS(MUT) to near-endogenous levels enhanced MAPK signaling in both models (fig. S2, E and F). All Type 1 and Type 2 RAFis demonstrated preferential MAPK pathway suppression in RAS(MUT) over RAS(WT) cells. ELV-3111 showed a modestly higher degree of separation between these contexts compared with GDC-0879, Naporafenib, or LY3009120. (Fig. 1H and fig. S2G). At concentrations within a therapeutic range, RAS(MUT) induction did not significantly enhance MAPK inhibition with Encorafenib. A slight differential effect was only observed at supraphysiological concentrations (~10μM), consistent with the inability of Type 1.5 inhibitors to potently inhibit RAF dimers.

To further substantiate these findings, we compared the activity of ELV-3111 to representative Type 1.5 (Encorafenib) and Type 2 (Naporafenib) RAFis, as well as a MEKi (Binimetinib) in Ba/F3 cells engineered to express distinct classes of BRAF mutations. In this system, ELV-3111 exhibited superior and broader inhibitory activity on cell viability across both monomeric and dimeric RAF-driven signaling contexts (fig. S3, A and B), indicating enhanced functional coverage compared to existing RAFi classes. We further used COLO 201 cells endogenously expressing BRAF(V600E) and engineered them to also express oncogenic RAS(MUT) to model acquired resistance to Type 1.5 RAFis mediated by RAS-driven RAF dimerization(24, 25). While the Type 1.5 RAFi Encorafenib exhibited reduced potency in these RAF dimer-dependent models, ELV-3111 remained highly effective in suppressing growth and MAPK signaling of BRAF(MUT)/RAS(MUT) cells (fig. S3, C to F).

We finally evaluated ELV-3111 across a broader panel of tumor cell lines, including RAS(WT) and various RAS(MUT) and BRAF(MUT) models. ELV-3111 demonstrated high potency in cells driven by Class 1, Class 2, and Class 3 BRAF mutants, with GI50 and IC50 values under 100nM, and was also highly effective against multiple NRAS- and KRAS-mutant cell lines and cell lines harboring dual mutations in both BRAF and NRAS (fig. S4, A to F and Table S3).

Both Type 1 and Type 2 kinase inhibitors bind in the αC-helix-IN mode(4), which is associated with the active dimeric RAF conformation. We thus asked whether the high potency of ELV-3111 in RAS(MUT) tumor cells is the result of strong inhibitor binding to the highly active RAF conformation in RAS(MUT) cells. To this end, we developed a live-cell NanoBRET BRAF engagement assay(18, 26) to monitor target engagement of RAF inhibitors (Fig. 1I). HEK293 cells were co-transfected with BRAF-NanoLuc with or without NRAS(Q61K) and KRAS(G12C) for 24 hours, followed by treatment with increasing concentrations of a fluorescent tracer that we designed and synthesized based on the potent RAFi LY3009120 (LY-Tracer – fig. S5A). The apparent dissociation constant (LY-Tracer KD,app) was first determined to select the LY-Tracer concentration for subsequent competition assays (~50% of RAF-binding saturation) (fig. S5B). Ectopic RAS(MUT) expression resulted in a significant increase in NanoBRET signal by LY-Tracer, at much higher levels compared to non-transfected cells, consistent with active RAS(MUT) promoting the active conformation of RAF that enhances RAFi binding. ELV-3111 was consistently highly potent in RAS(MUT)-expressing cells, demonstrating a ~2-fold greater potency than LY3009120, 4-fold greater than Naporafenib, and 5-fold greater than Belvarafenib - all Type 2 RAFis (Fig. 1, J to M and fig. S5, C to E). Thus, ELV-3111 exhibits high activity in RAS(MUT) cells, through enhanced target engagement and potent inhibition of fully activated RAF.

Taken together, these data demonstrate that ELV-3111 is broadly active across many different classes of MAPK alterations, including settings of known resistance to approved Type 1.5 RAFis, and maintains activity in both the monomer and dimer RAF settings, with a measurable therapeutic window in dimeric RAF-driven, including RAS(MUT) tumors, over normal cells.

ELV-3111 potently suppresses MAPK in BRAF(MUT) and RAS(MUT) tumor models while inducing MAPK hyperactivation in normal tissues as drug concentrations decline in vivo.

To assess MAPK pathway inhibition by ELV-3111 in vivo, we carried a pharmacodynamic/pharmacokinetic study (PK/PD) in mice engrafted with RAS(MUT) (SK-MEL-30 melanoma) xenograft tumors (Fig. 2A). As expected, after a single dose of ELV-3111, initial MAPK inhibition was followed by recovery in both tumor and normal tissues, consistent with relief of negative feedback and adaptive response to MAPK-targeting inhibitors previously described by us and others(27–30). All treatments potently inhibited the pathway with a modest MAPK recovery in RAS(MUT) tumors (Fig. 2B). Type 2 RAFi Naporafenib and the FDA-approved MEK inhibitor (MEKi) Binimetinib, as single treatments, resulted in MAPK pathway inhibition in normal tissues and recovery close to baseline, as expected (Fig. 2C and fig. S6A). Surprisingly, treatment with ELV-3111 resulted in marked MAPK hyperactivation above basal levels in normal tissue (~ 3 and ~2 fold above baseline in lung and kidney, respectively - Fig. 2C and fig. S6A). This hyperactivation emerges as ELV-3111 plasma concentration declined following peak exposure (Fig. 2D), suggesting a relationship between drug concentration and MAPK signaling dynamics. Naporafenib, despite exhibiting a similar PK profile and in vivo half-life (Fig. 2D), did not induce MAPK hyperactivation. Thus, ELV-3111 effectively suppressed MAPK signaling in RAS(MUT) tumors, similar to Type 2 RAFis, but paradoxically hyperactivated MAPK signaling in normal tissues. This phenomenon is reminiscent of that observed with first-generation Type 1.5 RAFis, and indicates an obvious opportunity for substantially expanding the therapeutic index using this RAFi. We next assessed the combination of either ELV-3111 or Naporafenib with Binimetinib in the PK/PD study. The combination of Type 2 RAFi with Binimetinib enhanced MAPK inhibition in both tumor and normal cells, consistent with the narrow therapeutic window of these drugs. ELV-3111 combined with Binimetinib strongly inhibited MAPK in the tumor; however, in normal cells, the opposing effects of the two drugs restored MAPK levels close to normal (Fig. 2, B and C and fig. S6A). These data indicate that paradoxical MAPK hyperactivation by Type 1 RAFi selectively in normal tissue enables expanding the therapeutic index through combinations with MAPK inhibitors.

Fig. 2. ELV-3111 potently suppresses MAPK in BRAF(MUT) and RAS(MUT) tumor models while inducing MAPK hyperactivation in normal tissue, in vivo.

Fig. 2.

Fig. 2.

(A) Schematic representation of the experimental design used to assess in vivo pharmacodynamic/pharmacokinetic effects of ELV-3111 (Type 1 RAFi) or Naporafenib (Type 2 RAFi) in combination or not with Binimetinib (MEKi). SK-MEL-30 tumor xenograft-bearing mice were dosed via oral gavage with a single dose of 3mg/kg Binimetinib, 3mg/kg ELV-3111, 25mg/kg Naporafenib, or combinations of Binimetinib + ELV-3111 or Binimetinib + Naporafenib. (B) Tumor and (C) lung tissues were collected after 2, 4, 6 or 8 hours, lysed, and analyzed by immunoblotting for pERK to assess MAPK pathway activity. Data are represented as the mean relative pERK levels normalized to vehicle-treated controls (n = 5 mice per group). (D) Plasma concentration of free RAFis was measured 2, 4, 6 or 8 hours after dosing. (E) Schematic representation of the experimental design used for the complete drug washout time-course studies. (F) RAS(WT) HEK293 cells were treated with 1μM of either Type 1 RAFis (ELV-3111, SB-590885), Type 2 RAFis (Naporafenib, LY3009120, Belvarafenib) or 10μM of the Type 1 RAFi GDC-0879. Following treatment, cells were washed and incubated in drug-free conditioned medium for the indicated time points. Cell lysates were analyzed by immunoblotting to assess MAPK signaling. MEK was used as loading control. (G) A panel of RAS(WT) cell lines (HEK293, Caco-2, HeLa, Hermes 2B, HPDE, MEFs and Human Fibroblast) was treated with 1μM of either Type 1 RAFi ELV-3111, Type 2 RAFis (Naporafenib or Belvarafenib), or 10μM of the Type 1 RAFi GDC-0879 for 8 hours. Following treatment, cells were washed and incubated in drug-free conditioned medium for 15 or 30 minutes, depending on the cell lines. Cell lysates were analyzed by immunoblotting to assess MAPK signaling. MEK was used as loading control. (H) Concentration of ELV-3111 was measured and adjusted for known plasma protein binding to determine the free drug concentration (nM) at the indicated time points from 3 pharmacokinetic studies in mice (n = 3 animals/group), or the PK/PD study in (A). (I) Schematic representation of the experimental design for the gradual drug washout studies. (J) RAS(WT) Human Fibroblast, (K) RAS(WT) HPDE, (L) NRAS(Q61R) SK-MEL-2, (M) KRAS(Q61K) Calu-6, or (N) NRAS(Q61K) WM3623 cells were treated with 500nM (High – H) of Type 1 RAFis (ELV-3111, GDC-0879, ELV-1974) or Type 2 RAFis (LY3009120, Naporafenib) for 8 hours. Following treatment, cells were washed and incubated with 50nM (Low – L) of Type 1 or Type 2 RAFis for 15 or 30 minutes depending on the cell line. Cell lysates were analyzed by immunoblotting (pMEK, MEK) to assess MAPK signaling. Ratios of phosphorylated MEK to MEK for each sample were calculated and represented as fold change pMEK bar graphs compared to DMSO-controls.

Unlike the paradoxical MAPK hyperactivation induced by Type 1.5 RAFis that is evident both in vitro and in vivo(8–10), MAPK hyperactivation by ELV-3111 appears more robust in vivo. In vitro studies with a single fixed concentration of compound over time do not faithfully replicate the dynamic nature of an in vivo pharmacokinetic (PK) profile, where compound concentrations spike and then decrease with inhibitor clearance (Fig. 2D). To better model the in vivo setting, we first treated cells with different RAFis followed by washout to simulate drug clearance. First, RAS(WT) HEK293 cells, as a model of normal cells, were treated with either Type 1 RAFis (ELV-3111, GDC-0879 or SB-590885), or Type 2 RAFis (including Naporafenib, used in the PK/PD studies) at saturating concentrations for 8 hours, followed by washout (Fig. 2E). Treatment with ELV-3111 or any Type 1 RAFi, followed by washout, resulted in robust MAPK hyperactivation (Fig. 2F), indicating that this is a shared property of Type 1 RAFis. In contrast, treatment with Type 2 RAFis resulted in a significantly lower MAPK rebound, consistent with the Naporafenib profile in vivo. To further extend these observations, we treated an expanded panel of RAS(WT) cells with two Type 1 RAFis (ELV-3111, GDC-0879) or two Type 2 RAFis (Naporafenib, Belvarafenib), followed by washout. Consistently, upon wash-out, each Type 1 RAFi induced pronounced MAPK hyperactivation, whereas MAPK rebound after Type 2 RAFi treatment was minimal (Fig. 2G and fig. S6B).

To simulate more faithfully the gradual reduction of drug concentration observed in vivo, we established two settings of high and low ELV-3111 concentrations - Cmax of 150 nM or 500 nM and Cmin of 4 nM or 50 nM - based on free drug concentrations we observed in pharmacokinetic studies in mice (Fig. 2H). RAS(WT) cells and RAS(MUT) were initially treated at a peak concentration (150 nM or 500 nM) with a panel of Type 1 RAFis , including ELV-3111, GDC-0879 and ELV-1974, a Type 1 RAFi sharing a closely related core scaffold with ELV-3111 (fig. S6C), or with Type 2 RAFis LY3009120 or Naporafenib. Following washout, cells were re-exposed to the corresponding low concentration (4 nM or 50 nM, respectively – Fig. 2I). Under these conditions, Type 1 RAFis induced significantly higher MAPK hyperactivation than Type 2 RAFis, selectively in RAS(WT) cells but not in RAS(MUT) cells (Fig. 2, J to N and fig. S6, D and E). GDC-0879 showed modest hyperactivation in RAS(MUT) cells(10), consistent with its low potency and incomplete suppression of primed RAF at the concentrations used, as shown for αC-IN RAF inhibitors(5). To validate these findings by an orthogonal approach, we tested the Type 1 RAFi ELV-3111 and the Type 2 RAFi Naporafenib in two RAS(WT) models and one RAS(MUT) model, using a TR-FRET assay to measure pERK. Consistent with the preceding results, ELV-3111 induced significantly greater MAPK hyperactivation than Naporafenib, selectively in RAS(WT) cells (fig. S6F).

Together, these results support a model in which MAPK hyperactivation in normal cells following ELV-3111 treatment is a property of Type 1 RAFis that separates them from Type 2 RAFis and is the consequence of declining drug concentrations in vivo.

Type 1 RAFi-induced MAPK hyperactivation is RAS-dependent

To understand the mechanism underlying the Type 1 RAFi-mediated MAPK hyperactivation phenotype in normal cells, we first investigated the potential role of RAS(WT) activity. We have shown previously that RAS activity is essential for MAPK paradoxical activation by Type 1.5 RAFis(5). To test whether this is also the case for Type 1 RAFis, we first used a SHP2 inhibitor (TNO155)(31) and the EGFR inhibitor Gefitinib, both of which suppress RAS(WT) activity. In RAS(WT) cells, high concentrations of either inhibitor in combination with ELV-3111 markedly suppressed MAPK signaling and prevented hyperactivation induced by ELV-3111 after washout (Fig. 3, A and B). We further validated this dependency using HEK293 cells in which all three RAS paralogs have been knocked out via CRISPR (“RAS-less cells”)(32). Consistent with our hypothesis, ELV-3111-induced MAPK hyperactivation was abolished in RAS-less cells after washout and was fully restored upon ectopic expression of RAS(WT) (Fig. 3C).

Fig. 3. Type 1 RAFi-induced MAPK hyperactivation is RAS-dependent.

Fig. 3.

(A) Human Fibroblast cells were treated with 500nM (High – H) of ELV-3111 in combination or not with increasing concentrations of SHP2 inhibitor (TNO155). After treatment, cells were washed and incubated with 50nM of ELV-3111 (Low – L) and the same concentrations of SHP2 inhibitor. Cell lysates were immunoblotted to assess MAPK signaling. MEK and actin were used as loading controls. (B) HPDE cells were treated with 500nM (High – H) of Type 1 ELV-3111 in combination or not with increasing concentrations of EGFR inhibitor (Gefitinib). After treatment, cells were washed and incubated with 50nM of ELV-3111 (Low – L) and the same concentrations of EGFR inhibitor. Cell lysates were subjected to immunoblotting to assess MAPK signaling. MEK and actin were used as loading controls. (C) HEK293, HEK293 RAS-less or HEK293 RAS-less cells ectopically expressing NRAS(WT) were treated with 500 nM (High – H) of Type 1 ELV-3111. Following treatment, cells were washed and incubated with 50 nM (Low – L) of ELV-3111. Cell lysates were immunoblotted with the indicated antibodies to assess MAPK signaling and RAS expression. Actin was used as loading control. (D) Schematic representation of the NanoBit assay used to assess BRAF-KRAS interaction in live-cells. HEK293 cells were co-transfected with LgBit-BRAF and either (E) SmBit-KRAS(WT) or (F) SmBit-KRAS(G12V) constructs for 24 hours. Cells were then treated with 300nM of different Type 1 RAFis (ELV-3111, GDC-0879), Type 2 RAFis (Naporafenib, LY3009120) or Type 1.5 Paradox Breaker (PLX8394). Immediately after treatment, cells were incubated with 20μM of Fluorofurimazine, and luminescence signal was measured over time. Data are presented as luminescence after 60 minutes of treatment/plate reading and represented as bar graphs (luminescence mean values ± SEM, n = 3 independent experiments; ns = not significant (P>0.05), *P<0.05, **P<0.01, ***P<0.001, ordinary one-way ANOVA). (G) COS-7 cells were co-transfected with LgBit-BRAF and SmBit-KRAS(WT) constructs for 24 hours. Cells were then treated with 500nM (High – H) of Type 1 RAFi ELV-3111 or Type 2 RAFi LY3009120 for 2 hours. Following treatment, cells were washed, treated with 50nM (Low – L) of ELV-3111 or LY3009120, and directly incubated with 20μM of Fluorofurimazine. Luminescent signal was measured over time. Data are presented as luminescence after 60 minutes of treatment/plate reading and represented as bar graphs (mean luminescence values ± SEM, n = 3 independent experiments; ns = not significant (P>0.05), ****P<0.0001, ordinary one-way ANOVA). (H) NRAS(Q61R) SK-MEL-2 cells were treated with different concentrations (0.03, 0.3μM) of the indicated RAFis for 1 hour. BRAF was immunoprecipitated and immunoblotted with the indicated antibodies to assess BRAF-CRAF dimerization and BRAF-MEK interaction. Whole cell lysates were immunoblotted with the indicated antibodies to assess MAPK signaling. (I) RAS(WT) Human Fibroblast cells were treated with 500nM (High – H) of Type 1 RAFi ELV-3111 or Type 2 RAFi Naporafenib. Following treatment, cells were washed and treated with 50nM (Low – L) of either ELV-3111 or Naporafenib. BRAF proteins were immunoprecipitated and immunoblotted with the indicated antibodies to assess BRAF-CRAF dimerization. Whole cell lysates were immunoblotted with the indicated antibodies to assess MAPK signaling.

Both Type 1 and Type 2 RAFis are potent RAF dimer inhibitors. However, Type 1 RAFis promoted significantly higher levels of paradoxical MAPK hyperactivation in normal cells compared to Type 2 RAFis. To understand the basis of the difference, we compared their biochemical properties in more detail. First, we assessed formation of the RAS-RAF complex by employing a NanoBiT complementation proximity assay(33). In this system, RAS and BRAF are tagged with the NanoLuc subunits SmBiT and LgBiT, respectively, and expressed in HEK293 cells (fig. S7A). Luminescence is generated upon complex assembly, providing a quantitative readout of protein-protein interactions in live-cells(33) (Fig. 3D). We measured RAS-BRAF interaction after treatment with Type 1 or Type 2 RAFis, as well as a Type 1.5 RAFi that does not induce paradoxical RAF/MAPK activation (Plixorafenib – PLX8394)(34) as a negative control. As expected, and consistent with previous reports(5, 35), both Type 1 and Type 2 RAFis enhanced RAS-BRAF interaction to a similar extent, whereas Plixorafenib did not (Fig. 3, E and F). To mimic drug clearance in vivo, we repeated the experiment by treating with RAFis, followed by drug washout. Treatment with the Type 1 RAFi ELV-3111 or the Type 2 RAFi LY3009120, followed by washout, resulted in a similar induction of the RAS-BRAF interaction in RAS(WT) setting (Fig. 3G). Thus, the induction of the RAS-BRAF interaction is a property shared by Type 1 and Type 2 RAFis and cannot account for their different phenotype in vivo.

To further characterize RAFi-induced signaling complexes, we examined the BRAF-CRAF and BRAF-MEK complex formation following treatment with different RAFis. Both Type 1 and Type 2 RAFis induced RAF transition to its active state, also known as RAFi-induced “priming”(5, 10), evidenced by an increase in S338 phosphorylation of CRAF, BRAF-CRAF dimerization, and BRAF-MEK complex formation at comparable levels in both RAS(MUT) (Fig. 3H and fig. S7, B and C) and RAS(WT) Human Fibroblast cells (Fig. 3I), after washout. ELV-3111 also promoted the BRAF-CRAF complexes at much lower concentrations (Fig. 3H), consistent with its higher RAF binding potency. We also profiled the formation of BRAF or CRAF homodimers or BRAF-CRAF heterodimers using NanoBRET technology and found ELV-3111 to be distinct from other known Type 1 RAFis in its ability to drive CRAF homodimers to a similar extent as Naporafenib (fig. S7D). Thus, induction of the RAF signaling complexes is similar between Type 1 and Type 2 RAFis, it is therefore unlikely to cause their different phenotype in vivo. Finally, to exclude the possibility that Type 1 RAFi induces MAPK hyperactivation through higher RAS-GTP levels resulting from RTK-RAS feedback reactivation, we treated RAS(WT) HPDE cells with low concentration of ELV-3111 (selected to avoid MAPK pathway inhibition and thereby limit relief of ERK-dependent negative feedback) or with an ERK inhibitor as a positive control, either alone or in combination (fig. S7, E and F). As expected, ERK inhibition caused classical feedback relief, resulting in increased RAS activity. In contrast, ELV-3111 washout induced MAPK hyperactivation without increasing RAS activity. Notably, combining ELV-3111 with the ERK inhibitor did not further increase RAS activity compared with ERK inhibition alone. Together, these findings demonstrate that ELV-3111-mediated MAPK hyperactivation is mechanistically distinct from classical RTK–RAS feedback reactivation.

Type 1 RAFis exhibit over an order-of-magnitude faster dissociation from RAF relative to Type 2 RAFis

As both Type 1 and Type 2 RAFis promoted the active RAF conformation and similar RAF signaling complexes, we next hypothesized that differences in inhibitor off-rate kinetics may promote inhibitor dissociation and activated RAF activity following Type 1, but not Type 2, RAFi withdrawal. To test this, we characterized the binding kinetics of two Type 1 RAFis (ELV-3111, GDC-0879) and two Type 2 RAFis (Naporafenib, LY3009120), using an in vitro TR-FRET binding assay based on competition with a fluorescent probe targeting purified activated BRAF (Fig. 4A). Probe occupancy occurred much more rapidly for Type 1 RAFis compared to Type 2 RAFis, indicating a faster dissociation rate (Fig. 4B). Calculation of dissociation rate constant (k2) and residence times (τ) revealed that ELV-3111 and GDC-0879 exhibited markedly faster - by about an order of magnitude - off-rates and thus possess substantially shorter residence times than Type 2 RAFis (fig. S8, A and B).

Fig. 4. Type 1 RAFis exhibit over an order-of-magnitude faster dissociation from RAF relative to Type 2 RAFis.

Fig. 4.

Fig. 4.

(A) Schematic representation of the in vitro TR-FRET binding assay based on competition with a fluorescent probe targeting purified activated BRAF. (B) The binding kinetics of Type 1 RAFis (ELV-3111, GDC-0879) or Type 2 RAFis (Naporafenib, LY3009120) were characterized at two concentrations (10nM and 100nM) against 0.3nM of activated BRAF(WT) in vitro. % of probe occupancy over time was represented as curves (C) Schematic representation of the NanoBRET assay used to assess RAFi residence time in live cells. (D) COS-7 cells were transfected with BRAF-NanoLuc for 24 hours. Cells were then treated for 2 hours with saturating concentrations of the indicated Type 1 (ELV-3111, GDC-0879, ELV-1974) or Type 2 (Naporafenib, LY3009120, Belvarafenib) RAFis. After treatment, cells were washed, treated with 62.5nM of LY-Tracer and incubated with 20μM of Fluorofurimazine. BRET ratio signals were measured over time using a plate reader (mean values ± SEM, n = 3 independent experiments). (E) Ligand RMSD plots showing the fluctuation of LY3009120 (Top - blue) and ELV-3111 (Top - red) over 1 μs MD simulations. (Bottom) Chemical structures of LY3009120 and ELV-3111 with atomic numbering alongside their respective ligand RMSF profiles, highlighting atomic fluctuations within the binding pocket. (F) Overlays of ligand conformations extracted from four time points (200 ns, 400 ns, 600 ns, and 800 ns) from the 1 μs MD trajectory (Left). Two-dimensional ligand interaction diagrams summarizing the key residues mediating contacts with LY3009120 and ELV-3111 (Right). Note: Residue numbering is shifted by one position in the BRAF sequence relative to the crystal structure/MD output. For example, Glu500 in the figure corresponds to Glu501 in the canonical BRAF sequence. (G) HeLa cells engineered to express doxycycline-inducible KRAS(Q61K) were treated with doxycycline (500ng/mL) for 48 hours, followed by treatment with 500nM (High – H) of Type 1 RAFi (ELV-3111) or Type 2 RAFi (Naporafenib) for 8 hours. Following treatment, cells were washed and treated with 50nM (Low – L) of either ELV-3111 or Naporafenib. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling. MEK and actin were used as loading controls. (H) pCRAF levels at baseline and after or without doxycycline induction were quantified from (G) and represented as a bar graph. (I) pERK and (J) pCRAF levels following RAFi treatment were quantified from the immunoblots shown in (G), normalized to DMSO control, and represented as bar graphs.

To validate these findings in live cells, we developed a NanoBRET-based RAFi competition assay(18). NanoLuc-BRAF was ectopically expressed and treated with saturating concentrations of Type 1 (ELV-3111, GDC-0879, and ELV-1974) or Type 2 (LY3009120, Naporafenib, and Belvarafenib) RAFis, followed by washout and addition of LY-Tracer (Fig. 4C). We found that the signal after Type 2 RAFis treatment with and without washout remained very low, indicating prolonged binding of RAFis to RAF after washout. In contrast, the signal after ELV-3111, ELV-1974 or GDC-0879 treatments remained low without washout, but dramatically increased after washout, indicating rapid inhibitor dissociation (Fig. 4D and fig. S8C). Notably, the signal after washing out either Type 1 RAFi examined was even higher than basal (DMSO-treated), consistent with prior RAF engagement by Type 1 RAFi enhancing LY-Tracer binding, due to induction of the RAF active conformation.

To better understand the structural basis of the difference in the rate of dissociation between Type 1 (ELV-3111 or GDC-0879) and Type 2 (LY3009120 or Sorafenib) RAFis, we carried out molecular dynamics (MD) simulations using their respective crystal structures bound to BRAF. Type 2 RAFis exhibited a lower ligand RMSD throughout the simulation compared to Type 1 RAFis, consistent with a more stable binding conformation (Fig. 4E and fig. S8, D and E). Analysis of ligand flexibility showed higher RMSF values for ELV-3111, particularly in solvent-exposed area near the hinge, where the fluorocyclopropane amide moiety adopted dynamic conformations (Fig. 4, E and F). While ELV-3111 and LY3009120 inhibitors formed comparable hydrophobic interactions within the ATP-binding cleft and maintained hydrogen bonds with conserved residues Cys531, Asp593 and Glu500, LY3009120’s urea group enabled persistent bifurcated hydrogen bonds with Glu500 over the course of the simulation (Fig. 4F and fig. S8F). Thus, the urea-mediated interactions and reduced flexibility of LY3009120 likely confer enhanced stabilization of the inhibitor–kinase complex consistent with its slower dissociation, whereas the increased conformational mobility of ELV-3111 is consistent with its faster dissociation kinetics.

To validate the priming requirement for Type 1 RAFi-mediated MAPK hyperactivation in a single background, we used doxycycline-inducible KRAS(Q61K) systems (Fig. 4, G to J; Fig. S9, A to C). KRAS(MUT)-induction raises baseline MAPK signaling and pCRAF, indicating enhanced priming. Without doxycycline (RAS(WT)), Type 1 RAFi ELV-3111 and both Type 2 RAFis LY3009120 and Naporafenib increased pCRAF, but only ELV-3111 produced strong MAPK hyperactivation after washout. With KRAS(Q61K) induced, baseline pCRAF was already elevated; neither inhibitor further increased it, and no hyperactivation occurred. Hyperactivation therefore requires induction of priming in a non-primed background and is attenuated where RAF is already primed by oncogenic RAS. Further, we expressed dimerization-deficient RAF mutants (CRAF(R401H) or BRAF(R509H)) or RAF(WT) in HEK293 and assess MAPK hyperactivation after ELV-3111 treatment and washout. ELV-3111-induced MAPK activation was lower in RAF(MUT) settings compared to wild-type RAF, shown by lower pERK and pMEK. Induction of pCRAF by ELV-3111 was also reduced in the dimerization-deficient mutants, indicating impaired priming (fig. S9D). RAF dimerization is therefore required for inhibitor-induced priming and the subsequent hyperactivation, and disrupting it blocks the effect.

Collectively, these data support a model in which both Type 1 and Type 2 RAFis promote RAF association with RAS, RAF dimerization, and transition into the active conformation. However, the rapid dissociation of Type 1 RAFis from activated RAF generates partially or fully drug-free RAF dimers capable of hyperactivating MAPK signaling in normal cells. In RAS(MUT) cells, RAF is already primed, and Type 1 RAFi suppresses RAF and MAPK signaling (fig. S9E).

ELV-3111 - based MAPK targeting combinations are highly potent in BRAF(MUT) and RAS(MUT) tumors and well tolerated in vivo

MAPK hyperactivation in normal tissues by Type 1 RAFis indicates that MAPK-targeting vertical combinations would enhance antitumor control and mitigate toxicities by restoring basal levels of MAPK signaling in normal tissue. To evaluate the therapeutic potential of combining MAPK inhibitors with ELV-3111 in BRAF(MUT) or RAS(MUT) tumors, we first assessed the combination of ELV-3111 with the MEKi Binimetinib in vitro. The ELV-3111+Binimetinib combination consistently resulted in superior viability inhibition in KRAS(MUT) cell lines compared to treatment with either inhibitor alone (Fig. 5, A and B). To investigate the mechanistic basis underlying the enhanced efficacy of the ELV-3111 + Binimetinib combination compared to either agent alone in vitro, we assessed MAPK signaling dynamics and apoptotic responses at both early (1 hour) and later (8, 24 or 48 hours) time points. While ELV-3111 or Binimetinib alone induced only transient suppression of pERK, followed by partial rebound of MAPK signaling, the combination induced a more pronounced and sustained inhibition of the MAPK pathway, with minimal or no evidence of pathway reactivation. Consistent with these signaling dynamics, prolonged combination treatment resulted in stronger suppression of Cyclin D1, and greater induction of apoptosis as evidenced by the increased levels of cleaved PARP and cleaved caspase-3/7 activity, compared to single-agent treatments (Fig. 5, C and D, and Fig. S10, A to F). Furthermore, ELV-3111 demonstrated synergism when combined with Binimetinib in a dose range around its single-agent GI50 in both BRAF- and NRAS-mutant cell lines (Fig. S10G).

Fig. 5. ELV-3111 – based MAPK targeting combinations are highly potent in BRAF(MUT) and RAS(MUT) tumors and well tolerated in vivo, including tumor insensitive to current MAPK-targeting therapies.

Fig. 5.

Fig. 5.

Fig. 5.

KRAS(G12C) (A) Calu-1 and (B) MIA-PaCa-2 cell lines were treated for 6 days with 500 nM of Type 1 RAFi ELV-3111, 110nM or 12.5nM of Binimetinib (MEKi) or the combination. Cell viability was measured and represented as relative viability compared to DMSO-treated controls (mean values ± SEM, n = 3 independent experiments). KRAS(G12C) (C) Calu-1 or (D) MIA-PaCa-2 were treated with 500nM of Type 1 RAFi ELV-3111, 100nM of Binimetinib (MEKi) or the combination for 8, 24 or 48 hours. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling and cleaved PARP. MEK and actin were used as loading controls. (E) Parental NRAS(Q61R) SK-MEL-2 or Daraxonrasib-resistant NRAS(Q61R) SK-MEL-2 (DAR-R SK-MEL-2) cells were treated for 6 days with 100nM of Type 1 RAFi ELV-3111, 30nM of Binimetinib (MEKi), or the combination. Cell viability was measured and represented as relative viability compared to DMSO-treated controls (mean values ± SEM, n = 3 independent experiments). (F) Parental NRAS(Q61R) SK-MEL-2 or Daraxonrasib-resistant NRAS(Q61R) SK-MEL-2 (DAR-R SK-MEL-2) cells were treated for 24 hours with increasing concentrations of Binimetinib (MEKi) in combination or not with 20nM of Type 1 RAFi ELV-3111. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling. MEK and actin were used as loading controls. (G) Parental KRAS(Q61K) Calu-6 or Daraxonrasib-resistant KRAS(Q61K) Calu-6 (DAR-R Calu-6) cells were treated for 6 days with 100nM of Type 1 RAFi ELV-3111, 30nM of Binimetinib (MEKi), or the combination. Cell viability was measured and represented as relative viability compared to DMSO-treated controls (mean values ± SEM, n = 3 independent experiments). (H) Parental KRAS(Q61K) Calu-6 or Daraxonrasib-resistant KRAS(Q61K) Calu-6 (DAR-R Calu-6) cells were treated for 24 hours with increasing concentrations of Binimetinib (MEKi) in combination or not with 100nM of Type 1 RAFi ELV-3111. Cell lysates were analyzed by immunoblotting with the indicated antibodies to assess MAPK signaling. MEK and actin were used as loading controls. (I) RAS(WT) Human Fibroblast cells were treated for 8 hours with 500nM (High – H) of ELV-3111 in combination or not with increasing concentrations of the MEKi Binimetinib. After treatment, cells were washed and incubated with 50nM (Low – L) of ELV-3111 and same concentrations of MEKi. MAPK signaling was assessed by immunoblotting. MEK and actin were used as loading controls. (J) RAS(WT) HPDE cells were treated with increasing concentrations of Type 1 RAFi ELV-3111 in combination or not with increasing concentrations of Binimetinib (MEKi) for 8 hours. Following treatment, cells were washed and incubated in drug-free conditioned medium for 15 minutes. Cell lysates were analyzed by pERK immunoblotting to assess MAPK signaling. MEK served as a loading control. pERK was quantified, normalized to DMSO control, and represented as a concentration-response heatmap. (K) HaCaT (RAF/RAS(WT)) or SK-MEL-30 (NRAS(MUT)) cells were treated with the indicated concentrations of ELV-3111 or Naporafenib with or without Binimetinib for 2 hours, fixed and pERK was measured by in-cell western. pERK percent was normalized to DMSO treated controls (n = 3 independent experiments). Mice with established COLO-205 (BRAF(V600E)) or SK-MEL-30 (NRAS(Q61K)) tumors (~150mm3), were treated p.o. with vehicle, 3 mg/kg Binimetinib BID, 6 mg/kg (COLO-205) or 10 mg/kg (SK-MEL-30) ELV-3111 QD or the combination (n = 8 mice per group). Individual tumor volume changes to baseline at the end of the study are represented as waterfall plot for (L) COLO-205 and (M) SK-MEL-30. Body weight of mice was followed over time and normalized to baseline for (N) COLO-205 and (O) SK-MEL-30 xenograft-bearing mice. Mice with established NCI-H2122 (KRAS(G12C)) tumors (~250mm3), were treated p.o. with vehicle, 3 mg/kg Binimetinib BID, 25 mg/kg Naporafenib BID, 25 mg/kg ELV-3111 QD, or Naporafenib + Binimetinib or ELV-3111 + Binimetinib combinations (n = 8 mice per group). (P) Individual tumor volume changes to baseline at the end of the study are represented as waterfall plot and (Q) tumor growth changes for both combinations over time were followed and normalized to the baseline. Mice with established NCI-H2122 (KRAS(G12C)) tumors (~500mm3), were treated p.o. for 72 hours with vehicle, 3 mg/kg Binimetinib BID, 25 mg/kg Naporafenib BID, 25 mg/kg ELV-3111 QD, or Naporafenib + Binimetinib or ELV-3111 + Binimetinib combinations (n = 2 mice per group). Tumor and skin tissues were collected, fixed, and stained for pERK and Ki-67. (R and T) Representative IHC staining of tumor and skin sections for pERK and Ki-67. Whole slide images were analyzed for pERK and Ki-67. Five non-overlapping regions of equal area spanning the entire tumor or skin section were quantified to ensure unbiased sampling of the tissue. (S and U) Percentage of positive cells for each staining was represented as bar graphs (mean values ± SEM, n = 10; ns = not significant (P>0.05), **P<0.01, ****P<0.0001, ordinary one-way ANOVA). (V) Therapeutic hypothesis. As a single agent, ELV-3111 demonstrates high potency and selectivity for RAF(MUT) or RAS(MUT) tumors, while induces MAPK hyperactivation in normal cells. This unique pharmacological phenotype creates an opportunity for tumor-selective combination strategies. Combining ELV-3111 with MEK, RAS or SHP2 inhibitors can suppress MAPK hyperactivation in normal tissues, thereby enhancing tumor selectivity and improving the therapeutic index (TI) for RAF/RAS(MUT)-targeted therapies.

The recent advent of direct RAS inhibitors represents a major advance in the treatment of RAS(MUT) cancers, producing meaningful clinical responses across several tumor types(36, 37). As with other targeted agents, however, their benefit is frequently limited by acquired resistance and eventual disease progression(38). We therefore evaluated the efficacy of the ELV-3111 + Binimetinib combination in NRAS- and KRAS-mutant models with acquired resistance to the multi-RAS(ON) inhibitor Daraxonrasib(39, 40). We derived Daraxonrasib-resistant (DAR-R) NRAS(MUT) and KRAS(MUT) cells, which exhibited increased RAS expression and enhanced downstream MAPK signaling. Similar mechanisms of resistance have been reported for mutant-selective RAS inhibitors(41–43). These changes conferred resistance to both Daraxonrasib-mediated growth inhibition and MAPK pathway suppression (fig. S10, H to L). In contrast, the ELV-3111 + Binimetinib combination suppressed cell growth and MAPK signaling comparably in parental and DAR-R cells (Fig. 5, E to H). These results indicate that this combination can be an effective strategy for both treatment-naïve RAS(MUT) and those that have developed resistance to RAS inhibitors, a scenario of increased clinical relevance.

To model the effect of the ELV-3111+MEKi combination on normal cells, we examined MAPK signaling dynamics following treatment and washout in RAS(WT) human fibroblast cells. ELV-3111+MEKi attenuated MAPK hyperactivation after washout, restoring signaling to basal levels (Fig. 5I). To define the robustness of MAPK hyperactivation and its buffering using MEK inhibition in normal cells, we performed a checkerboard-style concentration matrix of ELV-3111 and Binimetinib in RAS/RAF(WT) HPDE cells following washout and quantified MAPK dynamics. These data demonstrate that ELV-3111 induces MAPK hyperactivation across a broad concentration range, indicating that this effect is not limited to a narrow exposure window. Importantly, Binimetinib consistently buffers this hyperactivation, maintaining MAPK signaling within an intermediate, physiologically relevant range across clinically relevant concentrations (Fig. 5J). Together, these findings define a broad therapeutic buffering window that is robust to variations in drug exposure, supporting the translational feasibility of this combination strategy.

On-target toxicity from MAPK inhibitors in the clinic commonly appears in the skin. We compared the combination of either ELV-3111 or Naporafenib with Binimetinib in a normal human keratinocyte immortalized cell line, HaCaT, and compared it to the NRAS(MUT) melanoma cell line, SK-MEL-30. Binimetinib alone potently suppressed the MAPK pathway in both normal and RAS(MUT) cells (Fig. 5K). Naporafenib alone potently inhibited MAPK in the NRAS(MUT) context and modestly suppressed the pathway in the normal cells. However, combined Naporafenib+Binimetinib markedly inhibited MAPK in both contexts, consistent with the dose-limiting toxicities of the Naporafenib+MEKi or other Type 2 RAFi+MEKi combinations observed in the clinic(17, 44). ELV-3111 alone inhibited MAPK in the NRAS(MUT) cells but showed MAPK activation in the normal cells. When combined with Binimetinib, ELV-3111 strongly inhibited MAPK in the NRAS(MUT) cells but partially buffered the MAPK suppression in the normal cells (Fig. 5K). Thus, the modest MAPK hyperactivation by ELV-3111 without washout in vitro, suggests a therapeutic window of the ELV-3111+MEKi compared to Naporafenib+MEKi combination, which can be further expanded with the selective robust MAPK hyperactivation in normal cells in vivo.

To assess the effectiveness and tolerability of combinatorial strategies in vivo, we tested ELV-3111–based combinations in BRAF(MUT) and RAS(MUT) tumor models. We examined the effect of ELV-3111, alone or in combination with Binimetinib, in mice bearing BRAF(V600E) COLO-205 or NRAS(MUT) SK-MEL-30 xenograft tumors. At a low dose where the free drug concentrations target the in vitro pERK IC50 in SK-MEL-30 for only 6 hours (0.75mg/kg) (Fig. 2H), ELV-3111 alone modestly inhibited tumor growth, while its combination with Binimetinib achieved complete tumor stasis, with modest regressions observed in individual animals, and maintaining good tolerability (fig. S11, A to C). ELV-3111 at higher dose, well above the in vitro pERK IC50 in SK-MEL-30, or Binimetinib alone promoted either growth inhibition or tumor stasis, respectively. High dose of ELV-3111 together with Binimetinib promoted deep tumor regressions in both BRAF(MUT) and RAS(MUT) mutant contexts, with 75% and 100% of overall response rate (ORR), respectively (Fig. 5, L and M and fig. S11, D and E). High doses of ELV-3111 alone were associated with body weight loss, likely a result of MAPK hyperactivation in normal tissues, as it was reversed with the addition of Binimetinib (Fig. 5, N and O). To corroborate this tolerability effect with another Type 1 RAFi, we used ELV-1974 (fig. S6C). ELV-1974 treatment also induced body weight loss at high concentrations. Similar to ELV-3111 + Binimetinib, the ELV-1974 + Binimetinib combination resulted in deep tumor regressions and completely restored normal body weight (Fig. S11, F and G). These results show that Type 1 RAFi-mediated MAPK hyperactivation in normal tissues contributes to reduced tolerability in vivo, but can be exploited to design safe and effective combinatorial strategies for BRAF/RAS(MUT) tumors.

ELV-3111, in combination with a MEK inhibitor, induces deep objective responses in a RAS(MUT) tumor model insensitive to current MAPK-targeting therapies.

Exclusive reliance on highly responsive preclinical models can substantially overstate the efficacy of RAS/MAPK pathway inhibition, as most RAS(MUT) tumors in both experimental models and clinical settings exhibit, at best, cytostatic responses. This persistent “cytostatic ceiling” reflects a fundamental limitation of current therapeutic strategies, including combined Type 2 RAF and MEK inhibitors, and has posed a major barrier to translating pathway blockade into durable clinical benefit. To determine whether Type 1 RAFi can circumvent this intrinsic resistance and elicit more profound antitumor effects, we employed the NCI-H2122 KRAS(MUT) lung cancer xenograft model, which has consistently demonstrated only growth arrest in response to Type 2 RAFi + MEKi therapy(15, 45).

Consistent with previous reports, Naporafenib alone or Naporafenib + Binimetinib combination resulted in progressive disease, or stasis at best (Fig. 5, P and Q). ELV-3111 monotherapy likewise resulted in progressive disease, underscoring the necessity of combination treatment. Strikingly, the combination of ELV-3111 with Binimetinib demonstrated markedly superior potency compared to Naporafenib + Binimetinib, reaching an 87.5% Objective Response Rate (ORR), while the Naporafenib + Binimetinib combination elicited no objective responses (Fig. 5, P and Q). Although high-dose ELV-3111 monotherapy exhibited reduced tolerability, necessitating dosing holidays and resulting in treatment-related deaths, co-treatment with Binimetinib was well tolerated, with no observed toxicity or excessive body-weight loss. In contrast, the Naporafenib-Binimetinib combination exhibited reduced tolerability, causing rash and necessitating dosing holidays and resulting in animal deaths (4 in the Naporafenib+Binimetinib combinations versus none in the ELV-3111 + Binimetinib treatments) (fig. S12, A to C). Reducing the dose of Binimetinib (1.5 mg/kg) or Naporafenib (12.5 mg/kg) resulted in decreased toxicity, as expected; however, it also compromised the antitumor efficacy of the Naporafenib + Binimetinib combination (Fig. S12, D and E).

In a paired tumor-and-normal-skin pharmacodynamic (PD) analysis performed in the NCI-H2122 xenograft model under the same dosing conditions as the efficacy study, we assessed pERK and proliferative output by immunohistochemistry (IHC) and downstream MAPK signaling by immunoblotting in tumor xenografts and normal skin (Fig. 5, R to U and fig. S12F). Across every tumor readout examined, ELV-3111 + Binimetinib matched or exceeded Naporafenib + Binimetinib, while in the same animals it preserved normal-tissue pERK and Ki-67 near baseline, whereas Naporafenib + Binimetinib suppressed them below baseline, and ELV-3111 as a single agent induced MAPK hyperactivation. These PD data link the proposed mechanism, deep tumor suppression with normal-tissue sparing, directly to the drug combination used in the efficacy studies.

Together, these results demonstrate that the combination of ELV-3111 with MEKi induces deep tumor-MAPK-signaling and proliferative output inhibition, while restoring normal MAPK signaling in normal tissues, significantly improving tolerability, which results in markedly increased antitumor efficacy in RAF/ RAS(MUT) tumors (Fig. 5V).

Discussion

Here we uncover a mechanism of MAPK hyperactivation by Type 1 RAFis, selectively in normal cells. Both Type 1 and Type 2 RAFis bind RAF in the αC-helix-IN conformation and thereby drive a shift in the equilibrium towards the active RAF conformation (priming). However, we found that the position of DFG (IN for Type 1 and OUT for Type 2) is associated with markedly different inhibitor dissociation rates from RAF. Thus, drug clearance in vivo results in rapid inhibitor dissociation from RAF primed by Type 1 RAF inhibitors, but not Type 2 RAF inhibitors, generating drug-free RAF dimers capable of paradoxically activating MAPK signaling in normal tissues. In contrast, the prolonged residence time of Type 2 RAF inhibitors maintains inhibition of primed RAF despite declining drug concentrations. Importantly, MAPK hyperactivation does not occur in RAS(MUT) tumors because RAF is constitutively primed by oncogenic RAS, rendering it unresponsive to further priming by Type 1 RAF inhibitors. Collectively, these findings identify inhibitor dissociation kinetics, rather than equilibrium target engagement alone, as a key determinant of the biological consequences of RAF inhibition.

A central challenge in targeting the RAS/MAPK pathway has been achieving complete suppression of oncogenic MAPK signaling in tumors while preserving physiological MAPK signaling in normal tissues. This therapeutic ceiling has limited the efficacy of most MAPK-targeted therapies. Current RAF and MEK inhibitor combinations for BRAF(V600X) tumors represent a notable exception. Their success relies on complementary pharmacology, whereby paradoxical MAPK activation induced by Type 1.5 RAF inhibitors is buffered by MEK inhibition in normal tissues while maintaining profound pathway suppression in tumors. The therapeutic opportunity therefore does not derive from paradoxical MAPK activation itself, which represents a dose-limiting liability, but from the ability to buffer this otherwise deleterious effect through vertical MEK inhibition, thereby enabling greater dose intensity and deeper tumor MAPK suppression. Our findings demonstrate that this therapeutic principle is not restricted to Type 1.5 RAF inhibitors or BRAF(V600X) tumors. Instead, the Type 1 RAF inhibitor + MEK inhibitor combination recreates this complementary pharmacology while retaining the ability to inhibit RAF dimers, thereby extending this therapeutic strategy to RAF dimer-driven tumors, including those driven by oncogenic RAS.

Type 1 RAFis were among the first RAF-targeting agents developed, but none advanced to clinical development. This may have reflected insufficient potency or selectivity, as is clearly the case for GDC-0879, rather than an inherent limitation of the Type 1 binding mode itself. ELV-3111 distinguishes itself by exhibiting markedly greater potency and selectivity than earlier Type 1 RAF inhibitors, allowing the unique pharmacology of this class to be therapeutically exploited by combining with a MEK inhibitor. We provide strong preclinical evidence that combining ELV-3111 with a MEK inhibitor induces profound and durable tumor regressions across multiple RAF- and RAS(MUT) models, including a RAS(MUT) model refractory to current MAPK-targeted therapies, while maintaining a favorable tolerability profile. The improved therapeutic index results from complementary pharmacologic effects: enhanced MAPK suppression in tumors together with restoration of physiological MAPK signaling in normal tissues.

Mechanistically, our findings indicate that paradoxical MAPK activation by Type 1 RAF inhibitors is distinct from previously described mechanisms. Multiple alternative explanations, including differential induction of RAF-RAS complexes, RAF dimerization, RAF priming, or ERK-dependent feedback reactivation, were excluded experimentally. Instead, both Type 1 and Type 2 RAF inhibitors similarly promote RAF association with RAS, RAF dimerization, and transition to the primed state, whereas the rapid dissociation of Type 1 RAF inhibitors uniquely permits activation of these primed RAF complexes as drug concentrations decline. This also explains why paradoxical MAPK activation by Type 1 RAF inhibitors is primarily evident in vivo, but less apparent under conventional in vitro conditions employing constant drug concentrations. Based on its mechanism of action, ELV-3111 provides a versatile backbone for MAPK-targeted combination therapies. Beyond MEK inhibition, combinations with upstream pathway inhibitors, including SHP2(31), SOS1(46), or RAS(47) inhibitors - may be both well tolerated and highly effective, depending on tumor context.

The recent clinical success of direct RAS inhibitors has transformed the treatment of several RAS-mutant cancers. Nevertheless, many RAS-driven tumors remain without effective direct inhibitors, and even in tumors with approved RAS inhibitors, adaptive and acquired resistance commonly restore MAPK signaling through mechanisms involving both mutant and wild-type RAS. In this context, Type 1 RAF + MEK inhibition may provide a complementary therapeutic strategy by targeting RAF downstream of both mutant and wild-type RAS while retaining the therapeutic advantages conferred by selective modulation of MAPK signaling in normal tissues. Consistent with this concept, the ELV-3111 + MEK inhibitor combination remained effective in cellular models with acquired resistance to a multiselective RAS(ON) inhibitor. Although these observations are currently limited to in vitro models and require validation in vivo and ultimately in clinical studies, they suggest that this strategy may have particular utility in the increasingly relevant post-RAS inhibitor setting, while also providing a therapeutic option for RAS-driven tumors that currently lack effective direct RAS-targeted therapies.

The unique property we discovered may not be exclusive to Type 1 RAFis. Other kinase inhibitors, including those targeting AKT(48) or S6K1(49), have also been shown to induce target priming. It is conceivable that fast off-rate inhibitors targeting these kinases may similarly promote transient target hyperactivation during drug clearance, creating opportunities for tumor-selective vertical pathway inhibition. More broadly, our findings suggest that signaling dynamics and inhibitor binding kinetics, in addition to target selectivity, can be leveraged to improve the therapeutic index of targeted therapies. Thus, therapeutically exploiting target priming together with rapid inhibitor dissociation through appropriately designed vertical pathway combinations may represent a broadly applicable strategy for overcoming the therapeutic ceiling that has limited therapies targeting oncogenic signaling pathways.

Our study demonstrates that combining a Type 1 RAF inhibitor with MEK inhibition integrates the complementary therapeutic properties of current Type 1.5 and Type 2 RAF inhibitor strategies while broadening the scope of RAF-targeted therapy beyond BRAF(V600X) cancers. By coupling potent inhibition of RAF dimers in tumors with paradoxical MAPK activation selectively in normal tissues, this combination strategy overcomes the therapeutic ceiling imposed by physiological MAPK pathway inhibition, enabling deeper suppression of oncogenic MAPK signaling without a corresponding increase in normal tissue toxicity. Beyond treatment-naïve RAS-driven cancers, this strategy has the potential to expand therapeutic options for tumors that currently lack effective direct RAS inhibitors and for patients whose tumors have progressed following RAS inhibitor therapy, an increasingly common clinical challenge. Together, these findings provide a strong rationale for evaluating Type 1 RAF inhibitor/MEK inhibitor combinations as a broadly applicable therapeutic strategy across RAS/MAPK-driven cancers.

Supplementary Material

Supplement 1

Acknowledgments

We acknowledge the Cancer Center support grant: NIH/NCI 1P30CA196521-0. We acknowledge Proteros Biosciences for generating the co-crystal of ELV-3111 with BRAF, and Pharmaron for their essential research services in support of this project.

Funding

P.I.P. is supported by the US National Institutes of Health (NIH) (R01 CA285713, R01 CA240362, R01CA238229), a V Foundation Translational Award, the Melanoma Research Alliance, the Breast Cancer Alliance, the Melanoma Research Foundation, the Irma T. Hirschl Trust and Tisch Cancer Institute developmental awards. M.D is supported by Melanoma Research Foundation (Career Development Award). E.M and E.G. were supported by a National Institutes of Health grant R01CA223243 (E.G), a Montefiore Einstein Comprehensive Cancer Center Support grant P30CA013330, and the Albert Einstein College of Medicine Ph.D. Student Stipend Fund to E.M.

Competing interests

P.I.P reports research funding to his institution from Verastem Oncology and Enliven Therapeutics. P.I.P reports consulting fees from Nuvalent Inc, Blueprint Medicines and Belharra Therapeutics. E.G. reports compensation for consulting or equity ownership from BaxGen Therapeutics, BeanPod Biosciences, Comorin Therapeutics, Life Biosciences, Stelexis Biosciences for projects unrelated to this work. T.L.P., L.R., J.R.L., M.R.B., D.D., S.D.G. are employees and shareholders of Enliven Therapeutics. All other authors declare no competing interests.

Data and Materials availability:

All data needed to evaluate the conclusions in the paper are present in the paper or the supplementary materials. The atomic coordinates and structure factors of kinase BRAF (chain A and chain B) bound to ELV-3111 have been deposited in the Protein Data Bank and can be accessed using the accession number 9ZBO.

References

  • 1.Young A. et al. , Ras signaling and therapies. Advances in cancer research 102, 1–17 (2009). [DOI] [PubMed] [Google Scholar]
  • 2.Moore A. R., Rosenberg S. C., McCormick F., Malek S., RAS-targeted therapies. Nature reviews. Drug discovery, (2021). [DOI] [PubMed] [Google Scholar]
  • 3.Samatar A. A., Poulikakos P. I., Targeting RAS-ERK signalling in cancer: promises and challenges. Nature reviews. Drug discovery 13, 928–942 (2014). [DOI] [PubMed] [Google Scholar]
  • 4.Zhang J., Yang P. L., Gray N. S., Targeting cancer with small molecule kinase inhibitors. Nature reviews. Cancer 9, 28–39 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Karoulia Z. et al. , An Integrated Model of RAF Inhibitor Action Predicts Inhibitor Activity against Oncogenic BRAF Signaling. Cancer cell 30, 485–498 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Karoulia Z., Gavathiotis E., Poulikakos P. I., New perspectives for targeting RAF kinase in human cancer. Nature reviews. Cancer 17, 676–691 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Tsai J. et al. , Discovery of a selective inhibitor of oncogenic B-Raf kinase with potent antimelanoma activity. Proceedings of the National Academy of Sciences of the United States of America 105, 3041–3046 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Poulikakos P. I., Zhang C., Bollag G., Shokat K. M., Rosen N., RAF inhibitors transactivate RAF dimers and ERK signalling in cells with wild-type BRAF. Nature 464, 427–430 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Joseph E. W. et al. , The RAF inhibitor PLX4032 inhibits ERK signaling and tumor cell proliferation in a V600E BRAF-selective manner. Proceedings of the National Academy of Sciences of the United States of America 107, 14903–14908 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hatzivassiliou G. et al. , RAF inhibitors prime wild-type RAF to activate the MAPK pathway and enhance growth. Nature 464, 431–435 (2010). [DOI] [PubMed] [Google Scholar]
  • 11.Flaherty K. T. et al. , Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. The New England journal of medicine 367, 1694–1703 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Robert C. et al. , Improved overall survival in melanoma with combined dabrafenib and trametinib. The New England journal of medicine 372, 30–39 (2015). [DOI] [PubMed] [Google Scholar]
  • 13.Dummer R. et al. , Encorafenib plus binimetinib versus vemurafenib or encorafenib in patients with BRAF-mutant melanoma (COLUMBUS): a multicentre, open-label, randomised phase 3 trial. The lancet oncology 19, 603–615 (2018). [DOI] [PubMed] [Google Scholar]
  • 14.Kopetz S. et al. , Encorafenib, Binimetinib, and Cetuximab in BRAF V600E-Mutated Colorectal Cancer. The New England journal of medicine 381, 1632–1643 (2019). [DOI] [PubMed] [Google Scholar]
  • 15.Yen I. et al. , ARAF mutations confer resistance to the RAF inhibitor belvarafenib in melanoma. Nature 594, 418–423 (2021). [DOI] [PubMed] [Google Scholar]
  • 16.Sullivan R. J. et al. , A Phase I Study of LY3009120, a Pan-RAF Inhibitor, in Patients with Advanced or Metastatic Cancer. Molecular cancer therapeutics 19, 460–467 (2020). [DOI] [PubMed] [Google Scholar]
  • 17.de Braud F. et al. , Initial Evidence for the Efficacy of Naporafenib in Combination With Trametinib in NRAS-Mutant Melanoma: Results From the Expansion Arm of a Phase Ib, Open-Label Study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 41, 2651–2660 (2023). [DOI] [PubMed] [Google Scholar]
  • 18.Robers M. B. et al. , Target engagement and drug residence time can be observed in living cells with BRET. Nature communications 6, 10091 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Monaco K. A. et al. , LXH254, a Potent and Selective ARAF-Sparing Inhibitor of BRAF and CRAF for the Treatment of MAPK-Driven Tumors. Clinical cancer research : an official journal of the American Association for Cancer Research 27, 2061–2073 (2021). [DOI] [PubMed] [Google Scholar]
  • 20.Hoeflich K. P. et al. , Antitumor efficacy of the novel RAF inhibitor GDC-0879 is predicted by BRAFV600E mutational status and sustained extracellular signal-regulated kinase/mitogen-activated protein kinase pathway suppression. Cancer research 69, 3042–3051 (2009). [DOI] [PubMed] [Google Scholar]
  • 21.Adamopoulos C. et al. , Exploiting Allosteric Properties of RAF and MEK Inhibitors to Target Therapy-Resistant Tumors Driven by Oncogenic BRAF Signaling. Cancer discovery 11, 1716–1735 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Poulikakos P. I. et al. , RAF inhibitor resistance is mediated by dimerization of aberrantly spliced BRAF(V600E). Nature 480, 387–390 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Peng S. B. et al. , Inhibition of RAF Isoforms and Active Dimers by LY3009120 Leads to Anti-tumor Activities in RAS or BRAF Mutant Cancers. Cancer cell 28, 384–398 (2015). [DOI] [PubMed] [Google Scholar]
  • 24.Kopetz S. et al. , Molecular profiling of BRAF-V600E-mutant metastatic colorectal cancer in the phase 3 BEACON CRC trial. Nature medicine 30, 3261–3271 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Nazarian R. et al. , Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature 468, 973–977 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Vasta J. D. et al. , Quantitative, Wide-Spectrum Kinase Profiling in Live Cells for Assessing the Effect of Cellular ATP on Target Engagement. Cell Chem Biol 25, 206–214 e211 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Ahmed T. A. et al. , SHP2 Drives Adaptive Resistance to ERK Signaling Inhibition in Molecularly Defined Subsets of ERK-Dependent Tumors. Cell Rep 26, 65–78 e65 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Lito P. et al. , Relief of profound feedback inhibition of mitogenic signaling by RAF inhibitors attenuates their activity in BRAFV600E melanomas. Cancer cell 22, 668–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Duncan J. S. et al. , Dynamic reprogramming of the kinome in response to targeted MEK inhibition in triple-negative breast cancer. Cell 149, 307–321 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Corcoran R. B. et al. , EGFR-mediated re-activation of MAPK signaling contributes to insensitivity of BRAF mutant colorectal cancers to RAF inhibition with vemurafenib. Cancer discovery 2, 227–235 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Liu C. et al. , Combinations with Allosteric SHP2 Inhibitor TNO155 to Block Receptor Tyrosine Kinase Signaling. Clinical cancer research : an official journal of the American Association for Cancer Research 27, 342–354 (2021). [DOI] [PubMed] [Google Scholar]
  • 32.Cuevas-Navarro A. et al. , RAS-dependent RAF-MAPK hyperactivation by pathogenic RIT1 is a therapeutic target in Noonan syndrome-associated cardiac hypertrophy. Sci Adv 9, eadf4766 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Dixon A. S. et al. , NanoLuc Complementation Reporter Optimized for Accurate Measurement of Protein Interactions in Cells. ACS Chem Biol 11, 400–408 (2016). [DOI] [PubMed] [Google Scholar]
  • 34.Zhang C. et al. , RAF inhibitors that evade paradoxical MAPK pathway activation. Nature 526, 583–586 (2015). [DOI] [PubMed] [Google Scholar]
  • 35.Jin T. et al. , RAF inhibitors promote RAS-RAF interaction by allosterically disrupting RAF autoinhibition. Nature communications 8, 1211 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hong D. S. et al. , KRAS(G12C) Inhibition with Sotorasib in Advanced Solid Tumors. The New England journal of medicine 383, 1207–1217 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Janne P. A. et al. , Adagrasib in Non-Small-Cell Lung Cancer Harboring a KRAS(G12C) Mutation. The New England journal of medicine 387, 120–131 (2022). [DOI] [PubMed] [Google Scholar]
  • 38.Ebright R. Y., Dilly J., Shaw A. T., Aguirre A. J., Response and Resistance to RAS Inhibition in Cancer. Cancer discovery 15, 1325–1349 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Holderfield M. et al. , Concurrent inhibition of oncogenic and wild-type RAS-GTP for cancer therapy. Nature, (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Cregg J. et al. , Discovery of Daraxonrasib (RMC-6236), a Potent and Orally Bioavailable RAS(ON) Multi-selective, Noncovalent Tri-complex Inhibitor for the Treatment of Patients with Multiple RAS-Addicted Cancers. Journal of medicinal chemistry 68, 6064–6083 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Awad M. M. et al. , Acquired Resistance to KRAS(G12C) Inhibition in Cancer. The New England journal of medicine 384, 2382–2393 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Yaeger R. et al. , Molecular Characterization of Acquired Resistance to KRASG12C-EGFR Inhibition in Colorectal Cancer. Cancer discovery 13, 41–55 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Dilly J. et al. , Mechanisms of Resistance to Oncogenic KRAS Inhibition in Pancreatic Cancer. Cancer discovery 14, 2135–2161 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Cassier P. et al. , 613MO A phase I clinical trial evaluating exarafenib, a selective pan-RAF inhibitor in combination with binimetinib in NRAS-mutant (NRASMut) melanoma (Mel) & BRAF-altered solid tumors. Annals of Oncology 35, S491–S492 (2024). [Google Scholar]
  • 45.Yen I. et al. , Pharmacological Induction of RAS-GTP Confers RAF Inhibitor Sensitivity in KRAS Mutant Tumors. Cancer cell 34, 611–625 e617 (2018). [DOI] [PubMed] [Google Scholar]
  • 46.Hofmann M. H. et al. , BI-3406, a Potent and Selective SOS1-KRAS Interaction Inhibitor, Is Effective in KRAS-Driven Cancers through Combined MEK Inhibition. Cancer discovery 11, 142–157 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Tedeschi A. et al. , Pan-KRAS Inhibitors BI-2493 and BI-2865 Display Potent Antitumor Activity in Tumors with KRAS Wild-type Allele Amplification. Molecular cancer therapeutics 24, 550–562 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Okuzumi T. et al. , Inhibitor hijacking of Akt activation. Nature chemical biology 5, 484–493 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Pearce L. R. et al. , Characterization of PF-4708671, a novel and highly specific inhibitor of p70 ribosomal S6 kinase (S6K1). Biochem J 431, 245–255 (2010). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1

Data Availability Statement

All data needed to evaluate the conclusions in the paper are present in the paper or the supplementary materials. The atomic coordinates and structure factors of kinase BRAF (chain A and chain B) bound to ELV-3111 have been deposited in the Protein Data Bank and can be accessed using the accession number 9ZBO.


Articles from bioRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES