Targeting Ras-, Rho-, and Rab-family GTPases via a conserved cryptic pocket

Abstract

The family of Ras-like GTPases consists of over 150 different members, regulated by an even larger number of guanine exchange factors (GEFs) and GTPase-activating proteins (GAPs) that comprise cellular switch networks that govern cell motility, growth, polarity, protein trafficking, and gene expression. Efforts to develop selective small molecule probes and drugs for these proteins have been hampered by the high affinity of guanosine triphosphate (GTP) and lack of allosteric regulatory sites. This paradigm was recently challenged by the discovery of a cryptic allosteric pocket in the switch II region of K-Ras. Here, we ask whether similar pockets are present in GTPases beyond K-Ras. We systematically surveyed members of the Ras, Rho, and Rab family of GTPases and found that many GTPases exhibit targetable switch II pockets. Notable differences in the composition and conservation of key residues offer potential for the development of optimized inhibitors for many members of this previously undruggable family.

Figure 1.. Covalent SII pocket inhibition of K-Ras(G12C), H-Ras(G12C), and N-Ras(G12C)

(A) X-ray structure of K-Ras(G12C) bound to sotorasib (PDB: 6OIM). (B) Chemical structures of optimized K-Ras(G12C) inhibitors tested in our screen. (C) Time-dependent covalent modification of K-Ras(G12C) by various compounds (5 μM). (D) Time-dependent covalent modification of H-Ras(G12C) by various compounds (5 μM). (E) Time-dependent covalent modification of N-Ras(G12C) by various compounds (5 μM). (F) Crystal structure of H-Ras(G12C),GDP, sotorasib adduct. (G) Comparison of the structures of K-Ras(G12C),GDP,sotorasib (PDB: 6OIM) and H-Ras(G12C)·GDP·sotorasib (yellow). (H) Intrinsic or SOS- or EDTA-mediated nucleotide exchange of BODIPY-GDP with N-Ras(G12C) 00B7GDP and N-Ras(G12C)·GDP·sotorasib adduct. (I–K) Relative growth of MOLM-13-KRAS-G12C (I), MOLM-13-NRAS-G12C (J), and MOLM-13-KRAS-G12D (K) cells after treatment with K-Ras(G12C) inhibitors for 72 h. Data are presented as mean ± SD (n = 3) and are representative of three independent experiments. See also Figure S1.

Figure 2.. Targeting Ras-family GTPases.

(A) Family tree of human superfamily of Ras-like GTPases. (B) Sequence alignment of various Ras-family GTPases. Residues mediating drug resistance to adagrasib are highlighted in light orange (rare) and orange (common). (C) Covalent modification of RalA(G23C) with compounds 1–10 (50 μM, 12 h). (D) Intact protein mass spectra of RalA(G23C)·GDP and RalA(G23C)·GDP·MRTX1257 adduct. (E) Time-dependent covalent modification of RalA(G23C) with different compounds (50 μM). (F) Differential scanning fluorimetry of RalA(G23C)·GDP and RalA(G23C)·GDP·divarasib adduct. (G) Covalent modification of Rap1A(G12C) with compounds 1–10 (50 μM, 12 h). (H) Intact protein mass spectra of Rap1A(G12C, L96F)·GDP and Rap1A(G12C,L96F)·GDP·divarasib adduct. (I) Time-dependent covalent modification of Rap1A(G12C) and Rap1A(G12C, L96F) with different compounds (50 μM). (J) Differential scanning fluorimetry of Rap1A(G12C, L96F)·GDP and Rap1A(G12C, L96F)·GDP·divarasib adduct. See also Figure S2.

Figure 3.. Cellular targeting of Ras-family GTPases.

(A) Peptides showing significant differences in HDX at any time point (>0.35 Da and >4.5%) mapped onto a homology model of RalA based on adagrasib-bound K-Ras(G12C) (PDB: 6USZ) according to the legend. (B) Intrinsic or RAPGEF5- or EDTA-mediated nucleotide exchange of BODIPY-GDP with Rap1A(G12C, L96F)·GDP and Rap1A(G12C, L96F)·GDP·divarasib adduct. (C) Immunoblot of HeLa cells transiently overexpressing EGFP-RalA(WT) and EGFP-RalA(G23C). (D) RalA activity measured by RalA G-LISA. HeLa cells were transiently transfected, treated with different concentrations of MRTX1257 for 12 h, and lysates were tested at 0.5 mg/mL. Data are presented as mean ± SEM (n = 2) and are representative of three independent experiments. (E) IP of active GTP-bound Rap1 using GST-RalGDS-RBD of HeLa cells transiently overexpressing EGFP-Rap1A(WT) and EGFP-Rap1A(G12C, L96F) and treated with different concentrations of divarasib. See also Figure S2.

Figure 4.. Cellular targeting of Ras-family GTPases with RMC6291.

(A) X-ray structure of tricomplex of RMC-4998, K-Ras(G12C), and CypA (PDB: 8G9P). K-Ras residues involved in binding RMC-4998 are colored in blue. Negatively charged K-Ras residues involved in binding CypA are colored in orange. (B) Sequence alignment of Ras-family GTPases K-Ras, M-Ras, R-Ras, and Rheb. Residues mediating drug resistance to RMC-4998 are highlighted in light orange (mild effect) and orange (strong effect). (C) Immunoblot of HeLa cells transiently overexpressing K-Ras or K-Ras(G12C). HeLa cells were transiently transfected, treated with different concentration ofRMC-6291 for 3 h, and blotted for Ras. (D) Immunoblot of HeLa cells transiently overexpressing M-Ras or M-Ras(G22C). HeLa cells were transiently transfected, treated with different concentration ofRMC-6291 for 3 h, and blotted for M-Ras. (E) Immunoblot of HeLa cells transiently overexpressing R-Ras1 or R-Ras1(G38C). HeLa cells were transiently transfected, treated with different concentration ofRMC-6291 for 3 h, and blotted for R-Ras. (F) Immunoblot of HeLa cells transiently overexpressing Rheb or Rheb(R15C). HeLa cells were transiently transfected, treated with different concentration of RMC-6291 for 3 h, and blotted for Rheb. Data are representative of three independent experiments.

Figure 5.. Targeting Rho- and Rab-family GTPases.

(A) Covalent modification of Rac1(G12C) with compounds 1–10 (50 μM, 12 h). (B) Covalent modification of RhoA(G14C) with compounds 1–10 (50 μM, 12 h). (C) Time-dependent covalent modification of Rac1(G12C), RhoA(G14C), and Rac1(WT) with divarasib (50 μM). (D) Time-dependent covalent modification of various Rac1 mutants with divarasib (50 μM). (E) Covalent modification of Rab1A(S20C) with compounds 1–10 (50 μM, 12 h). (F) Covalent modification of Rab5C(S30C) with compounds 1–10 (50 μM, 12 h). (G) Time-dependent covalent modification of Rab1A(S20C), Ypt1(S17C), Rab5C(S30C), and RabL5(WT) with MRTX1257 (50 μM). (H) Time-dependent covalent modification of various Rab1A mutants with MRTX1257 (50 μM). (I) Peptides showing significant differences in HDX at any time point (>0.35 Da and >4.5%) mapped onto a homology model of Rab1A based on adagrasib-bound K-Ras(G12C) (PDB: 6USZ). (J) Differential scanning fluorimetry of Rac1(G12C, K96H)·GDP and Rac1(G12C, K96H)·GDP·divarasib adduct. (K) Rac1 activity measured by Rac1 G-LISA. HeLa cells were transiently transfected, treated with different concentrations of divarasib for 12 h, and lysates were tested at 0.5 mg/mL. Data are presented as mean ± SEM (n = 2) and are representative of three independent experiments. See also Figures S2, S3, and S4.

Figure 6.. Ligand optimization for Rab and Rho GTPases.

(A and B) Representative binding poses from covalent MD simulations of divarasib, and selected ligands in Rac1(G12C) (A) and Rab5C(S30C) (B), respectively. (C) Chemical structures of novel SII pocket inhibitors to improve targeting of Rab and Rho GTPases. (D) Covalent modification of Rac1(G12C) with compounds 1–22 (50 μM, 1 h). (E) Covalent modification of Rab1A(S20C) with compounds 1–22 (50 μM, 1 h). (F) Covalent modification of Rab5C(S30C) with compounds 1–22 (50 μM, 12 h). See also Figures S5 and S6.