Review Papers Claim by Claim

 Quick Explanation Copied

 Long Explanation

Core findings across four orthogonal methods

The paper integrates ³¹P NMR, ¹⁵N CPMG relaxation dispersion, HDX-MS, and 10 × 1 µs MD replicates per protein (Amber20, ff19SB) to quantify the state-1/state-2 switch-1 equilibrium in GTP-analog-bound KRAS and NRAS RSC Chem Biol KRAS/NRAS dynamics paper. The headline surprise: codon-12 KRAS oncogenic mutants shift the equilibrium toward the inactive state, opposite to naive expectations of gain-of-function activation.

Values as reported in the paper's Fig. 2A–C; KRAS G12C plotted at wt levels per authors' statement that it "retains populations similar to wildtype." Exact G12C percentages are not given numerically in the text — treated as approximate.

The Q61R mechanism and the T35S dominance hierarchy

NRAS Q61R fully shifts to state-2, rationalized by MD as a transient Arg61–Thr35-backbone-carbonyl H-bond (~33% occupancy) consistent with Q61R co-crystal structures (6XGU, 6ZIZ). Strikingly, in the NRAS T35S/Q61R double mutant, T35S dominates — losing one methyl group outcompetes the arginine stabilization — showing epistasis in the allosteric network Q61R H-bond mechanism and T35S epistasis.

MRTX1133 mechanism of action

Using ¹H,¹³C-HMQC competition with RAF1-RBD, the authors show MRTX1133 allosterically prevents effector binding at physiological concentrations: the piperazine–Asp12 salt bridge displaces the Ala59/Gly60 loop, whose β-methyl pushes Thr35 and flips switch-1 open. RBD quenching only occurs at 100 µM RBD (non-physiological). This is a cleaner, more decisive allosteric MoA than the ambiguous BBO-8956 spectra of Sharma et al. MRTX1133 allosteric mechanism.

Critical caveats

• GMPPNP ≠ GTP: the rank order of state-1 population changes slightly between analogs (G12C vs wt swap positions); ³¹P work is GMPPNP-based, CPMG GTP-based — an acknowledged internal inconsistency the authors handle but cannot eliminate.
• In vitro, dilute, low-temperature (7 °C) NMR conditions and MD force-field assumptions (GMPPNP modeled as GTP; Mg²⁺ LJ parameters) mean absolute populations may differ in cells, where membrane anchoring, GEF/GAP, and crowding matter — none of which are modeled.
• HDX-MS coverage gaps at T35S and residue-61 peptides prevent direct switch-1 comparison for key mutants; the authors disclose this.
• Counterintuitive result demands caution: if G12 mutants slightly favor the inactive state, oncogenicity must reside in nucleotide loading/localization — a claim asserted but not tested here (no cellular assays).

What would falsify the core model: repeat ³¹P NMR with true GTP and physiological Mg²⁺/temperature showing reversed mutant ordering, or MD with corrected nucleotide parameters failing to reproduce the Arg61–Thr35 H-bond.

 BGPT Paper Review

Study Novelty

The state-1/state-2 paradigm is decades old (Kalbitzer/Hamasaki era) and Sharma et al. 2024 recently published overlapping KRAS 31P NMR mutant data; novelty rests on extending to NRAS Q61R, the Arg61–Thr35 H-bond mechanism, and the MRTX1133 allostery MoA — incremental but substantive.

Scientific Quality

Four orthogonal methods converge with quantified populations, uncertainties on CPMG fits, deposited HDX data (PXD054924), and clear disclosure of coverage gaps. Caveats: 7 °C NMR conditions, GMPPNP-as-GTP modeling, ~1.3 Å RMSD threshold for state classification in MD is heuristic, and no biological/cellular validation. Corporate authorship (Novartis) on inhibitor MoA warrants scrutiny, though declared no conflicts.

Study Generality

Findings center on two RAS isoforms and specific mutants; the state-1/state-2 framework and allosteric-node concept (Ala59, Thr35) generalize to the RAS family, but cellular-context generality is untested.

Study Usefulness

Directly informs G12D/Q61R drug design: conformational state populations offer a screening readout, and the MRTX1133 allosteric relay explains clinical inhibitor function and guides effector-blocking chemotypes.

Study Reproducibility

Detailed expression/purification/NMR/MD protocols; HDX data deposited (PRIDE PXD054924); CPMG WT data reusable from Dryad. MD depends on proprietary MOE preparation; exact 31P state populations rely on spectral deconvolution reproducible only with ESI data.

Explanatory Depth

Mechanistic depth is high: residue-level H-bond occupancies, crystallographic corroboration, epistasis between T35S and Q61R, and a structurally explicit allosteric relay for MRTX1133 — though causal links between occupancy percentages and populations remain inferential.