Drug Discovery Strategies for KRAS Small-Molecule Inhibitors—From Covalent Locking to Protein Degradation

The development of KRAS inhibitors spans over thirty years, with nearly two decades of stagnation followed by an explosion of progress in the last decade. The turning point came with new insights into the dynamic conformations of KRAS and advances in chemical technologies—particularly covalent chemistry, fragment-based drug design, and targeted protein degradation. Today, KRAS-targeted drug discovery has evolved into a diverse landscape with multiple mature strategies and varied chemotypes. This article systematically reviews the main inhibitor discovery strategies, screening technologies, and structural optimization pathways for targeting KRAS from a medicinal chemistry perspective.

The KRAS G12C mutation introduces a cysteine residue at position 12, located near the Switch II region. Covalent inhibitor design exploits the nucleophilicity of this cysteine by incorporating electrophilic warheads such as acrylamides, chloroacetamides, or epoxides, which form irreversible covalent bonds with the thiol group of Cys12. This mechanism allows the inhibitor to bind KRAS regardless of its GTP/GDP cycling state, effectively locking KRAS in place upon binding.

Typical covalent inhibitors (e.g., the ARS-1620 series) induce conformational rearrangements in the Switch II region through binding to the S-IIP pocket, favoring the GDP-bound inactive state of KRAS and blocking effector recruitment. Clinical candidate molecules optimized from this foundation have achieved breakthroughs in oral bioavailability, metabolic stability, and selectivity. Notably, the selectivity of covalent inhibitors relies on targeting specific cysteine residues, and other cysteines (e.g., Cys80, Cys118) are being explored as potential covalent modification sites to extend applicability beyond G12C mutants.

For mutants lacking reactive cysteine residues (e.g., G12D, G12V, Q61H), non-covalent inhibitors are the primary approach. Non-covalent design faces greater challenges due to the shallow binding pockets on the KRAS surface and the need to compete with millimolar intracellular GTP concentrations. Two main pathways exist:

(1) Direct competition with GTP binding. These nucleotide analogs compete with GTP for binding, but due to KRAS's extremely high affinity for GTP and high intracellular GTP levels, such inhibitors typically exhibit low activity and poor drug-like properties.

(2) Allosteric pocket inhibitors. Multiple inducible allosteric pockets have been discovered near the Switch I/II regions of KRAS. A representative example is the G12D inhibitor MRTX1133, which binds to a cleft between Switch I/II, stabilizing the GDP-bound inactive state through multiple hydrogen bonds and hydrophobic interactions. Despite non-covalent binding, it achieves sub-nanomolar affinity. Fragment-based NMR screening has identified additional allosteric binding sites on KRAS, including novel pockets beyond S-IIP, providing starting points for further optimization.

Covalent and non-covalent inhibitors primarily suppress KRAS's catalytic function without eliminating the protein itself. Resistance mutations often arise by altering residues at the inhibitor binding site or increasing KRAS expression levels. Protein degradation technologies offer an alternative to functional inhibition.

PROTAC (Proteolysis-Targeting Chimera) molecules consist of three components: a KRAS-binding ligand, an E3 ubiquitin ligase-recruiting ligand (e.g., CRBN or VHL ligands), and a linker chain. Preclinical KRAS G12C PROTAC molecules (e.g., LC-2) induce KRAS degradation at nanomolar concentrations and overcome certain resistance mutations. PROTACs targeting non-covalent mutants have also been reported. Since PROTACs catalytically degrade target proteins, the required drug concentrations are far lower than those of traditional inhibitors.

Molecular glues induce novel protein-protein interactions between KRAS and E3 ligases, enabling non-covalent cooperative degradation. Although KRAS molecular glues remain in early-stage discovery, this strategy avoids the complex chemical synthesis and high molecular weight issues of PROTACs, offering better drug-like potential.

KRAS inhibitor discovery heavily relies on advanced screening technologies. DNA-encoded compound libraries (DEL) and fragment-based screening using NMR/X-ray crystallography are common methods for hit identification. Fragment screening starts with low-molecular-weight compounds (<300 Da) that weakly interact with various sites on the KRAS surface, followed by fragment growth or linking to yield potent inhibitors. Structure-based virtual screening leverages high-resolution crystal structures of KRAS mutants to computationally dock large compound libraries, significantly reducing physical screening costs. Additionally, microfluidic capillary electrophoresis and TR-FRET high-throughput screening platforms enable direct quantification of KRAS-effector interactions, facilitating rapid screening of tens of thousands of samples per day.

Beyond directly targeting KRAS, modulating its upstream regulators (e.g., SOS1, NF1, SHOC2) has emerged as an indirect inhibition strategy. Small-molecule SOS1 inhibitors (e.g., BI-3406) block GEF-catalyzed nucleotide exchange, causing KRAS to accumulate in the GDP-bound inactive state. This approach shows synergy in KRAS-mutant tumors and can be combined with direct KRAS inhibitors. Similarly, inhibitors of the SHOC2-MRAS-PP1C complex, which block RAF activation, have proven effective in suppressing KRAS signaling.

Small-molecule KRAS inhibitor discovery has evolved into a multidimensional landscape featuring multiple strategies and modalities. Covalent inhibitors have led the way in clinical translation, non-covalent inhibitors address gaps for other mutants, degradation technologies show promise in overcoming resistance, and indirect strategies targeting upstream regulators expand the intervention window. Future KRAS drug development will increasingly adopt "combo approaches"—combining molecules with different mechanisms and simultaneously targeting KRAS and its compensatory pathways—to achieve deeper and more sustained pathway inhibition.

```