Search

Activating mutations in RAS proteins occur in ~1/3 of human cancers. These mutations impair the ability of the protein to hydrolyze GTP to GDP. As a result, mutant RAS proteins exist predominantly in the GTP-bound state, which directly activates aberrant downstream signaling via interaction with effectors such as RAF. Most RAS mutations occur at glycine 12 of the KRAS isoform. One such mutation, KRAS G12C, is particularly common in non-small cell lung cancer where it is found in ~15% of lung adenocarcinomas. Recent efforts have targeted KRAS G12C in the GDP-bound state; however, direct pharmacologic inhibition of active, GTP-bound KRAS G12C has proved challenging. Here, we deployed a novel SMARTTM (Small Molecule Assisted Receptor Targeting) platform to advance covalent compounds that selectively inhibit GTP-bound KRAS G12C. Using a mechanism reminiscent of the natural products rapamycin and cyclosporine, these compounds promote formation of a novel inhibitory ternary complex consisting of cyclophilin A (CypA, an abundant immunophilin present in all human cells), the SMART inhibitor, and GTP-KRAS G12C. Structure-based design of the SMART inhibitor yielded potent covalent inhibitors of GTP-KRAS G12C that exhibit >100-fold selectivity for mutant KRAS G12C over WT KRAS. Structural analysis of the ternary complex revealed that the covalent linkage between the SMART inhibitor and the mutant cysteine of KRAS occurred in the context of extensive interactions between CypA, the SMART inhibitor, and GTP-KRAS G12C that provide significant binding affinity (KI = 2.5 μM). The GTP-KRAS G12C|Inhibitor|CypA complex directly occluded effector binding, and as such, the compounds disrupted the KRAS-RAF interaction in biochemical assays. This activity was dependent on CypA, underlining the importance of the KRAS G12C| CypA protein-protein interaction in driving target engagement. In cell-based studies, SMART inhibitors crosslinked KRAS G12C and potently inhibited ERK phosphorylation and cell growth in G12C mutant tumor cell lines but had no effect on non-G12C bearing tumor cells. CRISPR knockout of cellular CypA confirmed that these activities were dependent on the presence of endogenous CypA. Importantly, SMART inhibitors bind directly to active, GTP-KRAS G12C and thus, their activity does not rely on trapping KRAS G12C in the inactive GDP-bound state. As a result, the cellular potency of SMART inhibitors with respect to crosslinking, pERK inhibition, and growth inhibition was maintained in the presence of growth factor treatments that reduce the cellular GDP-KRAS G12C pool. In contrast, we found that the activity of a previously described GDP-KRAS G12C targeting inhibitor was attenuated by growth factor treatment. To our knowledge, these are the first examples of mutant-selective KRAS inhibitors that target the active, GTP-bound state of KRAS G12C. We are currently optimizing the drug-like properties of these SMART inhibitors and evaluating their activity in in vivo models. Citation Format: Michelle L. Stewart, Nicholas R. Perl, Seung-Joo Lee, Linlong Xue, Minyun Zhou, Jonah Simon, Kathryn M. Luly, Siminia Grigoriu, Alex Yuzhakov, Alec Silver, Jason T. Lowe, Cindy C. Benod, Alan S. Mann, Gregory L. Verdine, Alan C. Rigby, Mark J. Mulvihill, Earl W. May, Anna Kohlmann, Sharon A. Townson, Roy M. Pollock, Meizhong Jin. Development of inhibitors of the activated form of KRAS G12C [abstract]. In: Proceedings of the AACR Special Conference on Targeting RAS-Driven Cancers; 2018 Dec 9-12; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(5_Suppl):Abstract nr B37.

RAS proteins are small GTPases involved in cell proliferation, survival, and differentiation, and are mutationally activated in about a third of all human cancers. These mutations drive cancer by impairing GTPase activity so that the RAS protein is found predominantly in its GTP-bound “on” conformation. Most KRAS isoform mutations are located at codon 12, a glycine in the P-loop of the GTPase active site. KRAS mutations in which the glycine is mutated to a cysteine (G12C) are particularly common in lung cancer. Despite its prevalence as an important oncogene and decades of research, the RAS protein remains an unexploited cancer target. We have developed a novel platform, SMART (Small Molecule Assisted Receptor Targeting), to disrupt the protein-protein interactions of so-called “undruggable” targets. Our compounds bind the immunophilin protein Cyclophilin A (CypA), and subsequently form a ternary complex with the target protein, KRAS. An electrophilic moiety on the ligand selectively forms a covalent bond with the cysteine of GTP-KRAS G12C, thus strengthening the stability of the ternary complex, partially occupying the effector face, and thus occluding the binding of downstream effector proteins, such as RAF. In a separate presentation, we describe a wide breadth of data to support the advancement of a small molecule specifically targeting the activated form of KRAS, with a focus on the cellular pharmacology. In this presentation, we highlight biophysical and biochemical findings that support our novel strategy for targeting GTP-KRAS G12C. SPR reveals the CypA-compound binary KD. A novel SPR protocol method was developed to examine CypA-compound-KRAS ternary complex formation, exploiting the A-B-A injection system of the Biacore 8K. LC-MS is used to characterize the crosslinking of our compounds to KRAS G12C in the presence of CypA, and full kinact/KI analysis is conducted with select compounds; in combination with SPR studies, the noncovalent KD of the ternary complex can be determined. Intrinsic warhead reactivity is measured through GSH adduct formation. TR-FRET studies probe the ability of ternary complexes to inhibit the binding of RAF to KRAS G12C. ITC and BLI are used to confirm ternary complex formation for selected compounds. In addition to guiding SAR, these studies characterize the degree of presenter (CypA) dependence, specificity of compound binding to the G12C mutant and to the GTP nucleotide-bound form of KRAS, and the ability to predict cellular efficacy. Our kinetic crosslinking experiments reveal single-digit micromolar KIs and a good kinact/KI ratio for our compounds relative to published literature on compounds targeting GDP-KRAS G12C. When these values are compared to SPR findings, the noncovalent KD contributes largely to KI, and thus noncovalent interactions play a bigger role in our compound’s interaction with RAS as compared to known GDP-KRAS G12C agents. Citation Format: Minyun Zhou, Alexander Yuzhakov, Cindy C Benod, Linlong Xue, Alec D Silver, Ganesh Iyer, Sharon A. Townson, Meizhong Jin, Nicholas R. Perl, Anna Kohlmann, Alan S. Mann, Gizem Akcay, Earl W. May. Biophysical and biochemical characterization of KRAS G12C inhibition through the SMARTTM platform [abstract]. In: Proceedings of the AACR Special Conference on Targeting RAS-Driven Cancers; 2018 Dec 9-12; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(5_Suppl):Abstract nr A06.