D3S-001, a KRAS G12C Inhibitor with Rapid Target Engagement Kinetics, Overcomes Nucleotide Cycling, and Demonstrates Robust Preclinical and Clinical Activities

Abstract

First-generation KRAS G12C inhibitors, such as sotorasib and adagrasib, are limited by the depth and duration of clinical responses. One potential explanation for their modest clinical activity is the dynamic "cycling" of KRAS between its guanosine diphosphate (GDP)- and guanosine triphosphate (GTP)-bound states, raising controversy about whether targeting the GDP-bound form can fully block this oncogenic driver. We herein report that D3S-001, a next-generation GDP-bound G12C inhibitor with faster target engagement (TE) kinetics, depletes cellular active KRAS G12C at nanomolar concentrations. In the presence of growth factors, such as epithelial growth factor and hepatocyte growth factor, the ability of sotorasib and adagrasib to inhibit KRAS was compromised whereas the TE kinetics of D3S-001 was nearly unaffected, a unique feature differentiating D3S-001 from other GDP-bound G12C inhibitors. Furthermore, the high covalent potency and cellular TE efficiency of D3S-001 contributed to robust antitumor activity preclinically and translated into promising clinical efficacy in an ongoing phase 1 trial (NCT05410145). Significance: The kinetic study presented in this work unveils, for the first time, that a GDP-bound conformation-selective KRAS G12C inhibitor can potentially deplete cellular active KRAS in the presence of growth factors and offers new insights into the critical features that drive preclinical and clinical efficacy for this class of drugs.

Figure 1.

D3S-001 is a KRAS G12C inhibitor with enhanced covalent potency and TE kinetics. A, Cellular activity of KRAS G12C inhibitors in NCI-H358 NSCLC cells was evaluated through active RAS-GTP pull-down followed by immunoblotting using a KRAS-specific antibody after a 2-hour treatment with different inhibitors at indicated concentrations. The levels of active KRAS were quantified by chemiluminescence intensity, and IC₅₀ values were calculated using GraphPad Prism for each compound. B, Maximal covalent inactivation rate (k_inact), the concentrations that achieve a half-maximal rate (K_I) and other kinetic parameters of different KRAS G12C inhibitors determined by SPR. The top doses tested were 20 μmol/L for ARS-853 and ARS-1620, 200 nmol/L for sotorasib and adagrasib, and 20 nmol/L for D3S-001, respectively. Three independent experiments were performed, and data are presented as mean ± SD. C, Free–cysteine proteome analysis of D3S-001. NCI-H358 whole-cell lysates were extracted and subjected to proteomics analysis after a 4-hour treatment with 10-nmol/L D3S-001 or DMSO. Criteria for covalent targets (highlighted in gray) were set as < −2.0 log₂ fold change with a P value of less than 0.001 across biological replicates (n = 3 replicates). D, Kinetics of cellular active RAS depletion in NCI-H358 cells after treatment with a time course and dose titration of KRAS G12C inhibitors were determined by RAS-GTP ELISA assay. The observed inhibition rate k_obs and concentrations to achieve a half-maximal rate [I]₅₀ were determined using GraphPad Prism. Cellular TE efficiency Max k_obs/[I]₅₀ values were calculated. E, Cellular kinetic parameters of different KRAS G12C inhibitors determined from time course and dose response studies as illustrated in D. Inhibition rate t_1/2 was calculated by ln2/Max k_obs.

Figure 2.

TE of D3S-001 is insusceptible to EGF stimulation. A, Kinetics of cellular active RAS depletion in NCI-H358 cells after treatment with ARS-853 or ARS-1620 at 1 μmol/L, sotorasib, adagrasib, or D3S-001 at 100 nmol/L, in the absence (solid line) or presence (dotted line) of 40 ng/mL EGF was determined by RAS-GTP ELISA assay. B, EGF treatment stimulates the transition of RAS to GTP-bound form. NCI-H358 cells were treated with 40 ng/mL EGF at different times as indicated. Cell extracts were prepared and subjected to (left) pull-down assay and (right) pERK HTRF assay at the indicated time post-treatment. For the pull-down assay, the levels of active KRAS were quantified by chemiluminescence intensity and normalized to vehicle. For pERK HTRF, the levels of pERK were quantified by HTRF and normalized to vehicle. C, NCI-H358 cells were treated with D3S-001, sotorasib, or adagrasib with or without concurrent EGF stimulation (40 ng/mL; top left), HGF stimulation (40 ng/mL; top middle), at 100 nmol/L for 2 hours, or with or without concurrent EGF stimulation (40 ng/mL) and KRAS G12C inhibitors at their corresponding [I]₅₀ concentrations for 1 hour (top right). Extracted cell lysates were subjected to pull-down assay to determine the effect on active KRAS. The levels of active KRAS were quantified by chemiluminescence intensity and normalized to DMSO (bottom). D, NCI-H358 cells were treated with EGF (40 ng/mL) with or without concurrent D3S-001, sotorasib, or adagrasib at their corresponding 5 × [I]₅₀ concentrations for a time course as indicated. Cell extracts were subjected to pERK HTRF assays, and the levels of pERK were quantified by HTRF and normalized to vehicle.

Figure 3.

D3S-001 demonstrated potent antitumor activity in vitro. A, Effect of sotorasib, adagrasib, and D3S-001 on phospho-ERK measured by HTRF assay in NCI-H358 cells (top) and MIA PaCa-2 cells (bottom). Cells were treated with sotorasib, adagrasib, and D3S-001 at indicated concentrations for 2 hours, and then cell lysates were extracted and subjected to pERK HTRF assay (n = 3 replicates). B, Effect of D3S-001 on cell proliferation in a panel of human cancer cell lines by 2D (top) and 3D (bottom) CellTiter-Glo assay (n = 3 replicates). C, IC₅₀ values of each compound tested in 2D (top) and 3D (bottom) assays. Each point represents an individual cell line. Folds of selectivity between the median IC₅₀ of KRAS G12C-mutant cell lines and non-KRAS G12C mutant cell lines were calculated.

Figure 4.

D3S-001 inhibited KRAS signaling and tumor growth in vivo. A, NCI-H358 xenograft model was treated with a single oral dose of D3S-001 at 3, 10, 30, and 100 mg/kg, and tumor samples were collected at 6 hours post-treatment (n = 4 mice). In vivo PD analysis was performed by immunoblotting. The total and phosphorylated levels of ERK1/2 and RSK were analyzed, and GAPDH was used as a loading control. Levels of active ERK1/2 and RSK were quantified by chemiluminescence intensity and normalized to total ERK1/2 and RSK, respectively. B, The effect of D3S-001 on 10 MAPK signature genes was assessed by RNAseq. Tumor tissue samples were collected at 6 hours after a single dose of D3S-001 at different dose levels from the same batch of experiments and subjected to RNA extraction followed by RNAseq analysis. Data are shown as mean ± SEM. C,In vivo depletion of active RAS (bar graph) was determined by RAS-GTP ELISA. Plasma concentrations (red triangles) or tumor concentrations (black open circles) were determined by LC-MS/MS. Plasma and tumor tissue samples were collected at 6 hours after a single dose of D3S-001 at different dose levels from the same batch of experiments of the dose-PD study. Data are presented as mean ± SEM, with n = 4 mice per dose group. D,In vivo tumor growth of the NCI-H358 xenograft model after treatment with D3S-001 at 3, 10, 30, and 100 mg/kg once daily. D3S-001 was administered via oral gavage daily until day 31. Data are shown as mean tumor volume ± SEM. E, Left, Summary of TGI and plasma PK from the NCI-H358 xenograft model in D. Mouse plasma samples were collected at the end of efficacy studies (0.25, 0.5, 1, 2, 4, 8, 24 hours post final dose). Right, Correlation of D3S-001 plasma PK exposure with TGI values was analyzed using a nonlinear regression model in GraphPad Prism.

Figure 5.

D3S-001 treatment resulted in tumor regressions in multiple cell line-derived and PDX models of different cancer types. A–C, Top, Tumor growth of MIA PaCa-2 pancreatic cancer xenograft model (A), SW837 colorectal model (B), and YUO142 NSCLC PDO model (C) after oral administration with sotorasib (30 mg/kg, QD), adagrasib (30 mg/kg, QD), and D3S-001 (10, 30, or 100 mg/kg, QD). Middle, TGI and plasma PK summary for each model. 6% and 1% free-drug fractions of sotorasib and adagrasib in mouse plasma were used for free-drug AUC calculation, respectively. Bottom, Individual mouse tumor volume changes at day 52 for the MIA PaCa-2 model, day 35 for the SW837 model, and day 21 for the YUO142 model, respectively. D, D3S-001 was administered via oral gavage at 30 mg/kg every day to mice bearing 10 different KRAS G12C-mutant CRC PDX models as indicated. Tumor volume changes from baseline were calculated after 21 days of treatment. E, D3S-001 (30 mg/kg, QD, p.o.) was co-administered with cetuximab (30 mg/kg, BIW. i.p.) to mice bearing the same panel of CRC PDX models. Tumor volume changes from baseline were calculated after 21 days of treatment.

Figure 6.

Evaluation of the effect of D3S-001 monotherapy or in combination with D3S-002, a clinical-stage ERK1/2 inhibitor, in brain metastatic model and sotorasib-resistant models. A, Plasma and cerebrospinal fluid (CSF) PK analysis of D3S-001 in male beagle dogs after a single oral dose at 30 mg/kg. Plasma and CSF samples were collected at 1, 4, 8, and 24 hours postdose. Drug levels in the plasma (nmol/L) and CSF (nmol/L) are shown from n = 4 beagle dogs as mean ± SD. B, Plasma and brain tissue free-drug analysis of D3S-001 in SD rats after a single i.v. bolus at 4 mg/kg. Plasma samples were collected at 0.25, 0.5, 1, 2, 4, and 6 hours postdose, and brain samples were collected at 0.5, 1, 2, and 6 hours postdose. Drug levels of D3S-001 in the plasma (nmol/L) and brain homogenate (nmol/kg) are shown from n = 3/sex SD rats as mean ± SD. C, Intracranial tumor growth of NC-H1373-Luc model after oral administration with sotorasib (30 and 100 mg/kg, QD), adagrasib (30 and 100 mg/kg, QD), or D3S-001 (30 and 100 mg/kg, QD). Tumor growth was imaged twice per week after grouping by bioluminescent imaging using a living image program (PerkinElmer, IVIS Lumina Series III). Data are shown as mean total flux ± SEM. D, Representative tumor image of NC-H1373-Luc intracranial tumor model presented in C. E, Effect of sotorasib (30 mg/kg, QD, p.o.), adagrasib (30 mg/kg, QD, p.o.), D3S-001 (30 mg/kg, QD, p.o.), D3S-002 (ERK1/2 inhibitor, 25 or 50 mg/kg, QD, p.o.), and the combination of D3S-001 (30 mg/kg, QD, p.o.) + D3S-002 (25 or 50 mg/kg, QD, p.o.) on in vivo tumor growth of sotorasib-resistant MIA PaCa-2 cells with KRAS G12C gene amplification. Data are shown as mean tumor volume ± SEM. F, Event-free survival of sotorasib-resistant MIA PaCa2 CDX model in E using tumor volume >1,000 mm³ as a surrogate endpoint. Survival data were analyzed by Kaplan–Meier statistical analysis. G, Effect of sotorasib (30 mg/kg, QD, p.o.), adagrasib (30 mg/kg, QD, p.o.), D3S-001 (30 or 100 mg/kg, QD, p.o.) on in vivo tumor growth of sotorasib-resistant PDX model. Data are shown as mean tumor volume ± SEM. H, Event-free survival of sotorasib-resistant PDX model in G using tumor volume >500 mm³ as a surrogate endpoint. Survival data were analyzed by Kaplan–Meier statistical analysis.

Figure 7.

Activity of D3S-001 in patients with NSCLC. A, The top shows baseline and C3D1 scans of a pretreated NSCLC patient with a KRASG12C mutation indicating a 47.4% reduction of a target lesion (liver metastasis). Partial response was confirmed on subsequent scans. The patient has been on treatment for more than 16 months as of March 2024 with intrapatient dose escalation in cycle 11 to 100 mg daily then to 200 mg daily in cycle 16. The patient is currently maintaining PR (target lesion decrease of −86.8%) at 400 mg daily from cycle 19. The bottom shows CT scans of the brain at baseline and at C3D1. A decreased metastatic lesion (nontarget lesion) in the cerebellum was observed. B, Baseline, C1D8, and C7D1 ctDNA of the patient. KRAS VAF of 8.76% at baseline decreased to nondetectable at C1D8 and remained undetected at C7D1. C, Plasma PK of the patient on C1D1 (day 1) and C2D1 (steady state). D, The top shows baseline and C3D1 scans of a pretreated NSCLC patient with a target lesion in the left hilar lymph node. The bottom shows baseline and C3D1 scans of another target lesion in the left axilla lymph node, indicating an overall 49% reduction of a target lesion (liver metastasis). Partial response was confirmed on subsequent scans. The patient has been on treatment for a total of 11.1 months with intra-patient dose escalation to 100 mg daily in cycle 10, then to 200 mg daily in cycle 15. The patient experienced disease progression after 8.3 months. E, Baseline, C1D8, and C7D1 ctDNA of the patient. KRAS VAF of 6.49% at baseline decreased to 2.47% at C1D8 and to undetected at C4D1. F, Plasma PK of the patient on C1D1 (day 1) and C2D1 (steady state).