Jacqueline V. Aredo

Comutations converge on the RAS pathway to increase KRAS codon 13 oncogenicity. A, Relative proliferation rates of advanced NSCLC cell lines cultured in medium containing 10% FBS. Cell lines are grouped and averaged by mutant KRAS genotype, with comparisons made between (left) KRASG12C and KRASG13C cell lines and (right) KRASG12D and KRASG13D cell lines. Each dot represents a time point, which is normalized to 0 hours. n = 3–6 cell lines per genotype, with three biological repeats carried out for each cell line; mean ± SEM. B, Relative proliferation rates of advanced NSCLC cell lines cultured in medium containing 0.5% FBS. Cell lines are grouped and averaged by mutant KRAS genotype, with comparisons made between (left) KRASG12C and KRASG13C cell lines and (right) KRASG12D and KRASG13D cell lines. Each dot represents a time point, which is normalized to 0 hours. n = 3–6 cell lines per genotype, with three biological repeats carried out for each cell line; mean ± SEM. C, Western blotting of advanced cancer cell lines 48 hours after plating in 3D ULA conditions in medium containing 10% FBS, seeded at equal cell density. Cell lines were run on two gels, with MOR cells run on both to allow quantification and comparison between both blots. Data are representative of 2–4 biological repeats. β-Actin was used as a protein loading control (sample processing control). Dashed lines separate cell lines belonging to each KRAS-mutant genotype. D, Quantification of C. Phosphorylated protein and total protein run on separate blots were normalized to the sample processing control, and then phospho-protein was normalized to the total protein. The data were then normalized to MOR cells to allow comparisons between blots. n = 2–4 biological repeats per cell line; mean ± SEM. The comutations associated with each cell line are listed underneath. Black + indicates pathogenic alteration, gray + indicates a variant of unknown significance, and – indicates no alteration. E, Western blotting of MOR-LKB1 overexpressing (OE) cells 24 hours after DOX-induced expression (1,000 ng/mL) of LKB1. Cells were plated in 3D ULA conditions. Data are representative of 3–4 biological repeats for every protein probed. Vinculin was used as a protein loading control. The dashed line indicates that representative blots from two different biological replicates are presented. F, Quantification of E. Phosphorylated protein and total protein run on separate blots were normalized to their respective loading controls on the same blot, and then phospho-protein was normalized to the total protein. The data were then normalized to untreated (no DOX) control. n = 3–4 biological repeats for every protein probed; mean ± SEM. G, Relative spheroid area of MOR parental and MOR-LKB1 OE cells 8 days after DOX-induced expression (1,000 ng/mL) of LKB1. Cells were plated in ULA conditions. Spheroid area was analyzed using ImageJ. Data are expressed relative to untreated (no DOX) control. n = 3 biological repeats; mean ± SEM. H, Western blotting of HCC515 and H647 cells 72 hours after exposure to a cocktail of four siRNAs targeting NF1 (siNF1) pooled together or nontargeting siRNA (siNTC). Data are representative of 3 or 4 biological repeats. Vinculin was used as a protein loading control. I, Quantification of H. Phosphorylated protein and total protein run on separate blots were normalized to their respective loading controls on the same blot, and then phospho-protein was normalized to the total protein. Data within each cell line were then normalized to siNTC control. n = 3 or 4 biological repeats; mean ± SEM. J, Relative spheroid area of HCC515 and H647 cells 96 and 144 hours after exposure to a cocktail of four siRNAs targeting NF1 (siNF1) pooled together or nontargeting siRNA (siNTC). Cells were plated in ULA conditions. Spheroid area was analyzed using ImageJ. Data within each cell line are expressed relative to siNTC. n = 3 biological repeats per cell line; mean ± SEM. Statistical significance between comparisons was calculated using a two-way ANOVA (A, B, and G) and an unpaired Student t test (F, I, and J) and ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns > 0.05.

Examining the response and molecular determinants of PD-1 immune checkpoint blockade in KRASG13X NSCLC. A, Schematic illustrating the use of RAS-PM to explore ICI-specific outcomes, TMB, and PD-L1 in patients affected by advanced/inoperable KRAS-mutant NSCLC. B, Percentage of patients within the codon 13 cohort harboring KRASG13C, KRASG13D, and other codon 13 mutations. C, Analysis of TMB between KRASG12C and KRASG13C patients and between KRASG12D and KRASG13D patients from cBioPortal. D, PFS between KRAS codon 12 and 13 mutant NSCLC patients receiving immunotherapy as first-line treatment. E, OS between KRAS codon 12 and 13 mutant NSCLC patients receiving immunotherapy first-line treatment. F, The percentage of comutations associated with KRASG12C-, KRASG12D-, KRASG13C-, and KRASG13D-mutant patients taken from the 100,000 Genomes Project. Comparisons are as follows: KRASG12X (n = 155) vs. KRASG13X (n = 26), KRASG12C (n = 115) vs. KRASG13C (n = 14), and KRASG12D (n = 40) vs. KRASG13D (n = 12). G, The percentage of comutations associated with KRASG12C-, KRASG12D-, KRASG13C-, and KRASG13D-mutant patients taken from RAS-PM. Comparisons are as follows: KRASG12X vs. KRASG13X, KRASG12C vs. KRASG13C, and KRASG12D vs. KRASG13D. Not every patient was tested for all 5 comutations simultaneously. Refer to Supplementary Table S3 for the numbers of patients tested for each comutation. H, Schematic illustrating factors contributing to the lack of difference in immunotherapy response between KRASG12X- and KRASG13X-driven NSCLC patients from the RAS-PM database. Statistical significance between comparisons was calculated using an unpaired Student t test (C), log-rank (Mantel–Cox; D and E), and Fisher exact test (F and G). ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns > 0.05. A, Created in BioRender. Tinsley, K. (2026) https://BioRender.com/vyu8cs2; H, Created in BioRender. Tinsley, K. (2026) https://BioRender.com/4ju5v3u.

KRAS codon 13 mutations confer reduced oncopotency. A, Schematic outlining the generation of DOX-inducible isogenic MLE-12 and BEAS-2B cells to compare the effect of expressing different mutant KRAS variants on NSCLC initiation. B, Relative viability of isogenic MLE-12 cells 96 hours after DOX-induced expression (100 ng/mL) of each mutant KRAS variant in three different serum conditions – 2%, 0.5%, and 0.1% FBS. Cells were plated in ULA conditions. Viability was analyzed using CellTiter-Glo 3D. Data for each isogenic line are expressed relative to the untreated (no DOX) control of that particular line. n = 3 biological repeats; mean ± SEM. C, Western blotting of isogenic MLE-12 cells 24 hours after DOX-induced expression (100 ng/mL) of each mutant KRAS variant in 0.5% serum. Cells were plated in ULA conditions. Data are representative of five biological repeats for KRAS effector proteins and four biological repeats for KRAS target proteins. Dashed lines indicate that representative blots from three different biological replicates are presented. Vinculin and α-tubulin were used as protein loading controls. D, Quantification of C and associated replicates. For phospho-proteins (KRAS effector blots), phosphorylated protein and total protein run on separate blots were normalized to their respective loading controls on the same blot, and then phospho-protein was normalized to the total protein. Data for each isogenic line were then normalized to the untreated (no DOX) control of that particular line. KRAS target proteins were normalized to the loading control, and then the data were normalized to the untreated (no DOX) control. n = 5 biological repeats for KRAS effector proteins and four biological repeats for KRAS target proteins; mean ± SEM. E, (Left) Timeline of BEAS-2B oncogenicity in vivo experiment using chicken egg xenograft. (Right) Normalized tumor weights measured upon termination of the experiment. Number of eggs indicated per condition. Mean ± SEM reported. Statistical significance between comparisons was calculated using a two-way ANOVA (B, D, and E). ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns, > 0.05. A, Created in BioRender. Tinsley, K. (2026) https://BioRender.com/ojsukx1; E, (Left) Created in BioRender. Tinsley, K. (2026) https://BioRender.com/wrfehon.

RMC-8839 and taxanes synergize in KRASG13C-driven NSCLC models. A, H1734 and MOR cells were exposed to increasing concentrations of KRASG13C inhibition (RMC-8839) and either docetaxel (DOC) or paclitaxel (PAC) in both 2D and 3D conditions. After 72 hours, viability was analyzed by CellTiter-Glo (3D) and expressed as a percentage of DMSO control. Drug interaction was calculated using Bliss synergy (BS) analysis. n = 3 biological repeats per cell line; the mean is shown. BS less than −10 denotes antagonism; a BS score more than −10 but less than 10 denotes additivity; and a BS above 10 denotes synergy. In BS graphs, G13Ci indicates RMC-8839 B, (Left) Schematic: H1734 cells were transplanted subcutaneously into the flank of NOD-SCID mice. Treatments began when the mean tumor volumes reached approximately 183 mm3. A single intravenous dose of docetaxel at 5 mg/kg was administered on day 0; RMC-8839 was administered orally at 100 mg/kg daily for 28 days. Tumor volume and body weight were measured biweekly throughout the treatment period. (Right) Mean tumor volumes. n = 9 mice per group; mean ± SEM. Statistical significance between treatment groups was calculated using a two-way ANOVA. ****, P < 0.0001. B, (Left) Created in BioRender. Tinsley, K. (2026) https://BioRender.com/9dc6kig.

Examining a novel RAS(ON)-G13C selective inhibitor, RMC-8839. A, Western blotting of isogenic MLE-12 cells previously treated with DOX for 24 hours to induce KRAS variant expression in 0.5% serum followed by 6 hours of treatment with 50 nmol/L RMC-7977. Cells were plated in ULA conditions. Data are representative of three biological repeats. Vinculin was used as a protein loading control. B, Quantification of pERK 1/2 from A. Phosphorylated ERK 1/2 and total ERK 1/2 run on separate blots were normalized to their respective loading controls on the same blot, and then pERK 1/2 was normalized to total ERK 1/2. The data for each isogenic line were then normalized to the DMSO control of that line. n = 3 biological repeats per isogenic cell line; mean ± SEM. C, Dose-response analysis of isogenic MLE-12 cells previously treated with DOX for 24 hours to induce KRAS variant expression in 0.5% serum, followed by 72 hours of treatment with a concentration range of RMC-7977. Cells were plated in ULA conditions. Viability was analyzed by CellTiter-Glo 3D. Data are expressed as a % of DMSO control. n = 3 biological repeats per isogenic cell line; mean ± SEM. IC50 values were calculated for each cell line and presented as a table. D, Western blotting of isogenic MLE-12 cells previously treated with DOX for 24 hours to induce KRAS variant expression in 0.5% serum, followed by 6 hours of treatment with 50 nmol/L RMC-8839. Cells were plated in ULA conditions. Data are representative of three biological repeats. Vinculin was used as a protein loading control. E, Quantification of pERK 1/2 from D. Phosphorylated ERK 1/2 and total ERK 1/2 run on separate blots were normalized to their respective loading controls on the same blot, and then pERK 1/2 was normalized to total ERK 1/2. The data for each isogenic line were then normalized to the DMSO control of that line. n = 3 biological repeats per isogenic cell line; mean ± SEM. F, Dose-response analysis of isogenic MLE-12 cells previously treated with DOX for 24 hours to induce KRAS variant expression in 0.5% serum followed by a 72-hour treatment with a concentration range of RMC-8839. Cells were plated in ULA conditions. Viability was analyzed by CellTiter-Glo 3D. Data are expressed as a percentage of DMSO control. n = 3 biological replicates per isogenic cell line; mean ± SEM. IC50 values were calculated for each cell line and presented as a table. G, Viability in response to 10 nmol/L RMC-7977 in advanced KRAS-mutant NSCLC cell lines grouped by KRAS codon genotype. Cells were plated in ULA conditions. Viability was analyzed after 72 hours by CellTiter-Glo 3D. Each dot represents a cell line within that genotype. Data are expressed as a percentage of DMSO control. n = 3 biological repeats per cell line; mean ± SEM. H, Western blotting of KRASG13C advanced cancer cell lines 24 hours after exposure to 100 nmol/L RMC-8839. Cells were plated in ULA conditions. Data are representative of three biological repeats per cell line. Vinculin was used as a protein loading control. I, Dose response analysis of advanced cancer cell lines harboring KRASG13C (H1734, H1355, HCC4087, and MOR), KRASG12C (H2030), KRASG12D (SKLU-1), and KRASG13D (H647) exposed to increasing concentrations of RMC-8839 for 168 hours. Viability was measured by CellTiter-Glo 3D. Data are expressed as a percentage of DMSO control. n = 3–4 biological repeats per cell line; mean ± SEM. IC50 values were calculated for each cell line and presented as a table. J, Apoptosis detection (left) and quantification (right) of H1734, MOR, HCC4087, and H1355 cells after 120 hours of exposure to 100 nmol/L RMC-8839. Cells were stained with annexin V and propidium iodide to detect apoptotic cells. n = 3 biological repeats per cell line; mean ± SEM. Statistical significance between comparisons was calculated using an unpaired Student t test (B, E, and G) and a two-way ANOVA (J). ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns > 0.05.

Selective sensitivity of KRASG13C NSCLC to chemotherapy. A, Top 25 compounds demonstrating the highest sensitivity to KRASG13C-driven NSCLC cell lines compared with cell lines with other KRAS-mutant genotypes. Each dot represents a cell line and is the mean of two repeats. The dashed line indicates the midway point between the most resistant and the most sensitive cell line for every compound, with cell lines to the left of the dashed line deemed drug-sensitive and cell lines to the right deemed drug-resistant. Each cell line is considered a genotype-specific biological replicate: KRASG12C (blue), n = 4 cell lines; KRASG12D (red), n = 3 cell lines; KRASG13C (green), n = 2 cell lines; and KRASG13D (purple), n = 4 cell lines. The dot represents the mean of the technical duplicate. Black arrows indicate compounds of interest. B, 2D (left) and 3D (right) viability of advanced cancer cell lines expressing either KRASG12C (H358, H23, HCC1171, H2030, H1792, and HOP62), KRASG12D (SKLU-1, A427, and HCC461), KRASG13C (H1734, H1355, HCC4087, and MOR), or KRASG13D (H647, H1944, HCC515, and HCC1974) exposed to 1 μmol/L docetaxel, 1 μmol/L paclitaxel, 1 μmol/L vincristine, 1 μmol/L epirubicin, 1 μmol/L gemcitabine, and 100 μmol/L carboplatin. Viability was measured by CellTiter-Glo (3D) 72 hours later. Data are expressed as a percentage of DMSO control. n = 3 per cell line with the mean of three biological repeats shown as a dot. Error bars represent mean ± SEM of all cell lines within each condition (KRASG12C + KRASG12D + KRASG13D vs. KRASG13C). Two KRASG12C lines (H2030 and HCC1171) were excluded from vincristine response in 3D analysis due to a viability increase of more than 180% relative to DMSO control. C, PFS between patients with KRASG13C and those with KRASG12C, KRASG12D, KRASG12V, or KRASG13D NSCLC receiving first-line chemotherapy. Data are taken from the Francophone de Cancérologie Thoracique dataset (51). D, Dose-response analysis of isogenic MLE-12 cells previously treated with DOX for 24 hours to induce KRAS variant expression in 0.5% serum, followed by a 72-hour treatment with a concentration range of indicated chemotherapeutics. Cells were plated in ULA conditions. Viability was analyzed by CellTiter-Glo 3D. Data are expressed as a percentage of DMSO control. n = 3 biological repeats per isogenic cell line; mean ± SEM. IC values were calculated for each cell line and presented as a table. E, Apoptosis quantification of all cell lines used in B (except H1792) after 72 hours of exposure to either 100 nmol/L docetaxel or 100 nmol/L paclitaxel. Cells were stained with FITC-conjugated annexin V to detect apoptotic cells. n = 3 biological repeats per cell line with the mean of three biological repeats shown as a dot. Error bars represent mean ± SEM of all cell lines within each condition. F, Viability of advanced cancer cell lines expressing (left) KRASG13C and (right) KRASG12C, KRASG12D, and KRASG13D grouped together, exposed to 1 μmol/L vincristine or either 100 or 1,000 nmol/L navitoclax or both compounds in combination. Viability was measured by CellTiter-Glo (3D) 72 hours later. Data are expressed as a percentage of DMSO control. n = 3 per cell line with the mean of 2 or 3 biological repeats shown as a dot. Error bars represent mean ± SEM of all cell lines within each condition. G, Viability of advanced cancer cell lines expressing (left) KRASG13C and (right) KRASG12C, KRASG12D, and KRASG13D grouped together exposed to 1 μmol/L epirubicin or either 100 or 1,000 nmol/L navitoclax or both compounds in combination. Viability was measured by CellTiter-Glo (3D) 72 hours later. Data are expressed as a percentage of DMSO control. n = 3 per cell line with the mean of 2 or 3 biological repeats shown as a dot. Error bars represent mean ± SEM of all cell lines within each condition. Statistical significance between comparisons was calculated using an unpaired Student t test (B), log-rank (Mantel–Cox; C), a two-way ANOVA (E), and a one-way ANOVA (F and G). COMB, combination. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05; ns > 0.05.

Supplementary Figure S1: Analysis of prognostic variables and genomic features in KRAS-mutant NSCLC patient datasets. Supplementary Figure S2: The impact of co-mutations on immunotherapy outcomes in KRASG13X NSCLC patients. Supplementary Figure S3: KRAS-mutant isoform-specific immune signatures from TCGA Firehose Legacy dataset. Supplementary Figure S4: Analysis of KRASG13X oncogenicity using isogenic models. Supplementary Figure S5: KRAS-mutant NSCLC cell line panel co-mutations and growth rate analyses. Supplementary Figure S6: KRAS-mutant NSCLC cell line panel signaling. Supplementary Figure S7: Analysis of (K)RAS (ON) inhibition in NSCLC models. Supplementary Figure S8: KRASG13C-driven NSCLC is selectively vulnerable to chemotherapy. Supplementary Figure S9: Combination of RMC-8839 and docetaxel is synergistic in KRASG13C-driven NSCLC models.

Cell line viability responses to compound library screen

TCGA LUAD immune gene signatures based on KRAS-mutant genotype

List of compound library screen drugs