Sotorasib combined with 3-methyladenine for the treatment of KRAS G12C-mutant pancreatic cancer and its underlying mechanisms Authors: Yukun Li, Mengmeng Zhou, Jingjing Pan, Haochen Su, Mengke Song, Zujiao Guan, Danmai Zhao, Xiaoxin Zhao, Yiqi Zhao, Pu Chen, Shu Zhang, Ying Lv Published: 2026 ABSTRACT Highlights•We first systematically evaluated the antitumor activity of Sotorasib in KRAS G12C mutant pancreatic cancer through both in vitro and in vivo experiments, and clarified the therapeutic potential of this drug in pancreatic cancer.•We first demonstrate that Sotorasib triggers early protective autophagy in KRAS G12C mutant pancreatic cancer cells, highlighting autophagy as a key mediator of Sotorasib-induced cell death.•We show that combining Sotorasib with the early autophagy inhibitor 3-Methyladenine further enhances its antitumor efficacy, supporting this combination strategy as a promising therapeutic option for KRAS G12C mutant pancreatic cancer.Keywords: Pancreatic cancer, AMG510, 3-MA, Autophagy, Apoptosis Introduction Pancreatic cancer is one of the most common malignancies worldwide. Recent data show that its incidence and mortality rank 12th and 6th globally, respectively, and are continuing to increase. The 5-year survival rate of patients with pancreatic cancer is extremely low, approximately 10%, indicating that it poses a serious threat to human health [[1], [2]]. Because of the lack of obvious early symptoms and specific diagnostic biomarkers, the majority of patients have unresectable tumors at diagnosis [3]. Although certain advances have been made in imaging diagnosis, surgical techniques, and neoadjuvant therapies (such as radiotherapy, chemotherapy, and immunotherapy) in recent years, these advances have not significantly improved the prognosis [[4], [5]]. Currently, the treatment for pancreatic cancer relies mainly on cytotoxic chemotherapy. Standard regimens include gemcitabine plus nab-paclitaxel and the FOLFIRINOX regimen. While some patients achieve short-term disease control under these regimens, the overall treatment efficacy remains unsatisfactory and is characterized by a short median progression-free survival, short overall survival, and severe toxicity [[6], [7], [8]]. Therefore, further elucidating the mechanisms of pancreatic cancer initiation and progression, identifying new therapeutic targets, and developing or optimizing more effective treatment strategies are particularly important. In pancreatic cancer, KRAS mutations represent one of the most characteristic molecular events; these mutations are detected in approximately 90% of patients with pancreatic ductal adenocarcinoma and are frequently observed in various solid tumors [[9], [10], [11]]. In pancreatic ductal adenocarcinoma, KRAS G12D and KRAS G12V are the predominant mutational subtypes. Among patients with KRAS-mutant tumors, KRAS G12D accounts for approximately 39%–41% and KRAS G12V for approximately 31%–32%, whereas KRAS G12C is a relatively rare subtype, accounting for only approximately 1%–2% of KRAS-mutant pancreatic cancers[14]. Because the KRAS protein lacks classical small-molecule binding pockets and has very high affinity for guanosine triphosphate (GTP), KRAS has long been considered an “undruggable” target [[12], [13]]. In recent years, the development of covalent small-molecule inhibitors has enabled major breakthroughs in KRAS-targeted therapy. In particular, the identification of the KRAS G12C mutation provides an opportunity for the direct pharmacological inhibition of KRAS. This mutation involves the substitution of glycine (Gly) at codon 12 with cysteine (Cys), the latter of which contains a unique thiol group that can be specifically recognized and covalently bound by small molecules [14]. KRAS G12C inhibitors function by locking KRAS in the guanosine diphosphate (GDP)-bound state, thereby preventing the activation of the downstream MAPK/ERK and PI3K–AKT–mTOR pathways and suppressing malignant proliferation [[15], [16]]. AMG510, the first FDA-approved KRAS G12C inhibitor, has shown favorable efficacy in solid tumors such as non-small cell lung cancer [[17], [18], [19]]. In a phase III randomized controlled trial (NCT04303780) enrolling 345 patients with advanced non-small cell lung cancer harboring the KRAS G12C mutation, AMG510 treatment achieved an objective response rate (ORR) of 28.1%, which was significantly higher than that in the control group, thereby validating its clinical effectiveness [20]. Clinical trials of AMG510 in patients with KRAS G12C-mutant pancreatic cancer are also underway. In a multicenter, open-label phase I/II trial (NCT03600883) involving 38 patients with locally advanced or metastatic pancreatic cancer harboring the KRAS G12C mutation, AMG510 treatment resulted in an ORR of 21% (8/38) and a median overall survival (OS) of 6.9 months, with no treatment-related deaths or treatment discontinuations due to treatment-related adverse events [21]. In patients with colorectal cancer, the objective response rate to AMG510 monotherapy is approximately 10%. Although AMG510 shows certain activity in patients for whom multiple prior therapies failed, its overall efficacy is limited. Further studies revealed that combining AMG510 with anti-EGFR antibodies significantly improved the objective response rate and progression-free survival, suggesting that the bypass of EGFR activation may represent a critical resistance mechanism limiting the efficacy of KRAS G12C inhibitors administered as a monotherapy [19]. At present, clinical trials evaluating AMG510 in combination with EGFR antibodies or immune checkpoint inhibitors are ongoing, with multiple studies aiming to increase AMG510 efficacy by blocking bypass activation and relieving immunosuppression [[22], [23], [24]]. Nevertheless, the mechanisms by which AMG510 suppresses the growth of KRAS G12C-mutant pancreatic cancer remain unclear. Moreover, whether AMG510 affects tumor drivers or signaling pathways in addition to KRAS mutations is still unknown. Therefore, elucidating the molecular mechanisms by which AMG510 regulates tumor biology in pancreatic cancer is important and represents an urgent scientific problem that remains to be addressed. AMG510 inhibits the proliferation of KRAS G12C-mutant pancreatic cancer cells and induces cell cycle arrest We first validated the mutation status of the KRAS gene by performing Sanger sequencing to ensure the genetic background of the cell models used in the drug sensitivity assays. As shown in Fig. 1A, both the pancreatic cancer cell line MiaPaCa-2 and the non-small cell lung cancer cell line H358 harbored the KRAS c.34G>T (p.G12C) mutation. Consistent base peaks at codon 12 in both forward and reverse chromatograms confirmed the presence and zygosity of the G12C mutation, establishing a molecular basis for the subsequent investigation of AMG510-mediated targeted inhibition (Fig. 1A). We first assessed the cytotoxicity of AMG510 in two KRAS G12C-mutant cell lines, H358 and MiaPaCa-2, using a CCK-8 assay to evaluate whether AMG510 affects the biological behavior of pancreatic cancer cells harboring the KRAS G12C mutation. CCK-8 assays showed that treatment with AMG510 for 48 and 72 h significantly suppressed the proliferation of both MiaPaCa-2 and H358 cells in a dose-dependent manner (Fig. 1B). The half-maximal inhibitory concentration (IC₅₀) of AMG510 in Mia PaCa-2 cells was 134.1 nM at 48 h and 71.0 nM at 72 h, whereas in H358 cells, the IC₅₀ values were 379.1 nM and 99.84 nM, respectively. A colony formation assay was subsequently performed to further evaluate the inhibitory effects of AMG510 on the proliferation of KRAS G12C-mutant cell lines (Fig. 1C). In MiaPaCa-2 cells, the number of colonies gradually decreased after treatment with increasing concentrations of AMG510, from 100.00 ± 4.44 colonies per well in the control group to 82.96 ± 6.42, 55.56 ± 3.85 and 16.30 ± 2.57 colonies per well at 50 nM, 100 nM and 500 nM, respectively. A similar trend was observed for H358 cells, with colony numbers decreasing from 100.00 ± 14.18 colonies per well in the control group to 75.00 ± 6.40, 41.94 ± 2.79 and 28.23 ± 3.70 colonies per well following treatment with 50 nM, 100 nM and 500 nM AMG510, respectively. These results demonstrate that AMG510 significantly inhibits the colony-forming ability of KRAS G12C-mutant cells in a dose-dependent manner. AMG510 suppresses the proliferation of KRAS G12C-mutant cells and induces cell cycle arrest. (A) Sanger sequencing confirmed that the pancreatic cancer cell line MiaPaCa-2 and the non-small cell lung cancer cell line H358 both harbor the KRAS codon 12 c.34G>T (p.G12C) mutation. (B) Dose–response curves and corresponding IC₅₀ values were determined using a CCK-8 assay after the treatment of MiaPaCa-2 and H358 cells with increasing concentrations of AMG510 for 48 h and 72 h. (C) Colony formation assays were performed to evaluate the effect of AMG510 on the clonogenic potential of MiaPaCa-2 and H358 cells. (D) Flow cytometry analysis of the cell cycle distribution of MiaPaCa-2 and H358 cells following treatment with varying concentrations of AMG510. (E) western blot analysis of the levels of cell cycle-related proteins (CDK2, Cyclin E, and Cyclin D3) in MiaPaCa-2 and H358 cells treated with the indicated concentrations of AMG510. β-Actin was used as a loading control. (F) Sanger sequencing validation revealed that PANC-1 cells carry the KRAS codon 12 c.35G>A (p.G12D) mutation. (G) CCK-8 assays were performed to evaluate the viability of PANC-1 cells at 48 h and 72 h after treatment with different concentrations of AMG510. (H) Colony formation assays were performed to assess the effects of different concentrations of AMG510 on the clonogenic capacity of PANC-1 cells. *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. Given that cell cycle arrest is a major mechanism contributing to the inhibition of cell proliferation, we next examined the effects of AMG510 on the cell cycle distribution of H358 and MiaPaCa-2 cells using flow cytometry (Fig. 1D). In untreated MiaPaCa-2 cells, the proportions of cells in G1/S, S and G2/M phases were 33.65 ± 1.91%, 46.00 ± 2.83% and 20.35 ± 0.92%, respectively. Upon treatment with 50 nM AMG510, the G1/S-phase population significantly increased to 56.55 ± 0.78%, which was accompanied by decreases in the S-phase population to 31.10 ± 1.27% and the G2/M-phase population to 12.50 ± 0.71%. Following treatment with 200 nM AMG510, the percentage of cells in G1/S phase further increased to 60.20 ± 2.55%, the percentage of cells in S phase was 33.70 ± 2.40%, and the percentage of cells in G2/M phase decreased further to 6.10 ± 0.14%. A similar trend was observed in H358 cells. In the untreated H358 cells, the distributions were 46.64 ± 2.34% of cells in G1/S phase, 38.78 ± 0.59% of cells in S phase and 14.58 ± 1.75% of cells in G2/M phase. Following treatment with 50 nM AMG510, the percentage of cells in G1/S phase increased to 50.81 ± 2.05%, the percentage of cells in S phase increased slightly to 39.04 ± 1.74%, and the percentage of cells in G2/M phase decreased significantly to 10.16 ± 0.30%. After treatment with 200 nM AMG510, the percentage of cells in G1/S phase further increased to 52.19 ± 1.44%, the percentage of cells in S phase was 40.27 ± 1.01%, and the percentage of cells in G2/M phase decreased to 7.54 ± 0.43%. These results indicate that AMG510 significantly increases the proportion of KRAS G12C-mutant cells in G1/S phase while reducing the proportion of cells entering G2/M phase, suggesting that AMG510 may inhibit the proliferation of KRAS G12C-mutant cells by inducing cell cycle arrest. These findings were further supported by western blotting, which showed that AMG510 reduced the expression levels of multiple cell cycle-related proteins, including CDK2, Cyclin E and Cyclin D3, in KRAS G12C-mutant cells (Fig. 1E). We further validated the specificity of AMG510 for the KRAS G12C mutation by performing the same assays in the pancreatic cancer cell line PANC-1, which does not harbor the G12C mutation. Sanger sequencing confirmed that PANC-1 cells carry the KRAS codon 12 c.35G>A (p.G12D) mutation (Fig. 1F). The results of the CCK-8 assay indicated that compared with no treatment, AMG510 treatment did not significantly decrease the viability of PANC-1 cells (Fig. 1G). Consistent with these findings, colony formation assays revealed no substantial reduction in clonogenic potential upon AMG510 exposure (Fig. 1H). In the untreated group, the number of colonies was 100.00 ± 14.55 per well; following treatment with 50 nM, 100 nM and 500 nM AMG510, the colony numbers were 92.12 ± 13.24, 87.27 ± 17.38 and 89.70 ± 2.10 per well, respectively, with no statistically significant differences observed (P > 0.05). Collectively, these results indicate that AMG510 selectively induces cell cycle arrest in G1/S phase in KRAS G12C-mutant pancreatic cancer cells in vitro, which may contribute to its antiproliferative effect in this genetic context. AMG510 promotes apoptosis and induces mitochondrial dysfunction in KRAS G12C-mutant pancreatic cancer cells We further investigated whether AMG510 induces apoptosis in KRAS G12C-mutant cell lines by performing annexin V/propidium iodide (PI) double staining followed by flow cytometry (FCM) to evaluate apoptotic cell populations (Fig. 2A). In MiaPaCa-2 cells, the percentage of apoptotic cells increased from 6.56 ± 1.22% in the control group to 10.02 ± 0.34%, 11.33 ± 0.21% and 12.75 ± 0.76% following treatment with 50 nM, 100 nM and 500 nM AMG510, respectively. A similar trend was observed in H358 cells, where the percentage of apoptotic cells increased from 2.59 ± 0.17% in the control group to 4.72 ± 0.07%, 6.13 ± 0.30% and 7.17 ± 0.13% in groups treated with the same concentrations of AMG510. These results indicate that AMG510 significantly induces apoptosis in KRAS G12C-mutant cells in a concentration-dependent manner. Western blotting further revealed that AMG510 upregulated the expression of the pro-apoptotic proteins cleaved PARP and Bax but downregulated the expression of the anti-apoptotic protein Bcl-2 in KRAS G12C-mutant cells (Fig. 2B). These results suggest that AMG510 specifically induces apoptosis in MiaPaCa-2 and H358 cells in vitro. AMG510 promotes apoptosis and induces mitochondrial dysfunction in KRAS G12C-mutant pancreatic cancer cells. (A) MiaPaCa‑2 and H358 cells were treated with the indicated concentrations of AMG510 for 48 h. Early and late apoptotic cells were quantified by flow cytometry using Annexin V‑FITC/PI staining. (B) western blot analysis of the levels of apoptosis‑related proteins (Bax, Bcl‑2, and cleaved PARP) in MiaPaCa-2 and H358 cells following treatment with different concentrations of AMG510. β‑Actin was used as a loading control. (C) PANC-1 cells were treated with the indicated concentrations of AMG510 for 48 h. Early and late apoptotic cells were quantified by flow cytometry using Annexin V-FITC/PI staining. (D) The mitochondrial membrane potential was evaluated in MiaPaCa-2 cells using a JC‑1 fluorescent probe and flow cytometry after 48 h of treatment with the indicated concentrations of AMG510. (E) Intracellular ROS levels were measured by flow cytometry using a DCFH-DA fluorescent probe following treatment with various concentrations of AMG510. The mean fluorescence intensity was quantified. (F) Fluorescence microscopy was performed to visualize ROS accumulation in MiaPaCa-2 cells treated with different concentrations of AMG510. The fluorescence intensity was quantified. *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. Apoptosis was also assessed in the KRAS G12D-mutant pancreatic cancer cell line PANC-1 by performing Annexin V/PI staining and flow cytometry to further validate the specificity of AMG510 (Fig. 2C). The percentage of apoptotic cells in the control group was 6.97 ± 0.50%, whereas treatment with 50 nM, 100 nM and 500 nM AMG510 resulted in 6.79 ± 0.15%, 6.39 ± 0.42% and 6.31 ± 0.16% apoptotic cells, respectively. No statistically significant differences were observed compared with the untreated control group (P > 0.05), indicating that AMG510 did not induce apoptosis in PANC-1 cells. We investigated whether AMG510 induces apoptosis by disrupting mitochondrial homeostasis in KRAS G12C-mutant pancreatic cancer cells by measuring the mitochondrial membrane potential (MMP) in MiaPaCa-2 cells using the JC-1 probe. JC-1 aggregates within mitochondria with a high MMP, emitting red fluorescence, whereas a loss of the MMP yields monomeric JC-1 that emits green fluorescence; thus, the red-to-green fluorescence ratio reflects mitochondrial depolarization. After the cells were incubated with JC-1, flow cytometry was used to quantify the proportion of JC-1 monomers, reflecting changes in the MMP. The proportion of JC-1 monomers increased after AMG510 treatment: 2.81 ± 0.51% in the control group and 23.83 ± 6.25%, 33.77 ± 1.39% and 37.50 ± 7.11% in the 50 nM, 100 nM and 500 nM groups, respectively. These results indicated that AMG510 disrupted mitochondrial homeostasis in MiaPaCa-2 cells (Fig. 2D). Given that mitochondrial dysfunction is often accompanied by elevated ROS levels, which can trigger oxidative stress, apoptosis and inflammatory responses, we next examined whether AMG510 induces ROS accumulation in KRAS G12C-mutant pancreatic cancer cells. Intracellular ROS levels were measured using the ROS-sensitive probe DCFH-DA, and the mean fluorescence intensity (MFI) was analyzed using flow cytometry (Fig. 2E). The MFI increased in MiaPaCa-2 cells after AMG510 treatment; it was 1166.7 ± 31.0 in the control group and increased to 1507.3 ± 55.5, 1525.0 ± 76.7 and 1942.3 ± 234.3 in the 50 nM, 100 nM and 500 nM groups, respectively. Confocal imaging (Fig. 2F) showed a similar pattern, with mean fluorescence intensities of 13.33 ± 1.98 in the control group that increased to 42.35 ± 8.07, 61.34 ± 13.62 and 53.46 ± 5.88 in the respective treatment groups. Collectively, these findings indicate that AMG510 disrupts mitochondrial homeostasis in KRAS G12C-mutant pancreatic cancer cells by inducing a loss of the mitochondrial membrane potential and the accumulation of ROS, thereby promoting apoptosis. AMG510 effectively inhibits the proliferation of KRAS G12C-mutant pancreatic cancer cells in vivo We established subcutaneous xenograft and orthotopic pancreatic cancer models in nude mice and conducted the corresponding in vivo experiments to further validate the findings from the in vitro experiments described above. First, a subcutaneous xenograft model was generated by subcutaneously injecting MiaPaCa-2 cells, and the mice were randomly assigned to a control group or an AMG510-treated group (n = 6 mice per group). AMG510 was administered for 21 consecutive days, and body weight and tumor volume were monitored at regular intervals to evaluate antitumor efficacy and systemic toxicity. At the end of treatment, the mice were euthanized, and the tumor tissues were harvested for subsequent analyses. Compared with the control group, the AMG510-treated group exhibited significantly reduced tumor volume and tumor weight, with concordant decreases in both measures (Fig. 3A–C). AMG510 effectively suppresses the growth of subcutaneous and orthotopic KRAS G12C-mutant pancreatic tumors in vivo. (A) Twelve mice were randomly assigned to two groups (n = 6 mice per group): control (DMSO) and AMG510 (10% AMG510 + 90% corn oil; 30 mg/kg/day). Representative images of the tumors after 21 consecutive days of oral gavage of DMSO or AMG510. (B) At the end of treatment, subcutaneous tumor weights were measured and compared between groups. (C) Body weights were monitored during the treatment period and compared between groups. (D) Subcutaneous tumor volumes were measured during the treatment period and compared between groups. (E) H&E staining and immunohistochemistry were performed on subcutaneous tumor samples to evaluate the levels of Ki-67, cleaved caspase-3, and LC3B. (F) At the end of treatment, the effects of AMG510 on complete blood counts and hepatic/renal function were assessed. (G) At the end of treatment, H&E staining was performed to evaluate histopathological changes in the heart, liver, spleen, lungs, and kidneys. (H) Ten mice were randomly assigned to one of two groups (n = 5 mice per group): control (DMSO) or AMG510 (10% AMG510 + 90% corn oil; 30 mg/kg/day). Representative images of orthotopic pancreatic tumors after 21 consecutive days of oral gavage of DMSO or AMG510. (I) Body weights were monitored during the treatment period and compared between groups. (J) At the end of treatment, orthotopic tumor weights were measured and compared between groups. (K) H&E staining and immunohistochemistry were performed on orthotopic tumor specimens to evaluate the levels of Ki-67, cleaved caspase-3, and LC3B. (L) At the end of treatment, the effects of AMG510 on complete blood counts and hepatic/renal function were assessed. (M) At the end of treatment, H&E staining was performed to evaluate histopathological changes in the heart, liver, spleen, lungs, and kidneys. *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. We further established an orthotopic pancreatic cancer model using MiaPaCa-2 cells to more faithfully recapitulate the primary pancreatic lesion and applied the same group allocation and dosing scheme for in vivo validation. Upon completion of the treatment cycle, body weight, tumor weight, and tissue specimens were systematically evaluated to further confirm the in vivo antitumor activity of AMG510. AMG510 also displayed significant antitumor activity in the orthotopic model, with significantly lower tumor weight and tumor volume in the treated group than in the control group (Fig. 3H–I), consistent with the findings from the in vitro and subcutaneous xenograft models. Immunohistochemical analyses further showed that compared with control treatment, AMG510 treatment significantly upregulated LC3B and cleaved caspase-3 levels, whereas the expression of the proliferation marker Ki-67 was markedly reduced in both xenograft and orthotopic tumors (Fig. 3E and K). Throughout the experiments, in vivo safety of AMG510 was assessed by continuously monitoring body weight changes in both the xenograft and orthotopic model mice. No significant differences in body weight were observed between the treatment and control groups (Fig. 3D and J). In addition, complete blood counts and indices of hepatic/renal function did not differ significantly between the groups (Fig. 3F and L). H&E staining of major organs, including the heart, liver, kidneys, and lungs, revealed no overt histopathological abnormalities (Fig. 3G and M), indicating a favorable in vivo biosafety profile for AMG510. Collectively, these in vivo findings demonstrate the significant antitumor activity of AMG510 in KRAS G12C-mutant pancreatic cancer models, which was characterized by the suppression of tumor growth, decreased proliferative activity, and the activation of autophagy- and apoptosis-related pathways without evident toxicity, indicating that AMG510 has promising therapeutic potential and a favorable biosafety profile. AMG510 induces early autophagy in KRAS G12C-mutant pancreatic cancer cells Previous studies have shown that AMG510 monotherapy has limited efficacy in patients with colorectal cancer, whereas AMG510 combined with EGFR inhibitors exhibits significantly increased efficacy, indicating that the antitumor effects of AMG510 monotherapy on certain KRAS G12C-mutant solid tumors may be constrained by resistance mechanisms or the activation of alternative signaling pathways [19]. We further investigated the mechanism of AMG510 in KRAS G12C-mutant cells and whether cell death pathways other than apoptosis are involved by performing transcriptome sequencing of MiaPaCa-2 cells treated with AMG510. A total of 3114 differentially expressed genes (DEGs) were identified at an FDR < 0.05, including 2021 upregulated genes and 1093 downregulated genes. Gene set enrichment analysis (GSEA) revealed the predominant enrichment of cell cycle-related pathways, particularly those associated with cell cycle arrest (Fig. 4A), consistent with our earlier results showing AMG510-induced G1/S phase arrest. Notably, the expression of several autophagy-related genes also significantly changed following AMG510 treatment (Fig. 4A). The GO enrichment analysis revealed that the “macroautophagy” pathway was significantly enriched (Fig. 4B). Based on these observations, we further examined the regulatory effect of AMG510 on autophagic activity in KRAS G12C-mutant pancreatic cancer cells. AMG510 induces autophagosome accumulation in KRAS G12C-mutant cancer cells. (A) Transcriptomic profiling indicated differentially expressed genes in MiaPaCa-2 cells following treatment with 100 nM AMG510. (B) The Gene Ontology (GO) enrichment analysis revealed the significant enrichment of macroautophagy. (C) Transmission electron microscopy (TEM) was used to visualize autophagosomes in MiaPaCa-2 and H358 cells after 24 h of treatment with 100 nM AMG510. Red arrows indicate autophagosomes. (D) western blot analysis of LC3B-I, LC3B-II and p62 levels in MiaPaCa-2 and H358 cells treated with the indicated concentrations of AMG510; the LC3B-II/LC3B-I ratio was quantified. (E) MiaPaCa-2 and H358 cells were treated with DMSO, 100 nM AMG510, 10 μM chloroquine (CQ), or AMG510 (100 nM) plus CQ (10 μM) for 24 h. LC3B-I, LC3B-II and p62 expression were analyzed using western blotting, and the LC3B-II/LC3B-I ratio was calculated. (F) Under the same conditions, the levels of early autophagy-related proteins, including Beclin-1, p-mTOR and mTOR, in MiaPaCa-2 and H358 cells were evaluated by western blotting. ⁎P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. We assessed whether AMG510 induces characteristic morphological features of autophagy by analyzing MiaPaCa-2 and H358 cells after 24 h of treatment using transmission electron microscopy (TEM) (Fig. 4C). Compared with control cells, AMG510-treated cells displayed a marked increase in the number of cytoplasmic double- or multilamellar vesicles that frequently enclosed damaged organelles such as mitochondria and endoplasmic reticulum—features consistent with those of typical autophagosomes (indicated by red arrows). In addition, some single-membrane vacuoles were observed, likely representing autolysosomes (also indicated by red arrows). These findings suggest that AMG510 promotes autophagosome formation in KRAS G12C-mutant cells, although the underlying mechanisms warrant further investigation. Previous studies have established LC3B and p62 as markers of distinct stages of autophagy, representing autophagosome formation and substrate degradation, respectively. Upon autophagy initiation, cytosolic LC3B-I undergoes lipidation to form membrane-bound LC3B-II, whose abundance indicates autophagosome formation. Conversely, p62 is selectively incorporated into autophagosomes and subsequently degraded; therefore, its expression inversely correlates with autophagic activity. Accordingly, a combined analysis of the LC3B-II/LC3B-I ratio and p62 level is typically used to dynamically evaluate autophagic flux. We next analyzed the expression of key autophagy markers, including p62, LC3B-I, LC3B-II and the LC3B-II/LC3B-I ratio, in MiaPaCa-2 and H358 cells treated with AMG510 using western blotting. The results showed that AMG510 increased the LC3B levels and decreased the P62 levels in both cell lines, suggesting that AMG510 can induce autophagy in KRAS G12C-mutant cells in vitro in a dose-dependent manner (Fig. 4D). We introduced the late-stage autophagy inhibitor chloroquine (CQ), which blocks autophagosome–lysosome fusion and thereby inhibits the degradation phase of autophagy, to further investigate the regulatory mechanism. Western blotting revealed that compared with either treatment alone, treatment with the combination of CQ and AMG510 markedly increased the LC3B-II/LC3B-I ratio and p62 protein levels in MiaPaCa-2 and H358 cells (Fig. 4E). These findings suggest that AMG510 primarily activates the early stage of autophagy rather than promoting autophagosome degradation. Based on these data, we examined key regulators involved in autophagy initiation. AMG510 treatment reduced p-mTOR and total mTOR levels, accompanied by the upregulation of Beclin-1 expression (Fig. 4F). These changes are consistent with the early-stage activation of autophagy by AMG510, suggesting that AMG510 may mediate autophagy initiation through mTOR inhibition and Beclin-1 activation. We performed immunofluorescence staining to assess the expression and subcellular localization of LC3 as a method to validate the effect of AMG510 on the early stage of autophagy. In DMSO-treated control cells, LC3 was diffusely distributed in the cytoplasm, whereas AMG510-treated cells exhibited a marked accumulation of LC3-positive puncta throughout the cytoplasm. The quantitative analysis revealed a significant increase in the number of LC3 puncta per cell in the AMG510 group (13.5 ± 0.5 puncta per cell) compared with that in the DMSO group (3.0 ± 1.0 puncta per cell). Moreover, cotreatment with AMG510 and CQ further increased the number of LC3 puncta (21.5 ± 1.5 puncta per cell) compared with treatment with CQ alone (16.0 ± 1.0 puncta per cell), and increased the fluorescence intensity (Fig. 5A, B), suggesting that the cotreatment exacerbated LC3 accumulation, potentially through increased autophagosome formation. These observations were consistent with the increased LC3B-II levels observed using western blotting. AMG510 induces early-stage autophagy in KRAS G12C-mutant cells without blocking autophagic flux. (A) MiaPaCa-2 and H358 cells were treated with DMSO, 100 nM AMG510, 10 μM CQ, or a combination of AMG510 (100 nM) and CQ (10 μM) for 24 h. Intracellular LC3B expression was assessed by performing immunofluorescence staining. (B) Quantification of the number of LC3B puncta per cell under the indicated treatment conditions. (C) MiaPaCa-2 cells were transduced with an mRFP-GFP-LC3 adenoviral vector and treated with DMSO, 100 nM AMG510, 10 μM CQ, or a combination of AMG510 (100 nM) and CQ (10 μM) for 24 h. Autophagic flux was evaluated by fluorescence microscopy. (D) H358 cells were transduced with an mRFP-GFP-LC3 adenoviral vector and treated with DMSO, 100 nM AMG510, 10 μM CQ, or a combination of AMG510 (100 nM) and CQ (10 μM) for 24 h. Autophagic flux was evaluated by fluorescence microscopy. *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. Given that autophagic flux is a dynamic process involving autophagosome formation, autolysosome generation and final substrate degradation, we transfected MiaPaCa-2 and H358 cells with an mRFP–GFP–LC3 adenovirus to monitor autophagic flux in real time. This dual-fluorescence system leverages the susceptibility of GFP to quenching in the acidic lysosomal environment; thus, unfused autophagosomes appear as yellow puncta (overlapping red and green), whereas autolysosomes formed after lysosomal fusion only display red fluorescence. Therefore, this system distinguishes different stages of the autophagic process. After 48 h of transfection, we quantified red and green puncta using confocal microscopy (Fig. 5C, D). Compared with the DMSO control treatment, AMG510 treatment led to a marked increase in red fluorescence and a concomitant reduction in green fluorescence, suggesting increased autophagosome formation and efficient fusion with lysosomes. In the CQ-alone group, both red and green fluorescence intensities were increased, reflecting autophagosome accumulation due to the blockade of autophagosome–lysosome fusion, consistent with the known mechanism of CQ as a late-stage autophagy inhibitor. In the AMG510+CQ group, both the red and green signals were further increased, accompanied by a significant increase in the number of yellow puncta, indicating that AMG510 promotes autophagosome formation without impeding autophagic flux. These observations, together with the western blotting and immunofluorescence data, support that AMG510 primarily modulates the early stage of autophagy rather than blocking the overall autophagic process. The role of autophagy in AMG510-induced cell death in pancreatic cancer cells harboring the KRAS G12C mutation Previous experiments have demonstrated that AMG510 induces autophagy in both MiaPaCa-2 and H358 cells in vitro, primarily by targeting the early phase of the autophagic process. We further investigated the biological effect of AMG510 on early-stage autophagy in pancreatic cancer cells harboring the KRAS G12C mutation by conducting functional inhibition assays using the early autophagy inhibitor 3-methyladenine (3-MA). Western blotting revealed that 3-MA markedly suppressed the AMG510-induced increase in the LC3B-II/LC3B-I ratio and prevented p62 degradation (Fig. 6A), indicating that 3-MA effectively blocked AMG510-mediated autophagic activation. Effects of AMG510 and 3-MA on autophagy, proliferation, migration and apoptosis in KRAS G12C-mutant cells. (A) MiaPaCa-2 and H358 cells were treated with DMSO, 100 nM AMG510, 1 mM 3-MA, or a combination of AMG510 (100 nM) and 3-MA (1 mM). The expression of LC3B-I, LC3B-II and p62 was analyzed using western blotting, and the LC3B-II/LC3B-I ratio was calculated. (B) The viability of MiaPaCa-2 and H358 cells was assessed by performing a CCK-8 assay after 48 h of treatment with various concentrations of AMG510, either alone or in combination with 1 mM 3-MA. (C) MiaPaCa-2 cells were treated with DMSO, 100 nM AMG510, 1 mM 3-MA, or a combination of AMG510 (100 nM) and 3-MA (1 mM) for 14 days. Colony formation assays were performed to evaluate the clonogenic capacity of cells under each condition. (D) MiaPaCa-2 cells were subjected to the same treatment conditions for 72 h, and wound-healing assays were conducted to evaluate relative cell migration. (E) The apoptosis of MiaPaCa-2 cells was assessed using Annexin V-FITC/PI double staining and flow cytometry following 72 h of treatment with DMSO, 100 nM AMG510, 1 mM 3-MA, or the combination of AMG510 (100 nM) and 3-MA (1 mM). *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. We performed CCK-8 assays to determine whether the inhibition of early-stage autophagy affects the viability of KRAS G12C-mutant cells. Compared with treatment with AMG510 alone, cotreatment with AMG510 and 3-MA further reduced the viability of both MiaPaCa-2 and H358 cells (Fig. 6B), suggesting that AMG510-induced autophagy may act as a drug-responsive cytoprotective mechanism. Based on this result, we evaluated the effect of the combination treatment on proliferation using colony formation assays (Fig. 6C). The average number of MiaPaCa-2 colonies in the combination group was 19.67 ± 2.08 colonies per well, which was significantly lower than that in the AMG510 alone group (67.33 ± 3.51 colonies per well), indicating that autophagy inhibition enhances the suppressive effect of AMG510 on clonal expansion. In addition, we conducted wound-healing assays to assess the effect on cell migration (Fig. 6D). The relative migration rate of MiaPaCa-2 cells was further reduced to 7.93% ± 0.52% in the AMG510 + 3-MA group compared with 16.81% ± 0.97% in the AMG510 monotherapy group, suggesting that autophagy inhibition further enhances the antimigratory effect of AMG510. Apoptosis was further evaluated using flow cytometry (Fig. 6E). Cotreatment with AMG510 and 3-MA significantly increased the percentage of apoptotic MiaPaCa-2 cells to 28.37 ± 2.24%, compared with 14.43 ± 0.75% in the AMG510 monotherapy group, indicating that autophagy inhibition potentiates the proapoptotic effect of AMG510. Notably, the results of the flow cytometry analysis also suggested that the combination treatment may accelerate the transition from early to late apoptosis. Collectively, these findings indicate that AMG510 induces early-stage autophagy with cytoprotective effects on pancreatic cancer cells harboring the KRAS G12C mutation. The inhibition of this autophagic process further enhances the proapoptotic effect of the drug. These results suggest that targeting autophagy in combination with AMG510 may represent a potential strategy to improve its antitumor efficacy. AMG510 combined with 3-MA further suppresses the growth of KRAS G12C-mutant pancreatic cancer cells in vivo We further evaluated the efficacy of AMG510 combined with 3-MA in vivo by establishing a subcutaneous xenograft model in nude mice. MiaPaCa-2 cells were inoculated subcutaneously to generate tumors, and the mice were randomly assigned to one of the following four groups (n = 5 mice per group): vehicle control, AMG510, 3-MA, and AMG510 plus 3-MA. During 21 consecutive days of treatment, the body weight and tumor volume were monitored at regular intervals to assess antitumor activity and systemic tolerance. At the end of treatment, the mice were euthanized, and tumor tissues were collected for subsequent analyses. Compared with the control treatment, AMG510 treatment significantly reduced tumor volume and tumor weight. Moreover, compared with AMG510 alone, the combination of AMG510 and 3-MA resulted in additional decreases in tumor size and weight (Fig. 7A–C), indicating that 3-MA potentiates the antitumor activity of AMG510 in vivo. Treatment with the combination of AMG510 and 3-MA further suppressed the growth of subcutaneous KRAS G12C-mutant pancreatic tumors in nude mice. (A) Twenty nude mice were randomly assigned to the following four groups (n = 5 mice per group): control (DMSO), AMG510 (10% AMG510 + 90% corn oil; 30 mg/kg/day), 3-MA (25 mg/kg/day), and AMG510 + 3-MA. Representative images of subcutaneous xenograft tumors after 21 consecutive days of treatment. (B) The tumor volumes were measured during the treatment period and compared among the groups. (C) Body weights were monitored during the treatment period and compared among groups. (D) At the end of treatment, subcutaneous tumor weights were measured and compared among the groups. (E) H&E staining and immunohistochemistry were performed on subcutaneous tumor samples to evaluate the levels of Ki-67, cleaved caspase-3, and LC3B. (F) Sanger sequencing of DNA from subcutaneous tumor specimens. (G) Sanger sequencing of cDNA derived from RNA extracted from subcutaneous tumor samples. *P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎⁎P < 0.0001. Throughout the study, the body weights of mice in the AMG510 group (25.08 ± 0.69 g), 3-MA group (24.38 ± 0.54 g), and AMG510 plus 3-MA group (24.52 ± 0.58 g) did not differ significantly from those of mice in the control group (24.32 ± 0.72 g) (P > 0.05) (Fig. 7D), indicating good tolerability without evident systemic toxicity. Immunohistochemical staining further showed that, compared with the control treatment, AMG510 treatment decreased the expression of the proliferation marker Ki-67 and increased the expression of LC3B and cleaved caspase-3 in both subcutaneous and orthotopic tumors. Notably, compared with AMG510 alone, the combination of AMG510 and 3-MA suppressed LC3B expression while further reducing Ki-67 expression and increasing cleaved caspase-3 levels (Fig. 7E). Sanger sequencing was performed on genomic DNA and cDNA from the tumor specimens from each group to determine whether AMG510 alone or in combination with 3-MA affects the KRAS G12C mutation itself. Although AMG510 administered alone or in combination with 3-MA significantly suppressed tumor growth, the KRAS G12C mutation in the DNA remained stable in tumor tissues across groups (Fig. 7F). Subsequent cDNA Sanger sequencing likewise confirmed persistent expression of the KRAS G12C mutation at the transcript level in the tumor tissues from each group (Fig. 7G). These results indicate that the pharmacological intervention did not alter the presence of the KRAS G12C mutation itself but exerted antitumor effects primarily by inhibiting the function of the mutant KRAS protein and its downstream signaling pathways. Discussion Despite recent advances in therapeutic strategies for pancreatic cancer, including combination chemotherapy, molecular targeted therapy and immunotherapy, the overall survival benefit for patients has not significantly improved [[4], [5]]. This dilemma is, to some extent, attributed to the sustained activation of early driver genes such as KRAS, on which pancreatic cancer is highly dependent. KRAS not only lacks classical small-molecule binding sites but also mediates an extremely complex signaling network and has long been considered an effectively ‘undruggable’ target, severely limiting the development of targeted therapeutic strategies and impairing the synergistic effects of chemotherapy and immunotherapy [[12], [13]]. However, the emergence of AMG510 represents a pivotal breakthrough in the development of KRAS G12C inhibitors [[14], [15], [16]]. Existing studies have demonstrated that AMG510 has favorable efficacy against non-small cell lung cancer and colorectal cancer [[17], [18], [19]]. Although KRAS G12C is not a major mutational subtype in pancreatic ductal adenocarcinoma and accounts for only approximately 1%−2% of all pancreatic cancers, it represents a relatively rare molecular subgroup [14]. However, therapeutic options remain very limited for the specific subset of patients harboring KRAS G12C mutations. The KRAS G12C mutation results in the substitution of glycine with cysteine at codon 12, creating a thiol group on the protein surface that can be specifically targeted by covalent small molecules. Therefore, KRAS G12C represents a well-defined therapeutic target, and small-molecule targeted strategies offer the potential for precision treatment in this small patient population, with the promise of improving clinical benefit and prognosis. Accordingly, research focusing on this subtype remains of important clinical significance. Notably, its application in pancreatic cancer remains at an early exploratory stage. This study is the first to systematically assess the antitumor activity of AMG510 in KRAS G12C-mutant pancreatic cancer cells and its potential molecular mechanisms. These findings provide new experimental evidence supporting the application of KRAS G12C-targeted therapy in pancreatic cancer and offer insights into how its efficacy can be optimized and increased. Our experimental results indicate that AMG510 can induce cell cycle arrest in KRAS G12C-mutant pancreatic cancer cells, thereby suppressing cell proliferation. In parallel, the drug also triggers mitochondrial dysfunction and ROS accumulation, further promoting apoptosis. Moreover, we observed that AMG510 has limited effects on the non-KRAS G12C-mutant pancreatic cancer cell line PANC-1, confirming the high heterogeneity of its targeted action across different KRAS mutation types in pancreatic cancer cells. A subsequent transcriptomic analysis revealed that autophagy plays a crucial role in AMG510-mediated cell death in KRAS G12C-mutant pancreatic cancer cells. Following AMG510 treatment, the degradation of LC3B and p62 increased in MiaPaCa-2 cells, and the formation of autophagosomes increased, indicating that autophagy plays a critical role in the effects of the drug. Further studies revealed that when AMG510 was combined with CQ, the LC3B level increased further, suggesting that AMG510 primarily activates the early stages of autophagy in KRAS G12C-mutant pancreatic cancer cells. Autophagy, an evolutionarily conserved intracellular degradation mechanism, plays a dual role in maintaining cellular homeostasis and responding to stress. In normal cells, autophagy maintains cellular stability through the engulfment and degradation of damaged organelles, such as mitochondria and the endoplasmic reticulum, as well as abnormal proteins [[25], [26]]. However, in tumor cells, the role of autophagy is more complex and dual-faceted. On the one hand, tumor cells rely on autophagy to survive under harsh microenvironmental conditions, such as hypoxia, nutrient deprivation and drug-induced stress. On the other hand, increasing evidence suggests that autophagy can also help tumor cells evade host immune surveillance by modulating the expression of major histocompatibility complex molecules, antigen processing and apoptosis-related signaling pathways [[27], [28], [29]]. Previous studies have shown that inhibiting autophagy reduces the proportion of CD44+CD117+ tumor cancer stem cells in ovarian cancer, thereby increasing their resistance to chemotherapy and tumorigenic potential [30]. Oxaliplatin treatment also induces autophagy in colorectal cancer cells. When CQ is administered, Beclin-1 is downregulated, or ATG5 is activated to inhibit autophagy, the sensitivity of colorectal cancer to oxaliplatin is increased synergistically [31]. These results suggest that during tumor treatment, the activation of autophagy is not only a survival mechanism through which tumor cells adapt to drug treatment but also a key factor in the resistance to certain chemotherapies or targeted drugs [[32], [33], [34]]. Our functional experiments indicated that treatment with the combination of the early autophagy inhibitor 3-MA significantly increased the cytotoxic effect of AMG510 on KRAS G12C-mutant pancreatic cancer cells, suggesting that autophagy may serve as an adaptive mechanism through which KRAS G12C-mutant pancreatic cancer cells cope with drug stress. Moreover, autophagy inhibition may increase the sensitivity of cells to AMG510, thereby increasing its antitumor efficacy. These findings suggest that targeting autophagy could serve as a reliable adjunct strategy for AMG510-based treatment of KRAS G12C-mutant pancreatic cancer. Apoptosis is a form of programmed cell death that is typically divided into early and late apoptotic stages. The main features of the early apoptotic stage include phosphatidylserine (PS) externalization, mitochondrial dysfunction and chromatin condensation, with the integrity of the cell membrane remaining intact, making this stage potentially reversible. However, in the late apoptotic stage, the integrity of the cell membrane is compromised, DNA fragmentation occurs, and the cell irreversibly progresses to death [[35], [36], [37]]. Using flow cytometry, we further assessed the effects of AMG510 combined with 3-MA on apoptosis. The results showed that the total percentage of apoptotic cells in the combination treatment group was significantly higher than that in the single-drug groups, exceeding the sum of their individual effects and suggesting a synergistic proapoptotic effect of the two drugs. More importantly, the combination treatment significantly increased the proportion of late apoptotic cells, promoting the transition from early to late apoptosis. These findings indicate that the combination strategy not only increases the intensity of apoptotic signaling but also accelerates the irreversible progression of the apoptotic process, potentially increasing therapeutic efficiency and durability. These results support the therapeutic potential of combining early autophagy inhibition with AMG510 in KRAS G12C-mutant pancreatic cancer, although the exact underlying molecular mechanisms require further investigation. Furthermore, our cellular experiments indicate that autophagy represents a distinct phenotype induced by drug stimulation in KRAS G12C-mutant pancreatic cancer cells. When AMG510 is combined with early autophagy inhibitors for the treatment of KRAS G12C-mutant pancreatic cancer, a thorough evaluation of the expression levels of autophagy-related proteins in tumor tissue is critical prior to treatment to determine the optimal timing and dosage. With advancements in digestive endoscopy and microsequencing technologies, obtaining genetic information from pancreatic cancer samples via endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) has become feasible [[38], [39], [40]]. In the future, integrating the KRAS G12C mutation burden, autophagy expression levels and other important clinical diagnostic results to develop predictive models will be critical for guiding personalized clinical treatment. Notably, we observed that the antitumor effect of AMG510 was weaker in the orthotopic tumor model than in the subcutaneous tumor model. This difference may suggest that the unique tumor microenvironment of pancreatic cancer, such as a highly fibrotic stroma, low vascular density and an immunosuppressive state, may limit drug penetration and efficacy to some extent [[41], [42], [43], [44]]. These findings highlight the potential effect of the tumor microenvironment on the efficacy of KRAS-targeted therapy and emphasize the need to fully consider the complexity of the tumor microenvironment when targeted therapies for pancreatic cancer are developed. The innovations of this study are as follows: (1) we systematically evaluated the antitumor activity of AMG510 in KRAS G12C-mutant pancreatic cancer through both in vivo and in vitro experiments, providing clear insights into the potential application of this drug in pancreatic cancer; (2) through transcriptomic sequencing and related functional assays, we further revealed that AMG510 induces early protective autophagy in KRAS G12C-mutant pancreatic cancer cells and that autophagy plays a synergistic antitumor role in AMG510-induced cell death and showed that combining AMG510 with early autophagy inhibitors further enhances the tumor-killing effect. However, this study has several limitations: (1) the number of cell models used was limited and we primarily focused on MiaPaCa-2 cells, which may not fully represent the heterogeneous subtypes of KRAS-mutant pancreatic cancer, and have not explored the response mechanisms in other KRAS-mutant (e.g., G12D or G12V) or wild-type backgrounds; (2) clinical tissue samples or organoid models were not included in this study, and experimental systems that more closely mimic the clinical physiological state to evaluate the efficacy and safety of AMG510 monotherapy or combination therapy are lacking; (3) the specific molecular mechanisms by which AMG510 activates early autophagy in KRAS G12C-mutant pancreatic cancer cells remain unclear; and (4) because the number of patients harboring KRAS G12C mutations is relatively small, related clinical studies are often limited by small sample sizes. This not only increases the difficulty of conducting large-scale, multicenter clinical trials, but may also affect the adequacy of efficacy evaluation, toxicity monitoring, assessment of long-term benefit, and investigation of resistance mechanisms. Therefore, the findings of the present study are more appropriately interpreted as mechanistic evidence for KRAS G12C-mutant pancreatic cancer and as a reference for subsequent translational studies with larger cohorts, rather than being directly generalized to all KRAS-mutant pancreatic cancers. Conclusions In summary, this study not only confirms the specific antitumor activity of AMG510 in KRAS G12C-mutant pancreatic cancer but also reveals for the first time that AMG510 regulates cellular stress responses and survival by inducing autophagy. Autophagy plays a critical regulatory role in the activity of AMG510, and combining early-stage autophagy inhibitors (3-MA) significantly enhances its therapeutic effect, suggesting the synergistic therapeutic potential of targeting autophagy. Future studies should further elucidate the interaction between the KRAS G12C pathway and autophagy in conjunction with tumor microenvironment characteristics to explore more translationally promising combination intervention strategies. Methods Cell culture and treatments The human pancreatic cancer cell lines MiaPaCa-2 and PANC-1 and the human non-small cell lung cancer (NSCLC) cell line NCI-H358 were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All the cell lines were authenticated by short tandem repeat (STR) profiling and routinely tested negative for mycoplasma contamination. MiaPaCa-2 and PANC-1 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Shanghai, China) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin‒streptomycin. NCI-H358 cells were cultured in RPMI-1640 medium (Gibco, Shanghai, China) supplemented with 10% FBS and 1% penicillin–streptomycin. All the cells were incubated at 37 °C in a humidified atmosphere containing 5% CO₂. AMG510 (sotorasib; HY-114277, MedChemExpress, USA), 3-methyladenine (3-MA; HY-19312, MedChemExpress, USA), and chloroquine diphosphate (CQ; HY-17589A, MedChemExpress, USA) were administered to the indicated treatment groups. Cell viability assay Exponentially growing cells were trypsinized, counted, and seeded at a density of 3 × 103 cells per well into 96-well plates (100 μL per well), with three independent biological replicates per group. After 24 h of attachment, the cells were treated with various concentrations of AMG510 or with AMG510 (100 nM) plus 3-methyladenine (3-MA, 1 mM) for 48 h and 72 h. The CCK-8 working solution was prepared at a 1:10 dilution according to the manufacturer’s instructions (Beyotime, Haimen, China) and added to each well; after 1 h of incubation, the absorbance at 450 nm was measured using a microplate reader (BioTek Instruments, USA). Colony formation assay Cells were seeded at a density of 500 cells per well in 6-well plates. After attachment, the cells were treated with AMG510 (0, 50, 100, or 500 nM) or AMG510 (100 nM) plus 3-methyladenine (3-MA, 1 mM) and cultured for 14 days, and the medium was replaced every 3 days. Colonies were fixed with methanol, rinsed with phosphate-buffered saline (PBS), and stained with 0.1% crystal violet (Beyotime, Haimen, China) for 30 min. Colonies containing >50 cells were counted. Wound-healing assay Confluent monolayers were scratched with a sterile 200-μL pipette tip, gently rinsed with PBS, and switched to serum-free medium. Cells were then treated as indicated (vehicle control, AMG510 (100 nM), 3-MA (1 mM), or their combination). Phase-contrast images were acquired at 0 h and 72 h using an inverted microscope (CKX41, Olympus, Tokyo, Japan). Relative cell migration was quantified using Adobe Photoshop based on changes in the wound width between time points. Annexin V–FITC apoptosis assay After drug treatment, the culture supernatants were collected to retain floating cells; adherent cells were trypsinized, pooled with the supernatants, washed with cold PBS, and stained using an Annexin V–FITC/PI apoptosis detection kit (Beyotime, Haimen, China) according to the manufacturer’s instructions (15 min at room temperature, protected from light). The samples were analyzed on a FACSCanto II flow cytometer (BD Biosciences, USA), and the data were analyzed with FlowJo v10. Cell cycle analysis MiaPaCa-2 and H358 cells were treated with the indicated concentrations of AMG510 for 48 h, harvested, washed three times with ice-cold PBS, and fixed with 70% ice-cold ethanol (500 μL per tube) at 4 °C overnight. The cells were then incubated with 500 μL of propidium iodide (PI)/RNase staining solution (BD Pharmingen, Franklin Lakes, NJ, USA) for 30 min at room temperature in the dark. The DNA content was analyzed using a FACSCanto II flow cytometer (BD Biosciences, USA) with a 488-nm excitation wavelength, and the cell cycle distribution was quantified using FlowJo v10. MMP detection The mitochondrial membrane potential (ΔΨm) was assessed using a JC-1 assay kit (Beyotime, Haimen, China). After treatment with AMG510, MiaPaCa-2 cells were incubated with JC-1 at 37 °C in a 5% CO₂ atmosphere for 20 min according to the manufacturer’s instructions, and fluorescence was analyzed using flow cytometry. ROS detection ROS production was measured using a Reactive Oxygen Species Assay Kit (Beyotime, Haimen, China). After AMG510 treatment, MiaPaCa-2 cells were incubated with 10 μM DCFH-DA at 37 °C for 30 min, washed three times with PBS, and analyzed using flow cytometry and confocal microscopy according to the manufacturer’s instructions. RNA-seq MiaPaCa-2 cells were treated with DMSO or AMG510 (100 nM) for 48 h. Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, Shanghai, China) according to the manufacturer’s instructions. RNA-seq library preparation and sequencing were performed by a commercial service provider. Differential expression was analyzed with DESeq2, and genes whose adjusted P value (false discovery rate, FDR) was < 0.05 were considered significantly differentially expressed. Analysis of KRAS mutations MiaPaCa-2 and NCI-H358 cells carrying the KRAS G12C mutation and PANC-1 cells carrying the KRAS G12D mutation were cultured under standard conditions (37 °C, 5% CO₂) to the exponential phase, after which the medium was removed, and the cells were gently washed twice with PBS. Genomic DNA was extracted using a genomic DNA extraction kit (Beyotime, Haimen, China) according to the manufacturer’s instructions. KRAS regions encompassing codon 12 were amplified by PCR with gene-specific primers; amplicons were subsequently purified from agarose gels and subjected to Sanger sequencing by a commercial service provider. Sequence chromatograms were aligned to the human KRAS reference sequence to confirm the expected variants: KRAS p.G12C (c.34G>T) in MiaPaCa-2 and NCI-H358 cells and KRAS p.G12D (c.35G>A) in PANC-1 cells. Genomic DNA was extracted from frozen subcutaneous xenograft tumor tissues using a genomic DNA extraction kit according to the manufacturer’s protocol. KRAS exon 2 fragments encompassing codon 12 were amplified by PCR with gene-specific primers, purified from agarose gels, and subjected to Sanger sequencing by a commercial service provider. Sequencing chromatograms were aligned to the human KRAS reference sequence to confirm the presence of the KRAS G12C mutation. The relative abundance of the mutant allele was estimated from the peak intensities of the chromatograms. Specific experimental reference sequences are shown in Supplementary Figures S2–S3. KRAS transcript sequencing in tumor tissues Total RNA was isolated from frozen subcutaneous xenograft tumor tissues collected after treatment and reverse-transcribed to cDNA. KRAS cDNA fragments encompassing codon 12 were amplified by PCR using human-specific primers spanning exon–exon junctions to avoid genomic DNA amplification. PCR amplicons were purified from agarose gels and submitted to a commercial provider for Sanger sequencing. The resulting sequences were aligned to the KRAS reference transcript NM_004985.3 to determine the presence of the KRAS G12C mutation at the transcript level. Western blot analysis Total protein was extracted using RIPA lysis buffer supplemented with a1% protease inhibitor cocktail and quantified with a BCA protein assay kit (Beyotime Biotechnology, Haimen, China). Equal amounts of protein (30 μg per lane) were separated on 10% or 12% SDS–PAGE gels and transferred onto PVDF membranes (Merck Millipore, USA). The membranes were blocked with 5% nonfat milk in TBST for 1 h at room temperature, followed by an overnight incubation at 4 °C with primary antibodies against cleaved PARP (5625, CST, USA), cleaved caspase-3 (Servicebio, Wuhan, China), Ki-67 (Servicebio, Wuhan, China), LC3 (ab51520, Abcam, UK), p62/SQSTM1 (18420-1-AP, Proteintech, USA), mTOR (66888-1-Ig, Proteintech, USA), phospho-mTOR (67778-1-Ig, Proteintech, USA), Bax (60267-1-Ig, Proteintech, USA), Bcl-2 (26593-1-AP, Proteintech, USA), Beclin-1 (11306-1-AP, Proteintech, USA), CDK2 (18048, CST, USA), Cyclin E (20808, CST, USA), and Cyclin D3 (2936, CST, USA) at the dilutions recommended by the manufacturers. After being washed, the membranes were incubated with HRP-conjugated goat anti-mouse or goat anti-rabbit secondary antibodies for 1 h at room temperature, and the signals were developed using an enhanced chemiluminescence (ECL) substrate. β-Actin served as a loading control. Immunofluorescence staining MiaPaCa-2 and NCI-H358 (KRAS G12C mutant) cells were seeded on glass coverslips in 24-well plates at 2 × 104 cells per well and treated as indicated. The cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with 0.5% Triton X-100 for 15 min, and blocked with 5% bovine serum albumin (BSA) for 1 h at room temperature. The samples were incubated with an anti-LC3 antibody (ab51520, Abcam, Cambridge, UK) overnight at 4 °C, followed by an incubation with fluorophore-conjugated secondary antibodies and nuclear counterstaining with DAPI. Images were acquired using a laser scanning confocal microscope (Leica Microsystems, Wetzlar, Germany). The fluorescence intensity was quantified with ImageJ. Immunohistochemistry (IHC) Formalin-fixed, paraffin-embedded (FFPE) tumor sections (4 μm) were deparaffinized, rehydrated, and incubated with 3% hydrogen peroxide for 10 min to quench endogenous peroxidase activity. The sections were then incubated overnight at 4 °C in a humidified chamber with primary antibodies against Ki-67 (Servicebio, Wuhan, China), LC3 (ab51520, Abcam, Cambridge, UK), and cleaved caspase-3 (Servicebio, Wuhan, China). After being washed, the slides were incubated with HRP-conjugated secondary antibodies (CST, USA) at 37 °C for 30 min, developed with DAB substrate, counterstained with hematoxylin, dehydrated, and mounted. Images were acquired using a light microscope. TEM MiaPaCa-2 and NCI-H358 cells were seeded in 6-well plates, treated with AMG510 for 48 h, harvested, and fixed in 2.5% glutaraldehyde. The samples were postfixed with 1% osmium tetroxide, dehydrated through a graded ethanol series, and embedded in resin. Ultrathin sections were stained with uranyl acetate and lead citrate and examined using a transmission electron microscope to assess mitochondrial and endoplasmic reticulum morphology. mRFP-GFP-LC3 adenoviral transduction MiaPaCa-2 and NCI-H358 cells were seeded in 6-well plates during the exponential growth phase. When the cells reached 30–50% confluence, they were infected with mRFP-GFP-LC3 adenovirus according to the manufacturer’s instructions (Hanbio, China). At twenty-four hours post-infection, the cells were treated with AMG510 and CQ as indicated. After an additional 24 h, the fluorescent signals were imaged using a laser-scanning confocal microscope to assess autophagic flux. Animal experiments Male BALB/c nude mice (5–6 weeks old; Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were used for the in vivo studies. All procedures were approved by the Institutional Animal Care and Use Committee of Nanjing Drum Tower Hospital and conducted in accordance with established guidelines. For the subcutaneous xenograft model, MiaPaCa-2 cells (3 × 106) suspended in PBS/Matrigel (1:1, 200 μL) were injected into the right thigh. The tumor volume was measured every 3 days and calculated as V = (length × width²)/2. When the tumors reached 100–200 mm3, the mice were randomized into treatment groups. In the AMG510 monotherapy cohort, 12 mice were divided into vehicle and AMG510 groups (30 mg/kg/day, oral gavage, n = 6 mice per group) and treated for 21 days. In the AMG510 plus 3-MA cohort, 20 mice were randomized into four groups (vehicle, AMG510 30 mg/kg/day, 3-MA 25 mg/kg/day, and AMG510 + 3-MA; n = 5 mice per group) and treated for 21 days. For the orthotopic pancreatic tumor model, MiaPaCa-2 cells (5 × 10⁵) in PBS/Matrigel (1:1, 50 μL) were injected into the pancreatic parenchyma under anesthesia. When the tumors reached 100–200 mm3, the mice were randomized into vehicle or AMG510 groups (30 mg/kg/day, oral gavage, n = 5 mice per group) and treated for 21 days. Body weight and tumor growth were monitored every 3 days throughout the treatment period. At the study endpoint, the mice were euthanized, the tumor weights were recorded, and the tumors, major organs (heart, liver, kidneys, and lungs), and peripheral blood were collected for further analyses. Statistical analysis All experiments were independently repeated at least three times. The data are presented as the means ± SEMs, unless indicated otherwise. Comparisons among more than two groups were performed using one-way ANOVA followed by appropriate post hoc multiple-comparison tests; comparisons between two groups were performed using two-tailed Student’s t tests. A value of P < 0.05 was considered to indicate statistical significance (*P < 0.05; ⁎⁎P < 0.01; ⁎⁎⁎P < 0.001; and ⁎⁎⁎P < 0.0001). Statistical analyses were conducted with SPSS v19.0 and GraphPad Prism v10.0. Abbreviations FDA: Food and Drug Administration ROS: reactive oxygen species AMG510: Sotorasib 3-MA: 3-methyladenine ORR: objective response rate OS: overall survival TEM: Transmission Electron Microscopy CQ: chloroquine MMP: Mitochondrial Membrane Potential Ethics approval and consent to participate The current study was approved by the Institutional Animal Care and Use Committee of Nanjing Drum Tower Hospital (20,201,105). Consent for publication All authors agree with the content of the manuscript. Availability of data and materials Not applicable. Funding Supported by fundings for Clinical Trials from Nanjing Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University (2022-YXZX-ZL-02) CRediT authorship contribution statement Yukun Li: Writing – original draft, Visualization, Validation, Software, Methodology, Formal analysis, Data curation, Conceptualization. Mengmeng Zhou: Validation, Data curation. Jingjing Pan: Software, Data curation, Conceptualization. Haochen Su: Visualization, Validation, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Mengke Song: Validation, Data curation. Zujiao Guan: Software, Methodology, Conceptualization. Danmai Zhao: Visualization, Validation. Xiaoxin Zhao: Validation, Software. Yiqi Zhao: Visualization, Validation, Software. Pu Chen: Writing – review & editing, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Shu Zhang: Writing – review & editing, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Ying Lv: Writing – review & editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Declaration of competing interest The authors have no conflicts of interest to declare. Acknowledgements Not applicable. Footnotes Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.tranon.2026.102821. Contributor Information Yukun Li, Email: 1552417488@qq.com. Mengmeng Zhou, Email: mmzhou0611@163.com. Jingjing Pan, Email: pjj5810800@163.com. Haochen Su, Email: suhaochen0@163.com. Mengke Song, Email: 2791913263@qq.com. Zujiao Guan, Email: 2081424771@qq.com. Danmai Zhao, Email: zdm15383216647@163.com. Xiaoxin Zhao, Email: zhaoxx11037@163.com. Yiqi Zhao, Email: 2406981647@qq.com. Pu Chen, Email: chenpu@bjmu.edu.cn. Shu Zhang, Email: zhangsgastro@nju.edu. Ying Lv, Email: lvying@njglyy.com. Supplementary materials References Associated Data Supplementary Materials Data Availability Statement Not applicable.