AimA decrease in skeletal muscle mass in patients with gastric cancer following surgery is closely related to tumor recurrence and poor prognosis. Postoperative enteral nutrition support can accelerate the postoperative recovery of patients with gastric cancer. We aimed to analyze the effectiveness of short-term oral nutritional supplements(ONS) in attenuating postoperative skeletal muscle mass loss in gastric cancer patients after hospital discharge and the impact on long-term prognosis.MethodsThis study included patients who underwent radical gastrectomy at our center between 2014 and 2019. Univariate and multivariate logistic regression analyses were used to analyze the effectiveness of short-term oral nutrition for reducing postoperative skeletal muscle mass loss in patients with gastric cancer. Using cox regression to analyze the effect of ONS and each covariate on overall survival (OS).ResultsONS attenuated postoperative skeletal muscle loss, and it can improve the prognosis of patients, especially the elderly, those with poor nutritional status, anemia, and patients who have undergone total gastrectomy.ConclusionsShort-term oral nutrition is an independent protective factor for preventing skeletal muscle mass loss in patients with gastric cancer after surgery. Oral nutrition after surgery can reduce short-term muscle mass loss and is helpful in lowering the risk of death for patients.

BACKGROUND: Interferon regulatory factor 1 (IRF1) plays a crucial role in the type I interferon (IFN) response. However, its functional role and underlying mechanisms in gastric cancer (GC) remain unclear. This study aims to investigate the biological significance of IRF1 in GC progression and its potential as a therapeutic target. METHODS: IRF1 expression was analyzed using The Cancer Genome Atlas (TCGA) and GTEx databases, validated by immunohistochemistry (IHC) in 366 GC patients. Functional experiments, including CCK-8, Transwell migration and invasion assays, and apoptosis analysis, were conducted in GC cell lines with IRF1 overexpression or knockdown. A subcutaneous xenograft model was established to evaluate the in vivo effects of IRF1 on tumor growth. Co-immunoprecipitation and western blotting were performed to explore the molecular interactions between IRF1 and Myxovirus resistance 2 (MX2), as well as its regulation of the PI3K signaling pathway. RESULTS: IRF1 expression was significantly higher in gastric cancer tissues than in adjacent normal tissues. Higher IRF1 levels were also associated with improved patient survival. Overexpression of IRF1 inhibited GC cell proliferation, migration, and invasion while promoting apoptosis, whereas IRF1 knockdown had the opposite effects. Mechanistically, IRF1 suppressed PI3K/p-AKT signaling while enhancing p-ERK1/2 activation. Moreover, IRF1 directly interacted with MX2, a protein involved in epithelial-mesenchymal transition (EMT), and this interaction was essential for suppressing MX2-mediated oncogenic activity. In vivo experiments confirmed that IRF1 overexpression significantly reduced tumor growth and metastasis. CONCLUSIONS: IRF1 functions as a tumor suppressor in GC by modulating the PI3K signaling pathway and interacting with MX2 to inhibit EMT. These findings highlight IRF1 as a potential therapeutic target for GC treatment.

Background Most gastric cancer (GC) patients receive uniform treatment due to the lack of predictive biomarkers for chemotherapy or radiotherapy. We previously identified epithelial-mesenchymal transition (EMT) and metabolism subgroups in GC cell lines based on kinomic profiles, but their clinical relevance was unknown. Methods We developed an ensemble model using 37 kinases that differed between cell line subgroups to classify stage II-IV GC into EMT and metabolism subgroups. Survival differences between those who received chemotherapy or radiotherapy and those who did not were compared within each subgroup to validate the model effectiveness in predicting therapeutic response. An iteration approach was further applied to optimise and validate the feature set via multiple publicly-available datasets. Findings An 11-kinase signature stratified 893 patients into two subgroups. The metabolism subgroup showed significantly better survival with chemotherapy (HRmultivariable = 0.56) and radiotherapy (HRmultivariable = 0.55), whereas no such improvement was observed in the EMT subgroup. Significant interaction between kinomic subgroups and treatments were noted. The chemotherapy benefits between subgroups were greater with 5-fluorouracil-based regimens than cisplatin-based ones. This kinomic taxonomy was distinct from Lauren classification and previous transcriptomic subtypes and also suggested differential therapeutic vulnerabilities between subgroups. Interpretation This model holds promise for optimising chemotherapy and radiotherapy decisions for GC. Funding Biomedicine Discovery Scholarship and Graduate Research Completion Award, Monash University; National Natural Science Foundation of China (81602165) (C.H.). Australian Research Council Centre of Excellence for the Mathematical Analysis of Cellular Systems (CE230100001) (L.K.N.).

Background A substantial amount of research has been dedicated to the mechanisms by which tumor cells evade immune system recognition and manipulate the immune microenvironment to facilitate immune escape. Recent studies have shown that viruses and tumors can protect themselves from immune cytotoxicity by remodeling the actin cytoskeleton. However, cytoskeleton-mediated immune resistance and the specific cytoskeleton-related proteins involved require further research. Methods Single-cell RNA sequencing was used to identify cytoskeleton-related genes associated with the response to anti-programmed cell death protein 1 (PD-1) therapy across four digestive tumors. Immunohistochemistry was used to detect LASP1 expression in gastric cancer and analyze its prognostic value for survival and anti-PD-1 response. The impact of LASP1 deficiency on tumor response to anti-PD-1 treatment and cytotoxic lymphocyte-mediated lysis was determined in vivo and in vitro. Live-cell imaging was used to compare actin cytoskeletal dynamics at the immunological synapse between LASP1-deficient and mock tumor cells. Molecular mechanisms underlying LASP1-mediated immune-resistance were dissected using co-immunoprecipitation, immunofluorescence, domain deletion complementation, and Laurdan staining. Results The deficiency of LASP1 in gastric cancer affected the sensitivity of tumor cells to immunotherapy and induced cytotoxic lymphocytes exhaustion. LASP1 may act as a scaffold protein to regulate the Arp2/3 complex and remodel the cytoskeleton at the immunological synapses. LASP1 deficiency in tumor cells impairs lytic immunological synapse function by disrupting cytoskeletal dynamics-mediated cell membrane lipid organization at the immunological synapse. Finally, simvastatin combined with anti-PD-1 therapy reversed immunotherapy resistance in LASP1-deficient tumors. Conclusions The deficiency of LASP1 in tumors mediates immunological synapse dysfunction by affecting cytoskeletal dynamics-mediated cell membrane lipid organization, thus enabling tumors to protect themselves from immune cytotoxicity and immunotherapy.

Basal-type muscle-invasive bladder cancer (BMIBC) is characterized by aggressive metastasis and poor prognosis but lacks molecularly defined therapeutic targets. Through integrative analyses of clinical cohorts and BBN-induced mouse models, we identified KRT14 as a core oncogenic driver that orchestrates the KRT14-IGF2BP1 signaling axis to promote BMIBC progression and lung metastasis. Mechanistically, residues K294 and E295 within the KH2 domain of IGF2BP1 specifically recognize conserved residues D226 and E227 within the nuclear export signal of KRT14. This direct interaction facilitates IGF2BP1-mediated cytoplasmic trafficking and auto-stabilization of its own mRNA, establishing a positive feedback loop that amplifies IGF2BP1-targeted pro-invasive gene expression. Functional disruption of this axis suppressed primary tumor progression and lung metastasis in BMIBC models. Collectively, our findings define the KRT14-IGF2BP1 axis as a previously unrecognized, potentially targetable vulnerability in BMIBC, whose inhibition may limit aggressive disease progression and inform future therapeutic strategies.

Robotic surgery is a well-validated treatment option for colorectal cancer (CRC). We aimed to evaluate the efficacy and safety of the Weigao (WG) robotic system compared with those of the da Vinci (DV) platform for the surgical resection of CRC. We retrospectively analyzed patients with stage I–III CRC who underwent robotic CRC resection using the Weigao or da Vinci Surgical System. Statistical analysis of perioperative clinical data, including preoperative, intraoperative, and postoperative parameters was conducted. A total of 103 patients were included and divided into the WG (n = 65) and DV (n = 38) groups. All patients achieved surgical success, and there were no significant differences in preoperative baseline characteristics. However, the WG group demonstrated a significantly longer operative time. Patients in the WG group experienced a shorter time to first flatus and a low incidence of postoperative deep vein thrombosis. Notably, the total hospitalization cost was significantly lower when the Weigao Surgical System was used. This study demonstrated that the Weigao robotic system was comparable to the da Vinci system in terms of safety and efficacy for CRC surgery. Moreover, the total hospitalization cost was significantly reduced with the Weigao Surgical System, highlighting its potential as a cost-effective surgical option.

Cancer management remains fragmented across its continuum, from late-stage diagnosis and salvage therapies to non-personalized surveillance. Here, we present Oncoformer, a unified multimodal transformer model trained on the China Oncology Multimodal Prediction and Surveillance Study (COMPASS) cohort (3.67 million individuals, 17.7 million clinical visits) and validated on independent external cohorts, including the UK Biobank. Oncoformer integrates longitudinal electronic health records with chest X-ray imaging to address multiple clinical tasks: pan-cancer diagnosis (area under the receiver operating characteristic curve [AUROC] = 0.956), future cancer prediction up to 1 year before diagnosis (AUROC = 0.869), tumor stage inference (mean AUROC > 0.90), patient-specific treatment-response forecasting, and recurrence-free survival stratification across ten cancer types (all p < 0.01). Staging predictions were independently validated against postoperative pathological endpoints and shown to converge on core cancer genomic pathways. By translating routine clinical data into a dynamic view of cancer evolution, Oncoformer provides a framework for risk-informed cancer prediction and treatment stratification using routine clinical data.

Pangenomes are revolutionizing our ability to resolve genomic regions with complex variations1. However, existing human pangenomes2,3, constrained by small sample sizes, provide limited utility for medical and population genetic applications. Here we generated 1,116 diploid genome assemblies (55 de novo and 1,061 pangenome-informed) with an average size of 2.98 Gb and a mean quality value of 46 as part of the 1000 Chinese Pangenome (1KCP) project. On the basis of these assemblies, we constructed a pangenome comprising 405.3 million base pairs of sequences absent from the current references GRCh38 and CHM13, including 26.2 million base pairs of functional genic and predicted regulatory elements. We catalogued a full spectrum of genetic variation, including 35.4 million small variants, 110,530 structural variants (SVs), 485,575 tandem repeats (TRs) and 0.86 million nested variants embedded in non-reference sequences. This extensive dataset enabled detailed characterization of multiscale genic variations relevant to medical genetics, including gene-altering SVs, TR expansions, gene cluster variations and HLA gene haplotypes. Coupled with the 1KCP gene expression data, we conducted pan-variant expression quantitative trait locus (eQTL) mapping to analyse diverse variant types. We identified 3,256 eQTLs involving complex variants (SVs, TRs and nested variants) and elucidated their regulatory complexity. Finally, we developed a 1KCP pan-variant imputation reference panel, which provides multitype genetic markers to enhance the resolution of future association studies. This resource advances our understanding of complex variants and their functional implications to provide new insights into human health.

ABSTRACT Purpose This study aimed to explore the clinical benefits of receiving nutritional supplementation (NS) throughout the whole course of chemotherapy. Methods This multicenter prospective cohort study totally included 251 cancer patients requiring nutritional support and scheduled for chemotherapy. Primary outcomes included energy intake (EI), protein intake (PI), body mass index (BMI), NRS 2002 and PG‐SGA scores. Short‐term efficacy was the secondary outcome. Results Among the study participants, 168 received NS, whereas 83 opted for dietary advice (DA) alone. In the NS group, EI and PI demonstrated a gradual upward trend across the six cycles of chemotherapy, with no significant changes in the DA group. The BMI remained stable in both groups. The proportion of patients with or at risk of malnutrition showed a declining trend in the NS group but an increasing trend in the DA group. The generalized estimating equation results indicated that NS significantly improved PI (coefficient = 0.19, 95% CI: 0.11 to 0.27, p < 0.001), NRS 2002 (coefficient = −0.13, 95% CI: −0.23 to −0.03, p = 0.005), and PG‐SGA (coefficient = −0.18, 95% CI: −0.28 to −0.08, p < 0.001). The improvements in PI, NRS 2002, and PG‐SGA scores occurred from cycle 1, cycle 3, and cycle 2, respectively. Multivariate logistic analysis confirmed NS as a favorable factor associated with higher disease control rate (OR = 4.65, 95% CI: [1.88, 12.01], p = 0.001). Conclusions The incorporation of NS yielded several clinical benefits beyond adequate EI and stable weight. It contributes to higher protein intake and good nutritional status in patients with cancer throughout the whole course of chemotherapy, ultimately improving treatment efficacy. Trial Registration Patient‐Reported Outcome Management Including Surveillance and Intervention in Nutritional Group (PROMISING) Study (registration number: ChiCTR2100047535)

Cancer cachexia is a systemic metabolic disorder, with body weight loss and adipose tissue wasting as key features, and adipose tissue remodeling often preceding weight loss. Using pre-cachexia and cachexia models in Lewis lung carcinoma (LLC) tumor-bearing mice, transcriptomic analysis of white adipose tissue (WAT) identified Otopetrin 1 (Otop1) as a dynamically regulated gene, increased in pre-cachexia and decreased in cachexia. In patients with cancer, OTOP1 expression in subcutaneous WAT was reduced in cachexia and positively correlated with BMI in cachectic patients. In adipocytes treated with tumor-conditioned medium, OTOP1 overexpression alleviated metabolic dysfunction, accompanied by suppression of NF-κB signaling and activation of PPARγ, leading to reduced lipolysis and enhanced adipogenesis; these effects were partially attenuated by the PPARγ antagonist. Moreover, overexpression of OTOP1 in adipose tissue of LLC tumor-bearing mice alleviated adipose tissue wasting and improved lipid metabolism. These findings suggest a role for OTOP1 in adipose tissue remodeling during cancer cachexia.

The cyclic guanosine monophosphate (GMP)-adenosine monophosphate (AMP) synthase-stimulator of interferon genes (cGAS-STING) pathway is crucial for tumor immunity. However, activation of the cGAS-STING pathway alone is seldom sufficient to eliminate established tumors. Here, we report the engineering of zinc/manganese (Zn/Mn)-based metal-organic framework (MOF) nanoparticles, that is, AMP@Zn/Mn-MOF, comprising Zn/Mn-MOF nanoparticles as the carrier and the STING agonist c-di-AMP diammonium as the therapeutic drug for reinforcing antitumor immune responses. These therapeutic nanoplatforms can significantly activate the cGAS‒STING pathway and facilitate the innate immune response. Furthermore, the peroxidase (POD)-mimetic and glutathione oxidase (GSHox)-mimetic activities of AMP@Zn/Mn-MOF can significantly potentiate tumor cell death and effectively induce robust immunogenic cell death (ICD), thereby amplifying the cGAS-STING pathway. Moreover, AMP@Zn/Mn-MOF reprogrammed the immunosuppressive tumor microenvironment by promoting intratumoral lymphocyte infiltration, thereby significantly suppressing the growth of murine MC38 tumors in mice. Notably, AMP@Zn/Mn-MOF amplified the therapeutic effect of anti-programmed death ligand 1 (αPD-L1) blockade by triggering systemic antitumor responses, resulting in a notable abscopal effect to effectively inhibit distant tumors. In summary, AMP@Zn/Mn-MOF offers a nanoplatform with enhanced antitumor effectiveness through activation of the cGAS-STING pathway and ICD, suggesting that enhanced immune checkpoint blockade-based immunotherapy is promising for colon cancer treatment.

Stress-generating therapeutic materials are commonly optimized to deliver cytotoxic inputs, but delivery alone does not determine whether intracellular stress becomes irreversible damage. Tumor cells can buffer metal-, redox- and mitochondrial stress through glucose-dependent metabolism, creating a gap between stress exposure and stress execution. Here, we report a tumor-state-gated metallo-DNAzyme framework, C@HGDz2 1/CaCu, that introduces metabolic licensing as a programmable material function. The framework couples a miR-21/APE1-gated GLUT1 DNAzyme with a Ca/Cu-partitioned metal-nucleic-acid architecture and co-assembled cystine. Tumor-state logic confines GLUT1 mRNA cleavage to dual-input cancer-associated conditions, whereas metal partitioning assigns Ca2 + to DNAzyme-compatible catalysis and Cu2 + to framework persistence and copper-associated stress within the metabolically rewired cellular state. After HER2-guided deployment, C@HGDz2 1/CaCu suppresses GLUT1 and glycolytic signaling, restores PTEN-associated signaling and remodels glucose-derived metabolic buffering. This metabolic remodeling is associated with enhanced engagement of copper- and cystine-dependent oxidative, mitochondrial and cytoskeletal stress. In an orthotopic gastric tumor model, the framework enhances tumor accumulation, shows coordinated metabolic rewiring and stress engagement, and suppresses tumor progression with apparent tolerability under the tested regimen. This work establishes state-gated metallo-DNAzyme framework engineering as a strategy for shifting stress-generating materials from cytotoxic delivery to vulnerability programming.

Chemotherapy is often accompanied by impaired dietary intake, which is linked to poor clinical outcomes if not effectively addressed. Predicting dietary changes can facilitate early nutritional support, but the dynamic patterns and interindividual variability in dietary intake during chemotherapy remain unclear. Our aim was thus to evaluate the feasibility of using the Simple Diet Self-Assessment Tool (SDSAT) to establish dietary intake patterns in patients undergoing chemotherapy. This is an ongoing, prospective cohort study. Patients receiving chemotherapy from the Cancer Hospital, Chinese Academy of Medical Sciences were enrolled. Diet was assessed using the SDSAT once before the treatment and at least three times during one chemotherapy cycle. We applied a group-based trajectory model to identify trajectories of dietary intake. The 24-hour dietary recall (24HR) was used to verify the validity of energy and protein intake estimated by SDSAT; compliance with SDSAT was also assessed. Five trajectory groups were identified for food intake based on 217 chemotherapy cycles by 56 patients. The compliance rate accounted for 81.11 https://www.chictr.org.cn/indexEN.html .

Epstein-Barr virus-associated gastric cancer (EBVaGC) is a unique subtype of gastric cancer (GC) with distinct molecular characteristics that generally has a better prognosis. BamHI-A leftward frame 4 (BALF4), an envelope glycoprotein encoded by the Epstein-Barr virus (EBV), plays an important role in EBV infection. However, its biological function and potential molecular mechanisms in EBVaGC remain unclear. This study aimed to investigate the impact of the highly expressed viral gene BALF4 on the progression of EBVaGC. Here, we detected the expression of BALF4 in GC tissue chips and validated that the presence of BALF4 might be associated with a favorable prognosis in EBVaGC. The results showed that BALF4 inhibited the proliferation, migration, and invasion of GC cells in vitro and in vivo. In addition, we discovered that BALF4 interacts with N-acetyltransferase 10 (NAT10). High expression of NAT10 in GC tissues promotes the malignant phenotype of GC cells. We discovered that BALF4 could inhibit the malignant progression of GC by promoting the ubiquitination and degradation of NAT10. In summary, our study revealed a possible mechanism explaining the favorable prognosis of the EBVaGC subtype, which contributes to a better understanding of this special type of GC.

Chemotherapy remains essential for treating colorectal cancer (CRC), with personalized approaches increasingly addressing patient heterogeneity. Organ-on-a-chip systems offer promise for modeling tumors and guiding drug selection, yet replicating the intestine’s nutrient-rich, 3D microenvironment remains challenging. In this study, we developed a microfluidic chip integrated with hydrogel microcapsules to culture 3D tumor models from primary CRC cells. Cells were encapsulated in microcapsules with carboxymethyl cellulose cores and alginate shells, forming uniform, viable spheroids. These spheroids were exposed to dynamic drug gradients within the chip, enabling high-throughput chemotherapy screening. The system preserved tumor heterogeneity and revealed patient-specific drug responses The sensitivity of tumor spheroids on the chip to various drug regimens closely reflected clinical treatment outcomes, demonstrating strong consistency across all nine patients in this study. These results demonstrate that the 3D tumor spheroids model and drug evaluation chip provide an effective method for selecting personalized chemotherapy regimens for CRC patients.

Infectious diseases caused by pathogenic microorganisms and tumors arising from uncontrolled cell proliferation can be intricately linked through the lens of viromics. This study aimed to delineate the landscape of viral infections in gastric cancer, focusing on the relationship between pathogenic microorganisms and tumorigenesis. By conducting transcriptomic sequencing of gastric cancer tissues, we identified gastric cancer-associated viruses from raw transcriptomic data. This methodology was validated using public databases and experimental evidence, ensuring the authenticity of viral detection. Single-cell data further identified specific cells harboring these viruses and elucidated the correlation between viral gene expression and the genesis and progression of gastric cancer. Common tumor viruses, such as human endogenous retrovirus (HERV), Epstein-Barr virus (EBV), and human cytomegalovirus (HCMV), along with unique DNA viruses and bacteriophages, were detected in gastric cancer. Additionally, these viruses have been detected in both tumor and immune cells, indicating a close association between viral gene expression and the development of gastric cancer. This study advances the technique of screening viruses from transcriptomes, charting the distribution of viruses within tumors, clarifying the interplay between viruses and cancer, and identifying previously unknown viruses related to the occurrence and progression of cancer.IMPORTANCEIn our study, we have carefully examined the viral landscape in gastric cancer, supported by a series of thorough experiments that verify its reliability. Our approach has not only confirmed the presence of viruses known to be associated with cancer development but also identified a range of additional viral entities, including human cytomegalovirus (HCMV) and various herpesviruses, along with numerous bacteriophages. The high incidence of these viruses within tumor samples suggests they could be considered as potential biomarkers for early cancer detection. This method enhances our understanding of the role viruses play in cancer, which may assist scientists and medical professionals in identifying viral presence in cancers and could offer new angles for cancer prevention and the development of related measures. By identifying specific viruses linked to different cancers, we aim to improve patient outcomes.

Alpha-fetoprotein-producing gastric cancer (AFPGC) is a rare but highly aggressive subtype of gastric cancer. Patients with AFPGC are at high risk of liver metastasis, and the tumor microenvironment (TME) is complex. A multicenter retrospective study is conducted from January 2011 to December 2021 and included 317 AFPGC patients. Using a multivariable logistic regression model, a nomogram for predicting liver metastasis is built. By combining AFP and the neutrophil-lymphocyte ratio (NLR), we developed a novel and easily applicable predictive indicator, termed ANLiM score, for liver metastasis in AFPGC. An integrated multi-omics analysis, including whole-exome sequencing and proteomic analysis, is conducted and revealed an immunosuppressive TME in AFPGC with liver metastasis. Single-cell RNA sequencing and multiplex immunofluorescence identified the potential roles of tumor-associated neutrophils and tertiary lymphoid structures in shaping the immune microenvironment. These findings are validated in a real-world cohort receiving anti-programmed cell death 1 (anti-PD-1) therapy, which showed concordant effectiveness. In addition, the ANLiM score is also identified as a promising biomarker for predicting immunotherapy efficacy. Overall, a blood biomarker-based predictive indicator is developed for liver metastasis and immunotherapy response in AFPGC. The findings on immune microenvironmental alterations for AFPGC with liver metastasis provide new insights for optimizing immunotherapy strategies.

Human gastric cancer (GC) is one of the most malignant cancers, and cisplatin (Cis)-based chemotherapy remains the main clinical treatment for GC. However, Cis resistance often occurs, largely limiting its therapeutic efficacy in tumors. Therefore, a better understanding of the drug resistance mechanism could reveal new approaches for improving GC treatment efficacy. Here, we define the integrative role of nucleolar and coiled-body phosphoprotein 1 (NOLC1), a molecular chaperone that is significantly upregulated in GC tissues and Cis-resistant GC cells. Knocking down NOLC1 increased GC sensitivity to Cis by regulating ferroptosis. Mechanistically, NOLC1 binds to the p53 DNA-binding domain (DBD), decreasing p53 nuclear accumulation stimulated by Cis and suppressing p53 transcriptional functions. Then, the p53-mediated ferroptosis is suppressed. Furthermore, the silence of NOLC1 promoted ferroptosis-induced immunogenic cell death (ICD) and reprogrammed the immunosuppressive tumor microenvironment, thereby increasing sensitivity to anti-programmed cell death-1 (PD-1) therapy plus Cis. The combination of anti-PD-1 plus Cis effectively inhibited GC growth without significant side effects. In summary, our findings reveal that targeting NOLC1 may be a novel therapeutic strategy for GC and may increase the efficacy of chemotherapy combined with immune checkpoint inhibitor (ICI) therapy.

NSUN2, a major methyltransferase that catalyzes m5C methylation in eukaryotes, is known to be implicated in the development of multiple cancers. However, its role in colorectal cancer (CRC) and the related molecular mechanisms have yet to be sufficiently determined. Here, we conducted an analysis of public database (722 CRC patients) and two distinct cohorts from our centre (1559 CRC patients), which revealed that NSUN2 is upregulated in CRC and correlates with unfavourable prognosis. Our analyses also showed that NSUN2 promotes the proliferation and metastasis capabilities of CRC cells. Intriguingly, NSUN2 was found to promote CRC via an m5C-independent mechanism, which has not been previously reported. Overexpression of both wild-type and m5C enzymatic-dead mutant NSUN2 upregulated and activated the ErbB-STAT3 signalling pathway. We also found that both wild-type and the m5C enzymatic-dead mutant NSUN2 closely interacted with CUL4B. Silencing of CUL4B effectively inhibited the m5C-independent function of NSUN2. Moreover, overexpression of NSUN2 enhanced the sensitivity of CRC cells to lapatinib. Taken together, our findings revealed a novel m5C-independent mechanism for NSUN2 in the malignancy and lapatinib sensitivity of CRC via activation of the CUL4B/ErbB-STAT3 pathway, which provides a potential therapeutic strategy for patients with CRC. HIGHLIGHTS: NSUN2 is upregulated in CRC and associated with poor prognosis of CRC patients. NSUN2 promotes CRC malignancy independently of its m5C-enzymatic activity, a mechanism that has not been previously reported. The non-m5C carcinogenic roles of NSUN2 may be mediated through interactions with CUL4B, thereby activating the ErbB-STAT3 signalling pathway. NSUN2-mediated upregulation of ErbB-STAT3 pathway enhances the sensitivity of CRC to lapatinib treatment.