PURPOSEThe neoadjuvant chemoradiotherapy (nCRT) might accentuate surgical complications and toxicity in the treatment of locally advanced rectal cancer (LARC) while neoadjuvant chemotherapy (nCT) alone shows promise as an alternative treatment. However, which patients deserve most from the nCT need further clarify. This trial aimed to assess the non-inferiority of nCT with capecitabine plus oxaliplatin (CAPOX) versus nCRT with capecitabine in LARC with uninvolved mesorectal fascia (MRF).METHODSPatients with LARC within 12 cm from the anal verge and uninvolved MRF were randomly assigned to receive 4 cycles of CAPOX chemotherapy alone (nCT group) or CRT with concurrent Capecitabine (nCRT group). The primary end point is 3-year locoregional recurrence-free survival (LRRFS). Secondary end points, such as 3-year disease-free survival (DFS), 3-year overall survival (OS), and adverse events (AEs), were also reported.RESULTSA total of 663 patients were enrolled and 589 patients received the allocated treatment (nCT, n = 300; nCRT, n = 289). LRRFS was analyzed with a median follow-up of 48 months. 3-year LRRFS was 97.4% (95% CI, 95.5 to 99.3) in the nCRT group and 96.3% (95% CI, 94.0 to 98.6) in the nCT group, resulting in a hazard ratio (HR) of 1.40 (95% CI, 0.53 to 3.68). The nCT and nCRT achieved similar 3-year DFS (89.2% v 87.9%; HR, 0.88 [95% CI, 0.54 to 1.44]) and 3-year OS (95.0% v 94.1%; HR, 0.86 [95% CI, 0.42 to 1.76]). The nCT group showed a lower incidence of grade 2 to 4 long-term AEs (16.0% v 26.3%, P = 0.002) and proctitis (33.6% v 41.7%, P = 0.049) compared with nCRT group.CONCLUSIONSThe non-inferiority of nCT was not confirmed with a very low incidence of local recurrence in both group. But nCT offers comparable DFS and OS while mitigating the burden of toxicity as compared to nCRT. These insights shed light on a potential paradigm shift in the treatment for LARC with uninvolved MRF.

Figure S6. The effects of P. copri on metabolic profiles in vivo and in vitro. (A) Intensities of GPC related metabolites in three P. copri exposed models. (B) GPC enzymatic gene expression in tumors from the intratumoral bacterial model. Volcano plots showing downregulated GPC in the serum of P.copri-gavaged APCmin/+ mice (C), serum of the intratumoral bacterial model (D), and Pc.CM (E). (F) KEGG pathway enrichment of differential metabolites in Pc.CM. Data are expressed as mean± SD.

Supplementary Table S3: RT-qPCR Primers and short hairpin RNAs (shRNAs) sequence of BHLHE40

HnRNPAB promotes the liver metastasis of PDAC cells

Cancer-associated fibroblasts (CAFs) drive key aspects of tumor malignancy. However, the profound heterogeneity in cytokine secretion and cell-cell communication orchestrated by CAFs renders therapeutic strategies targeting these molecules largely ineffective against cancer. In contrast, exploring the metabolic reprogramming mechanisms of CAFs could be conducive to therapeutic intervention in colorectal cancer (CRC). Here, we conducted metabolomic and single-cell RNA-sequencing analyses that uncovered increased glycolysis, intracellular lactate, and histone lactylation in CAFs compared to normal fibroblasts (NFs). Elevated histone 3 lysine 18 lactylation (H3K18la) in CAFs promoted the transition of CAFs to myofibroblastic CAFs (myCAFs), increased collagen deposition, and enhanced organoid growth. Furthermore, fibroblast-specific conditional lactic dehydrogenase A (Ldha) knockout in a mouse model mitigated CRC tumorigenesis and progression in vivo. Cancer cell-derived TGF-β accelerated metabolic reprogramming and histone lactylation in CAFs. NPM1, a histone chaperone functioning as a cofactor of P300, enhanced the H3K18la level of the ROCK1 promoter and activated ROCK1 transcription. The RhoA/ROCK1/MLC2/MRTF-A pathway and extracellular matrix (ECM) remodeling were indispensable for the tumor-promoting effect of CAF histone lactylation in CRC. Furthermore, CAF histone lactylation limited CD8+ T cell infiltration and aggravated CD8+ T cell exhaustion via ECM remodeling. Stiripentol, a clinical drug that targets lactylation, potentiated the efficacy of immunotherapy in CRC models. Together, this study demonstrates that elevated lactate levels in CAFs reprogram the epigenetic landscape to promote tumor progression and immunosuppression, highlighting the potential of targeting CAF histone lactylation as a therapeutic strategy for CRC.

Supplementary Table S5: The top50 differentially expressed genes of each CAFs and Malignant EPCs subgroups are displayed.

Figure S1. Display of Leiden clusters and different cluster marker genes. (A) UMAP of all the 166,180 single cells in 22 primary colorectal cancers, annotated by Leiden clusters. (B) Dot plot of common cell markers in different Leiden clusters.

Figure S8. Immunofluorescence assay of LoVo and HCT116 cells. Scale bar: 20 µm.

Figure S6. Survival analysis of the protein expression of BHLHE40 in the TMA. (A, B) The KM plots showing the OS and DFS of the BHLHE40 high and the BHLHE40 low groups in the TMA of 80 colorectal cancer patient samples.

Figure S3. Potential functions of CXCL1+ CAFs and SFRP2+ CAFs. (A, B) Dot plot displaying GO analysis of CXCL1+ CAFs and SFRP2+ CAFs.

Prediction of hnRNPAB binding to MYC mRNA and effects of hnRNPAB on MYC protein stability

Background:Colorectal mucinous adenocarcinoma (MAC) shows heterogeneous outcomes after curative resection, and the prognostic value of lymphovascular invasion (LVI) remains unclear. This study evaluated the association of LVI with clinicopathological features, overall survival (OS), and disease-free survival (DFS) in patients with colorectal MAC. Methods:This retrospective cohort study included 2,391 adults with stage I-III pathologically confirmed colorectal MAC who underwent curative resection across 21 hospitals in China between 2016 and 2021. Patients were excluded if they had a mucinous component <50%, age <18 or >80 years, multiple primary tumors, synchronous multiple colorectal cancers, stage IV disease, or incomplete clinical, pathological, or follow-up data. Baseline clinicopathological factors included sex, age, tumor size, tumor location, differentiation, LVI status, T category, N category, and tumor-node-metastasis (TNM) stage. LVI was defined as tumor cells within endothelial-lined spaces or destruction of a lymphovascular wall. Postoperative surveillance was performed every 3 months for 2 years and every 6 months thereafter up to 5 years. OS and DFS were obtained from follow-up records and analyzed using Kaplan-Meier, log-rank, and Cox proportional hazards methods. LVI was fixed as the primary exposure in multivariable Cox models, and collinearity among candidate covariates was assessed using the variance inflation factor (VIF). All statistical tests were two-sided. Results:Among 2,391 patients, 1,397 (58.4%) were male, 994 (41.6%) were female, and 1,231 (51.5%) had stage III disease. LVI was detected in 541 patients (22.6%). LVI positivity was significantly associated with poor differentiation, advanced T stage, N stage, and TNM stage (all P<0.001). The 3-year OS rate was lower in the LVI-positive group than in the LVI-negative group (74.2% vs. 86.5%, P<0.001), as was the 3-year DFS rate (73.8% vs. 86.3%, P<0.001). In fixed-LVI multivariable Cox models, LVI positivity remained independently associated with poorer OS [hazard ratio (HR) =1.855; 95% confidence interval (CI): 1.319-2.607, P<0.001] and DFS (HR =1.923; 95% CI: 1.366-2.706, P<0.001). Subgroup analyses showed that LVI-positive patients had poorer OS and DFS than LVI-negative patients in left-sided colon, rectum, T3 stage, lymph node metastasis, and TNM stage III subgroups. Conclusions:In patients with MAC after curative resection, LVI was associated with more advanced pathological features and independently predicted poorer OS and DFS. Routine assessment of LVI may improve postoperative risk stratification and help identify patients who require closer surveillance.

Robotic-assisted laparoscopic surgery has made progress in addressing many of the technical challenges associated with conventional laparoscopy. Recently, a new robotic surgical system, Carina Platform, has been developed. The aim of this study is to evaluate the feasibility and safety of Carina for proctectomy surgery in preclinical models. Port and cart positions were initially determined by surgical simulation based on three-dimensional reconstruction. Six pigs, divided equally into acute and chronic groups, underwent robotic-assisted proctectomy using the Carina robotic system. Operative time, device errors and perioperative complications were recorded. All animals were observed postoperatively to assess their mental status, and after four weeks, the chronic pigs were euthanized for necropsy to evaluate the recovery of the anastomotic stoma. The optimal port and patient cart positions were further assessed in three human cadavers. Robotic-assisted proctectomy was successfully completed in all porcine subjects. No device-related intraoperative complications were observed. An autopsy examination revealed that the chronic pigs, which had been followed up for 28 days, had made a favorable recovery. The configurations of ports and patient carts were successfully validated in human cadaver models that could simulate the completion of a rectal resection. The feasibility and safety of the newly developed Carina robotic system for performing proctectomy were successfully demonstrated in porcine and cadaveric models. Further studies are required to validate its clinical application in human patients.

Primers used for PCR amplification and Summary of clinicopathologic characteristics

Colorectal cancer is characterized by a complex tumor microenvironment (TME) shaped by intestinal microbiota. In this study, 16S rRNA sequencing of tissues from patients with colorectal cancer identified Prevotella, particularly the dominant species Prevotella copri, as a key intratumoral bacterium. The parenchymal invasion of P. copri was confirmed by FISH, and the abundance of P. copri correlated with advanced tumor stages and postoperative serologic markers. Notably, the reduced abundance of P. copri in paired normal tissues implied potential bacterial translocation during tumorigenesis. In multiple murine models, P. copri not only accelerated tumor growth but also reprogrammed tumor-associated macrophages (TAM) toward a protumoral state. Untargeted metabolomics revealed glycerophosphocholine (GPC) as the only conserved metabolite depleted by P. copri across murine models and bacterial cultures, a finding confirmed by spatial metabolomics in clinical specimens. Strikingly, GPC supplementation reprogrammed MARCO+ TAMs toward an antitumoral phenotype, effectively counteracting P. copri-mediated tumor progression. Overall, this study uncovers a paradigm in colorectal cancer pathogenesis in which P. copri creates an immunosuppressive niche by depleting GPC to manipulate macrophage polarization. These findings position P. copri as both a noninvasive diagnostic marker and druggable therapeutic target, with GPC restoration representing a promising immunometabolic intervention strategy. SIGNIFICANCE:Glycerophosphocholine depletion by intratumoral P. copri induces immunosuppressive polarization of MARCO+ macrophages in colorectal cancer, revealing a microbial-metabolic-immune axis that remodels the tumor microenvironment and represents a potential immunotherapeutic target.

Figure S4. The effect of P. copri on colorectal tumorigenesis. (A) APCmin/+ mice received oral gavage of P. copri starting at 5 weeks of age, while the control group received saline for 12 weeks (n=10 per group, Independent replicate experiment corresponding to Figure 3A.). (B) Number and volume of colorectal tumor, (C) Spleen weight, and (D) colon images of APCmin/+ mice. (E) Representative histological images of colon tissues by H&E staining of colon from APCmin/+ mice. (F) Representative images of P. copri colonization in colon by FISH. (G) In the subcutaneous tumor model, C57BL/6 mice were pre-treated with P. copri via gavage before MC38 cell injection (n=10 per group, Independent replicate experiment corresponding to Figure 3G.). (H) Volume and weight of tumor and (I) spleen weight of C57BL/6 mice. (J) Ki67 and (K) PCNA expression in tumors from P. copri-gavaged mice. (L) Subcutaneous tumors were established in C57BL/6 mice by inoculation of MC38 cells with P. copri (MOI=100 after 24h co-culture) (n=10 per group, Independent replicate experiment corresponding to Figure 3K.). (M) Weight and volume of tumor and (N) spleen weight of C57BL/6 mice. (O) Ki67 and (P) PCNA expression in tumors from the intratumoral bacterial model. i.g., intragastric administration (gavage); s.c., subcutaneous injection. Data are expressed as mean± SD. Compared with control (Ctrl) group, * P< 0.05; ** P< 0.01; *** P< 0.001.

Figure S2. Intratumoral microbiota in CRC tumor tissues. (A) Rarefraction curves for two groups. (B) Beta diversity showed by PCoA, and the statistic difference was analyzed by PERMANOVA (R²=0.012, F=3.723, P< 0.001). (C) Prevotella levels in clinical samples versus negative controls. (D) Indicator species analysis reveals Prevotella as a significant indicator for T3-4 stage. (E) No significant intergroup differences in postoperative AFP levels were detected among either T3-4 or T0-2 patients stratified by P. copri abundance. (F) The relative abundance of P. copri in fecal samples from the PRJEB7774 cohort (n=46) (G) and the PRJDB4176 cohort (n=196). (H) FISH images of P. copri in tumor tissues from a T2N0M0 and a T3N0M0 case. (I) Schematic illustration of the Tumor-Normal Neutral Model. Low: Prevotella-low (< median); High: Prevotella-high (> median). Data are expressed as mean± SD, * P< 0.05; *** P< 0.001.

Figure S9. BHLHE40 promoted the invasion and migration of colorectal cancer induced by the TGFβ pathway in vitro. (A, B) Transwell assays and wound healing assays were used to assess the migration and invasion abilities of LoVo cells and HCT116 cells with or without treatment of TGFβ or LY2109761. Scale bar: 100 µm. All data were presented as the means ± SD of three independent experiments. *, P < 0.05, **, P < 0.01, and ***, P < 0.001, Student t test.

Clinical characteristics of patients in cohort II.

Figure S5. The effect of P. copri on the viability of CRC cells and macrophage polarization. (A) Cell viability of HCT116/MC38 cells stimulated directly with P. copri, or (B) indirectly via macrophage medium conditioned by P. copri. (C-D) Cell viability of HCT116/MC38 cells stimulated directly with Pc.CM, heat-killed bacteria or (E) indirectly via macrophage medium conditioned by Pc.CM, heat-killed bacteria. Flow cytometry analysis of (F) CD3+ cells and (G) CD8+ cells in P. copri-treated APCmin/+ mice. (H) Cell cycle analysis of HCT116 cells co-cultured with THP-1 macrophages pre-treated with Pc.CM (bar graph shows cell cycle phase distribution). The effect of P.copri conditioned medium (Pc.CM) on (I) CD86+ macrophages (M1-like) and (J) CD206+ macrophages (M2-like). Data are expressed as mean± SD, ** P< 0.01; *** P< 0.001.