Age-, sex- and proximal-distal-resolved multi-omics identifies regulators of intestinal aging in non-human primates.

The incidence of intestinal diseases increases with age, yet the mechanisms governing gut aging and its link to diseases, such as colorectal cancer (CRC), remain elusive. In this study, while considering age, sex and proximal-distal variations, we used a multi-omics approach in non-human primates (Macaca fascicularis) to shed light on the heterogeneity of intestinal aging and identify potential regulators of gut aging. We explored the roles of several regulators, including those from tryptophan metabolism, in intestinal function and lifespan in Caenorhabditis elegans. Suggesting conservation of region specificity, tryptophan metabolism via the kynurenine and serotonin (5-HT) pathways varied between the proximal and distal colon, and, using a mouse colitis model, we observed that distal colitis was more sensitive to 5-HT treatment. Additionally, using proteomics analysis of human CRC samples, we identified links between gut aging and CRC, with high HPX levels predicting poor prognosis in older patients with CRC. Together, this work provides potential targets for preventing gut aging and associated diseases.

Prediction of symptomatic anastomotic leak after rectal cancer surgery: A machine learning approach.

INTRODUCTION: Anastomotic leakage (AL) remains the most dreaded and unpredictable major complication after low anterior resection for mid-low rectal cancer. The aim of this study is to identify patients with high risk for AL based on the machine learning method. METHODS: Patients with mid-low rectal cancer undergoing low anterior resection were enrolled from West China Hospital between January 2008 and October 2019 and were split by time into training cohort and validation cohort. The least absolute shrinkage and selection operator (LASSO) method and stepwise method were applied for variable selection and predictive model building in the training cohort. The area under the receiver operating characteristic curve (AUC) and calibration curves were used to evaluate the performance of the models. RESULTS: The rate of AL was 5.8% (38/652) and 7.2% (15/208) in the training cohort and validation cohort, respectively. The LASSO-logistic model selected almost the same variables (hypertension, operating time, cT4, tumor location, intraoperative blood loss) compared to the stepwise logistic model except for tumor size (the LASSO-logistic model) and American Society of Anesthesiologists score (the stepwise logistic model). The predictive performance of the LASSO-logistics model was better than the stepwise-logistics model (AUC: 0.790 vs. 0.759). Calibration curves showed mean absolute error of 0.006 and 0.013 for the LASSO-logistics model and stepwise-logistics model, respectively. CONCLUSION: Our study developed a feasible predictive model with a machine-learning algorithm to classify patients with a high risk of AL, which would assist surgical decision-making and reduce unnecessary stoma diversion. The involved machine learning algorithms provide clinicians with an innovative alternative to enhance clinical management.

Reactive oxygen species in colorectal cancer adjuvant therapies.

Colorectal cancer (CRC), a prevalent global malignancy, often necessitates adjuvant therapies such as chemotherapy, radiotherapy, targeted therapy, and immunotherapy to mitigate tumor burden in advanced stages. The efficacy of these therapies is significantly influenced by reactive oxygen species (ROS). Previous research underscores the pivotal role of ROS in gut pathology, targeted therapy, and drug resistance. ROS-mediated CRC adjuvant therapies encompass a myriad of mechanisms, including cell death and proliferation, survival and cell cycle, DNA damage, metabolic reprogramming, and angiogenesis. Preliminary clinical trials have begun to unveil the potential of ROS-manipulating therapy in enhancing CRC adjuvant therapies. This review aims to provide a comprehensive synthesis of studies exploring the role of ROS in CRC adjuvant therapies.

Metastasis organotropism in colorectal cancer: advancing toward innovative therapies.

Distant metastasis remains a leading cause of mortality among patients with colorectal cancer (CRC). Organotropism, referring to the propensity of metastasis to target specific organs, is a well-documented phenomenon in CRC, with the liver, lungs, and peritoneum being preferred sites. Prior to establishing premetastatic niches within host organs, CRC cells secrete substances that promote metastatic organotropism. Given the pivotal role of organotropism in CRC metastasis, a comprehensive understanding of its molecular underpinnings is crucial for biomarker-based diagnosis, innovative treatment development, and ultimately, improved patient outcomes. In this review, we focus on metabolic reprogramming, tumor-derived exosomes, the immune system, and cancer cell-organ interactions to outline the molecular mechanisms of CRC organotropic metastasis. Furthermore, we consider the prospect of targeting metastatic organotropism for CRC therapy.

Selenium metabolism heterogeneity in pan-cancer: insights from bulk and single-cell RNA sequencing.

BACKGROUND: Selenium, a natural microelement with both nutritional and toxicological properties, is intertwined with tumorigenesis and progression. However, it is not fully understood how selenium metabolism affects immune response and cancer biology. METHODS: We estimated selenium metabolism by Gene Set Enrichment Analysis (GSEA) to delineate the selenium metabolism landscape using The Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx), Cancer Cell Line Encyclopedia (CCLE) and a integrated pan-cancer single-cell dataset. We systematically explored the prognostic implications of selenium metabolism and selenium-related regulatory patterns. The therapeutic value of selenium metabolism was explored through machine learning and examined in several immunotherapy cohorts. The heterogeneity and underlying mechanism of selenium metabolism were investigated by cell‒cell communication analysis at the single-cell level. RESULTS: A GSEA analysis based on 86 genes was used to evaluate the selenium metabolism landscape. The selenium metabolism score exhibited prognostic value in predicting the lower risk of mortality, possibly due to its correlation with multiple cancer hallmarks, including a positive correlation with complement (R = 0.761, P < 0.001), inflammatory response (R = 0.663, P < 0.001), apoptosis (R = 0.626, P < 0.001), hypoxia (R = 0.587, P < 0.001), reactive oxygen species (ROS) (R = 0.558, P < 0.001), and interferon gamma response (R = 0.539, P < 0.001). We also observed heterogeneity in the relationship between selenium metabolism and immunity across different cancers. Based on selenium-related genes, we constructed a machine learning model with area under the ROC curve (AUC) of 0.82 in predicting immune checkpoint inhibitor (ICI)-based immunotherapy response. Single-cell selenium metabolism quantification revealed that adjacent and tumor tissues had higher selenium metabolism compared with normal tissues, especially in epithelial cells, fibroblasts and macrophages. The communication between high-selenium epithelium and high-selenium fibroblast was significantly higher than other cells, especially in cytokines, chemokines, collagen, Wnt, VEGF, IGF and FGF pathways. CONCLUSION: Our study provides a comprehensive landscape of selenium metabolism levels and diverse regulatory patterns in different cancers, deepening the understanding of selenium's roles in tumorigenesis and immunity.

Colorectal cancer liver metastasis: genomic evolution and crosstalk with the liver microenvironment.

Colorectal cancer (CRC) patients frequently develop liver metastases, which are the major cause of cancer-related mortality. The molecular basis and management of colorectal liver metastases (CRLMs) remain a challenging clinical issue. Recent genomic evidence has demonstrated the liver tropism of CRC and the presence of a stricter evolutionary bottleneck in the liver as a target organ compared to lymph nodes. This bottleneck challenging CRC cells in the liver is organ-specific and requires adaptation not only at the genetic level, but also at the phenotypic level to crosstalk with the hepatic microenvironment. Here, we highlight the emerging evidence on the clonal evolution of CRLM and review recent insights into the molecular mechanisms orchestrating the bidirectional interactions between metastatic CRC cells and the unique liver microenvironment.

Deciphering the Clinical Significance and Kinase Functions of GSK3α in Colon Cancer by Proteomics and Phosphoproteomics.

GSK3α and GSK3ß are two GSK3 isoforms with 84% overall identity and 98% identity in their catalytic domains. GSK3ß plays important roles in the pathogenesis of cancer, while GSK3α has long been considered a functionally redundant protein of GSK3ß. Few studies have specifically investigated the functions of GSK3α. In this study, unexpectedly, we found that the expression of GSK3α, but not GSK3ß, was significantly correlated with the overall survival of colon cancer patients in 4 independent cohorts. To decipher the roles of GSK3α in colon cancer, we profiled the phosphorylation substrates of GSK3α and uncovered 156 phosphosites from 130 proteins specifically regulated by GSK3α. A number of these GSK3α-mediated phosphosites have never been reported before or have been incorrectly identified as substrates of GSK3ß. Among them, the levels of HSF1S303p, CANXS583p, MCM2S41p, POGZS425p, SRRM2T983p, and PRPF4BS431p were significantly correlated with the overall survival of colon cancer patients. Further pull-down assays identified 23 proteins, such as THRAP3, BCLAF1, and STAU1, showing strong binding affinity to GSK3α. The interaction between THRAP3 and GSK3α was verified by biochemical experiments. Notably, among the 18 phosphosites of THRAP3, phosphorylation at S248, S253, and S682 is specifically mediated by GSK3α. Mutation of S248 to D (S248D), which mimics the effect of phosphorylation, obviously increased cancer cell migration and the binding affinity to proteins related to DNA damage repair. Collectively, this work not only discloses the specific function of GSK3α as a kinase but also suggests GSK3α as a promising therapeutic target for colon cancer.

Phenotypic Heterogeneity Analysis of APC-Mutant Colon Cancer by Proteomics and Phosphoproteomics Identifies RAI14 as a Key Prognostic Determinant in East Asians and Westerners.

Adenomatous polyposis coli (APC) is an important tumor suppressor and is mostly linked to the regulation of the Wnt/ß-catenin signaling pathway. APC mutation has been identified as an early event in more than 80% of sporadic colorectal cancers (CRCs). Moreover, prognostic differences are observed in CRC patients with APC mutations. Although previous genomics studies have investigated the roles of concomitant gene mutations in determining the phenotypic heterogeneity of APC-mutant tumors, valuable prognostic determinants for APC-mutant CRC patients are still lacking. Based on the proteome and phosphoproteome data, we classified APC-mutant colon cancer patients and revealed genomic, proteomic, and phosphoproteomic heterogeneity in APC-mutant tumors. More importantly, we identified RAI14 as a key prognostic determinant for APC-mutant but not APC-wildtype colon cancer patients. The heterogeneity and the significance of prognostic biomarkers in APC-mutant tumors were further validated in the Clinical Proteomic Tumor Analysis Consortium (CPTAC) colon cancer cohort. In addition, we found that colon cancer patients with high expression of RAI14 were less responsive to chemotherapy. Knockdown of RAI14 in cell lines led to reduced cell migration and changes in epithelial-mesenchymal transition (EMT)-related markers. Mechanistically, knockdown of RAI14 remodeled the phosphoproteome associated with cell adhesion, which might affect EMT marker expression and promote F-actin degradation. Collectively, this work describes the phenotypic heterogeneity of APC-mutant tumors and identifies RAI14 as an important prognostic determinant for APC-mutant colon cancer patients. The prognostic utility of RAI14 in APC-mutant colon cancer will provide early warning and increase the chance of successful treatment.

Amino acid metabolic reprogramming in tumor metastatic colonization.

Metastasis is considered as the major cause of cancer death. Cancer cells can be released from primary tumors into the circulation and then colonize in distant organs. How cancer cells acquire the ability to colonize in distant organs has always been the focus of tumor biology. To enable survival and growth in the new environment, metastases commonly reprogram their metabolic states and therefore display different metabolic properties and preferences compared with the primary lesions. For different microenvironments in various colonization sites, cancer cells must transfer to specific metabolic states to colonize in different distant organs, which provides the possibility of evaluating metastasis tendency by tumor metabolic states. Amino acids provide crucial precursors for many biosynthesis and play an essential role in cancer metastasis. Evidence has proved the hyperactivation of several amino acid biosynthetic pathways in metastatic cancer cells, including glutamine, serine, glycine, branched chain amino acids (BCAAs), proline, and asparagine metabolism. The reprogramming of amino acid metabolism can orchestrate energy supply, redox homeostasis, and other metabolism-associated pathways during cancer metastasis. Here, we review the role and function of amino acid metabolic reprogramming in cancer cells colonizing in common metastatic organs, including lung, liver, brain, peritoneum, and bone. In addition, we summarize the current biomarker identification and drug development of cancer metastasis under the amino acid metabolism reprogramming, and discuss the possibility and prospect of targeting organ-specific metastasis for cancer treatment.

circNEIL3 inhibits tumor metastasis through recruiting the E3 ubiquitin ligase Nedd4L to degrade YBX1.

Distant metastasis is a major contributor to cancer-related mortality. However, the role of circRNAs in this process remains unclear. Herein, we profiled the circRNA expression in a cohort of 68 colorectal carcinoma (CRC) primary tumors and their paired liver metastatic lesions. By overlapping with the TGFß-responsive circRNAs, circNEIL3 (hsa_circ_0001460) was identified as a TGFß-repressive and metastasis-related circRNA. Functionally, circNEIL3 effectively inhibited tumor metastasis in both and in vivo and in vivo models of various cancer types. Mechanistically, circNEIL3 exerts its metastasis-repressive function through its direct interaction with oncogenic protein, Y-box-binding protein 1 (YBX1), which consequently promotes the Nedd4L-mediated proteasomal degradation of YBX1. Importantly, circNEIL3 expression was negatively correlated to YBX1 protein level and metastatic tendency in CRC patient samples. Collectively, our findings indicate the YBX1-dependent antimetastatic function of circNEIL3 and highlight the potential of circNEIL3 as a biomarker and therapeutic option in cancer treatment.

Targeting PSAT1 to mitigate metastasis in tumors with p53-72Pro variant.

The single-nucleotide polymorphism (SNP) of p53, in particular the codon 72 variants, has recently been implicated as a critical regulator in tumor progression. However, the underlying mechanism remains elusive. Here we found that cancer cells carrying codon 72-Pro variant of p53 showed impaired metastatic potential upon serine supplementation. Proteome-wide mapping of p53-interacting proteins uncovered a specific interaction of the codon 72 proline variant (but not p5372R) with phosphoserine aminotransferase 1 (PSAT1). Interestingly, p5372P-PSAT1 interaction resulted in dissociation of peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α) that otherwise bound to p5372P, leading to subsequent nuclear translocation of PGC-1α and activation of oxidative phosphorylation (OXPHOS) and tricarboxylic acid (TCA) cycle. Depletion of PSAT1 restored p5372P-PGC-1α interaction and impeded the OXPHOS and TCA function, resulting in mitochondrial dysfunction and metastasis suppression. Notably, pharmacological targeting the PSAT1-p5372P interaction by aminooxyacetic acid (AOA) crippled the growth of liver cancer cells carrying the p5372P variant in both in vitro and patient-derived xenograft models. Moreover, AOA plus regorafenib, an FDA-proved drug for hepatocellular carcinoma and colorectal cancer, achieved a better anti-tumor effect on tumors carrying the p5372P variant. Therefore, our findings identified a gain of function of the p5372P variant on mitochondrial function and provided a promising precision strategy to treat tumors vulnerable to p5372P-PSAT1 perturbation.

PHGDH arginine methylation by PRMT1 promotes serine synthesis and represents a therapeutic vulnerability in hepatocellular carcinoma.

Serine synthesis is crucial for tumor growth and survival, but its regulatory mechanism in cancer remains elusive. Here, using integrative metabolomics and transcriptomics analyses, we show a heterogeneity between metabolite and transcript profiles. Specifically, the level of serine in hepatocellular carcinoma (HCC) tissues is increased, whereas the expression of phosphoglycerate dehydrogenase (PHGDH), the first rate-limiting enzyme in serine biosynthesis pathway, is markedly downregulated. Interestingly, the increased serine level is obtained by enhanced PHGDH catalytic activity due to protein arginine methyltransferase 1 (PRMT1)-mediated methylation of PHGDH at arginine 236. PRMT1-mediated PHGDH methylation and activation potentiates serine synthesis, ameliorates oxidative stress, and promotes HCC growth in vitro and in vivo. Furthermore, PRMT1-mediated PHGDH methylation correlates with PHGDH hyperactivation and serine accumulation in human HCC tissues, and is predictive of poor prognosis of HCC patients. Notably, blocking PHGDH methylation with a TAT-tagged nonmethylated peptide inhibits serine synthesis and restrains HCC growth in an HCC patient-derived xenograft (PDX) model and subcutaneous HCC cell-derived xenograft model. Overall, our findings reveal a regulatory mechanism of PHGDH activity and serine synthesis, and suggest PHGDH methylation as a potential therapeutic vulnerability in HCC.

Disrupting metformin adaptation of liver cancer cells by targeting the TOMM34/ATP5B axis.

Metformin, a well-known antidiabetic drug, has been repurposed for cancer treatment; however, recently observed drug resistance and tumor metastasis have questioned its further application. Here, we found that long-term metformin exposure led to metabolic adaptation of hepatocellular carcinoma (HCC) cells, which was characterized by an obvious epithelial-mesenchymal transition (EMT) phenotype and compensatory elevation of oxidative phosphorylation (OXPHOS). TOMM34, a translocase of the outer mitochondrial membrane, was upregulated to promote tumor metastasis in response to metformin-induced metabolic stress. Mechanistically, TOMM34 interacted with ATP5B to preserve F1 FO -ATPase activity, which conferred mitochondrial OXPHOS and ATP production. This metabolic preference for OXPHOS suggested a large requirement of energy supply by cancer cells to survive and spread in response to therapeutic stress. Notably, disturbing the interaction between TOMM34 and ATP5B using Gboxin, a specific OXPHOS inhibitor, increased sensitivity to metformin and suppressed tumor progression both in vitro and in vivo. Overall, this study demonstrates a molecular link of the TOMM34/ATP5B-ATP synthesis axis during metformin adaptation and provides promising therapeutic targets for metformin sensitization in cancer treatment.

Pan-cancer single-cell analysis reveals the heterogeneity and plasticity of cancer-associated fibroblasts in the tumor microenvironment.

Cancer-associated fibroblasts (CAFs) are the predominant components of the tumor microenvironment (TME) and influence cancer hallmarks, but without systematic investigation on their ubiquitous characteristics across different cancer types. Here, we perform pan-cancer analysis on 226 samples across 10 solid cancer types to profile the TME at single-cell resolution, illustrating the commonalities/plasticity of heterogenous CAFs. Activation trajectory of the major CAF types is divided into three states, exhibiting distinct interactions with other cell components, and relating to prognosis of immunotherapy. Moreover, minor CAF components represent the alternative origin from other TME components (e.g., endothelia and macrophages). Particularly, the ubiquitous presentation of endothelial-to-mesenchymal transition CAF, which may interact with proximal SPP1+ tumor-associated macrophages, is implicated in endothelial-to-mesenchymal transition and survival stratifications. Our study comprehensively profiles the shared characteristics and dynamics of CAFs, and highlight their heterogeneity and plasticity across different cancer types. Browser of integrated pan-cancer single-cell information is available at https://gist-fgl.github.io/sc-caf-atlas/ .

Hypoxia-induced lncRNA STEAP3-AS1 activates Wnt/ß-catenin signaling to promote colorectal cancer progression by preventing m6A-mediated degradation of STEAP3 mRNA.

BACKGROUND: Hypoxia, a typical hallmark of solid tumors, exhibits an essential role in the progression of colorectal cancer (CRC), in which the dysregulation of long non-coding RNAs (lncRNAs) is frequently observed. However, the underlying mechanisms are not clearly defined. METHODS: The TCGA database was analyzed to identify differential lncRNA expression involved in hypoxia-induced CRC progression. qRT-PCR was conducted to validate the upregulation of lncRNA STEAP3-AS1 in CRC cell lines and tumor-bearing mouse and zebrafish models under hypoxia. ChIP-qRT-PCR was used to detect the transcriptional activation of STEAP3-AS1 mediated by HIF-1α. RNA-seq, fluorescent in situ hybridization, RNA pulldown, RNA immunoprecipitation, co-immunoprecipitation, immunofluorescence and immunoblot experiments were used to ascertain the involved mechanisms. Functional assays were performed in both in vitro and in vivo models to investigate the regulatory role of STEAP3-AS1/STEAP3/Wnt/ß-catenin axis in CRC proliferation and metastasis. RESULTS: Here, we identified a hypoxia-induced antisense lncRNA STEAP3-AS1 that was highly expressed in clinical CRC tissues and positively correlated with poor prognosis of CRC patients. Upregulation of lncRNA STEAP3-AS1, which was induced by HIF-1α-mediated transcriptional activation, facilitated the proliferation and metastasis of CRC cells both in vitro and in vivo. Mechanistically, STEAP3-AS1 interacted competitively with the YTH domain-containing family protein 2 (YTHDF2), a N6-methyladenosine (m6A) reader, leading to the disassociation of YTHDF2 with STEAP3 mRNA. This effect protected STEAP3 mRNA from m6A-mediated degradation, enabling the high expression of STEAP3 protein and subsequent production of cellular ferrous iron (Fe2+). Increased Fe2+ levels elevated Ser 9 phosphorylation of glycogen synthase kinase 3 beta (GSK3ß) and inhibited its kinase activity, thus releasing ß-catenin for nuclear translocation and subsequent activation of Wnt signaling to support CRC progression. CONCLUSIONS: Taken together, our study highlights the mechanisms of lncRNA STEAP3-AS1 in facilitating CRC progression involving the STEAP3-AS1/STEAP3/Wnt/ß-catenin axis, which may provide novel diagnostic biomarkers or therapeutic targets to benefit CRC treatment. Hypoxia-induced HIF-1α transcriptionally upregulates the expression of lncRNA STEAP3-AS1, which interacts competitively with YTHDF2, thus upregulating mRNA stability of STEAP3 and consequent STEAP3 protein expression. The enhanced STEAP3 expression results in production of cellular ferrous iron (Fe2+), which induces the Ser 9 phosphorylation and inactivation of GSK3ß, releasing ß-catenin for nuclear translocation and contributing to subsequent activation of Wnt signaling to promote CRC progression.

Association of recurrent APOBEC3B alterations with the prognosis of gastric-type cervical adenocarcinoma.

OBJECTIVE: Gastric-type cervical adenocarcinoma (GCA) is a rare and aggressive type of endocervical adenocarcinoma (ECA) with distinct histopathologic features and unfavorable treatment outcomes, but no genomic prognostic factor has been revealed. We aimed to systematically investigate the somatic alterations of GCA at genome-wide level and evaluate their prognostic value. METHODS: We performed whole-exome sequencing (WES) on 25 pairs of tumor and matched normal samples to characterize the genomic features of Chinese patients with GCA and investigated their relations to histopathological characterizations and prognosis. The prognostic value of the genomic alterations was evaluated in a total of 58 GCA patients. RESULTS: Mutations were commonly observed in reported GCA-related driver genes, including TP53 (32%), CDKN2A (20%), SKT11 (20%), BRCA2 (12%), SMAD4 (12%), and ERBB2 (12%). Recurrent novel trunk mutations were also observed in PBRM1 (12%), FRMPD4 (12%), and NOP2 (8%) with high variant allele frequency. Moreover, enrichment of the APOBEC signature was attributed to frequent gain of somatic copy number alteration (SCNA) of APOBEC3B (20%), which perfectly matched the nuclear-positive staining of APOBEC3B through immunohistochemistry. In contrast, APOBEC3B alteration was absent in patients with conventional type of ECA (N = 52). Notably, positive APOBEC3B was consistently enriched in patients with favorable prognosis in both the discovery cohort and an additional 33 GCA patients, thus indicating a significant association with lower relapse risk of GCA independent of cancer stage (P = 0.02). CONCLUSION: Our results can aid understanding of the molecular basis of GCA in the Chinese population by providing genomic profiles and highlighting the potential prognostic value of APOBEC3B for GCA through routine clinical IHC.

Inhibition of NPC1L1 disrupts adaptive responses of drug-tolerant persister cells to chemotherapy.

Entering a drug-tolerant persister (DTP) state of cancer cells is a transient self-adaptive mechanism by which a residual cell subpopulation accelerates tumor progression. Here, we identified the acquisition of a DTP phenotype in multidrug-resistant (MDR) cancer cells as a tolerance response to routine combination treatment. Characterization of MDR cancer cells with a DTP state by RNA-seq revealed that these cells partially prevented chemotherapy-triggered oxidative stress by promoting NPC1L1-regulated uptake of vitamin E. Treatment with the NPC1L1 inhibitor ezetimibe further enhanced the therapeutic effect of combinatorial therapy by inducing methuosis. Mechanistically, we demonstrated that NRF2 was involved in transcriptional regulation of NPC1L1 by binding to the -205 to -215 bp site on its promoter. Decreased DNA methylation was also related partially to this process. Furthermore, we confirmed that a triple-combination of chemotherapeutic agents, verapamil, and ezetimibe, had a significant anti-tumor effect and prevented tumor recurrence in mice. Together, our study provides a novel insight into the role of DTP state and emphasizes the importance of disrupting redox homeostasis during cancer therapy.

Genomic evolution and diverse models of systemic metastases in colorectal cancer.

OBJECTIVE: The systemic spread of colorectal cancer (CRC) is dominated by the portal system and exhibits diverse patterns of metastasis without systematical genomic investigation. Here, we evaluated the genomic evolution of CRC with multiorgan metastases using multiregion sequencing. DESIGN: Whole-exome sequencing was performed on multiple regions (n=74) of matched primary tumour, adjacent non-cancerous mucosa, liver metastasis and lung metastasis from six patients with CRC. Phylogenetic reconstruction and evolutionary analyses were used to investigate the metastatic seeding pattern and clonal origin. Recurrent driver gene mutations were analysed across patients and validated in two independent cohorts. Metastatic assays were performed to examine the effect of the novel driver gene on the malignant behaviour of CRC cells. RESULTS: Based on the migration patterns and clonal origins, three models were revealed (sequential, branch-off and diaspora), which not only supported the anatomic assumption that CRC cells spread to lung after clonally expanding in the liver, but also illustrated the direct seeding of extrahepatic metastases from primary tumours independently. Unlike other cancer types, polyphyletic seeding occurs in CRC, which may result in late metastases with intermetastatic driver gene heterogeneity. In cases with rapid dissemination, we found recurrent trunk loss-of-function mutations in ZFP36L2, which is enriched in metastatic CRC and associated with poor overall survival. CRISPR/Cas9-mediated knockout of ZFP36L2 enhances the metastatic potential of CRC cells. CONCLUSION: Our results provide genomic evidence for metastatic evolution and indicate that biopsy/sequencing of metastases may be considered for patients with CRC with multiorgan or late postoperative metastasis.