Immune cell infiltration is important for predicting the clinical outcomes of colorectal cancer. Integrin β7 (ITGB7), which is expressed on the surface of leukocytes, plays an essential role in the homing of immune cells to gut-associated lymphoid tissue and facilitating the retention of lymphocytes in gut epithelium; however, its role in colorectal cancer pathogenesis is poorly explored. Here, we found that the number of β7+ cells decreased significantly in tumor tissue compared with adjacent normal tissue. β7 expression decreased in tumor-derived compared with normal tissue–derived CD8+ T cells. With bulk RNA expression data from public platforms, we demonstrated that higher ITGB7 expression correlated with longer patient survival, higher cytotoxic immune cell infiltration, lower somatic copy-number alterations, decreased mutation frequency of APC and TP53, and better response to immunotherapy. The possible cell–cell interactions mediated by ITGB7 and its ligands MAdCAM-1, VCAM-1, and CDH1 were investigated using public single-cell RNA sequencing data. ITGB7 deficiency led to exaggerated tumorigenesis and progression in both Apcmin/+ spontaneous and MC38 orthotopic models of colorectal cancer, which could be due to a reduced infiltration of activated CD8+ T cells, effector memory CD8+ T cells, IFNγ+ CD8+ T cells, IFNγ+ natural killer cells, CD103+ dendritic cells, and other immune cell subsets that are essential players in antitumor immunity. In conclusion, our data revealed that ITGB7 could inhibit the tumorigenesis and progression of colorectal cancer by maintaining antitumor immunity.

Colorectal cancer is the third most prevalent cancer worldwide and one of the major causes of cancer-related death (1, 2). With improvements in surgery, radiotherapy, chemotherapy, and immunotherapy, the survival of patients with colorectal cancer has greatly increased, with the 5-year survival rate reaching over 50% in most regions worldwide (3). However, the prediction of survival outcomes in colorectal cancer remains challenging owing to the complexity and heterogenicity of this disease. Thus, there is an urgent need to investigate the molecular mechanisms underlying colorectal cancer pathogenesis to identify novel prognostic biomarkers and potential therapeutic target.

Increasing evidence has shown that abundant immune cell infiltration is associated with improved outcomes in various cancers (4–6). The prognostic marker Immunoscore, which quantifies CD3+ and CD8+ T cells at both the tumor center and invasive margin, has been established to predict the clinical outcomes of colorectal cancer (7, 8). Immune cell infiltration has been shown to have superior prognostic value over the traditional tumor–node–metastasis staging system (8, 9). However, Fakih and colleagues identified a risk population of patients with colorectal cancer who showed abundant infiltration of CD8+ T cells but still had a poor clinical outcome (10). These findings illustrate the complex interaction between tumors and their immunologic microenvironment.

Integrins are heterodimeric cell adhesion molecules composed of α and β subunits, mediating cell–cell, cell–extracellular matrix (ECM), and cell–pathogen interactions (11). There are 24 distinct integrins in vertebrates, comprising 18 α-subunits and 8 β-subunits, which play vital roles in biological events and diseases (12). Integrin subunit β7 (ITGB7) can bind with the α4 subunit to form the α4β7 heterodimer, which mediates lymphocyte homing to gut-associated lymphoid tissue via interaction with the mucosal addressin cell adhesion molecule-1 (MAdCAM-1) on the intestinal vasculature (13), or with the αE subunit to form αEβ7, which facilitates the retention of lymphocytes in the gut epithelium via binding to E-cadherin (14). The role of ITGB7 has been demonstrated in the pathogenesis of inflammatory bowel disease (15, 16); however, the role of ITGB7 in the pathogenesis of colorectal cancer and whether it can serve as a potential prognostic marker remains poorly explored.

Herein, we comprehensively examined the expression pattern of ITGB7 and its correlation with survival, tumor-infiltrating immune cells, somatic copy-number alterations (SCNA) and mutations, and responses to immunotherapy using tissue samples from patients with colorectal cancer, bulk RNA expression data, and single-cell RNA sequencing (scRNA-seq) data from public platforms. We further used two colorectal tumor models, the Apcmin/+ spontaneous and the MC38 orthotopic models, to investigate the role of ITGB7 in the pathogenesis of colorectal cancer and verified its correlation with tumor-infiltrating immune cells in vivo. We revealed a central inhibitory role for ITGB7 in colorectal cancer tumorigenesis and progression. ITGB7 supports antitumor immunity by maintaining the infiltration of various subsets of immune cells that are essential for tumor immunosurveillance.

Patients and cohort datasets

Primary tumor tissues and adjacent normal tissues were obtained from patients with colorectal adenocarcinoma who underwent a surgical resection in Shanghai Tenth People's Hospital (Shanghai, P.R. China). A total of 118 formalin-fixed, paraffin-embedded specimens (one from each case) collected between 2008 and 2013 were used to construct the tissue microarray (SHSY cohort). In addition, 22 pairs of tumor and normal adjacent tissues, which were collected between 2018 and 2019, were collected at the time of surgical resection and were immediately snap frozen in liquid nitrogen before storage at −80°C. Exclusion criteria were: (i) radiotherapy or chemotherapy before surgery, (ii) presence of hereditary or inflammation‐associated colorectal cancer, (iii) presence of mucinous adenocarcinomas or other pathologic complications, (iv) incomplete follow‐up information, (v) non‐colorectal cancer–related cause of death, and (vi) poor quality of the tissue. Blood from three healthy donors were collected for T-cell isolation and culture. All tissue and blood samples were collected from participants with appropriate informed written consent. The study was approved by the Institutional Review Board of the Tenth People's Hospital affiliated with Tongji University (2017-KN22-03) and was performed in accordance with the ethical standards of the Declaration of Helsinki.

RNA expression and clinical data were downloaded from The Cancer Genome Atlas (TCGA) using R package TCGAbiolinks (17). The fragments per kilobase million gene expression data of TCGA Colon Adenocarcinoma (TCGA-COAD) and TCGA Rectum Adenocarcinoma (TCGA-READ) were converted to transcripts per million (TPM) after removing duplicated genes and zero expression genes. Sample type was not “Primary Tumor,” and samples with incomplete overall survival information and samples that were formalin fixed, paraffin embedded were excluded. A total of 613 tumor samples and 51 normal samples were included in the study and used to construct the TCGA Colorectal Cancer (TCGA-CRC) cohort.

Microarray data, with at least 100 samples of patients with colorectal cancer, were downloaded from Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) with the following accession numbers: GSE14333 (n = 290), GSE17536 (n = 177), GSE21815 (n = 132), GSE37892 (n = 130), and GSE39582 (n = 566). Samples with “primary adenocarcinoma” were included. The scRNA-seq data were also downloaded from cohort GSE146771 (n = 10, Smart-seq2 samples). Characteristics of the SHSY, TCGA, and GEO cohorts are summarized in Supplementary Table S1.

Immunotherapy datasets

Three transcriptomic datasets were used to study the association between ITGB7 expression and response to immunotherapy: lung cancer (GSE136961, anti–PD-1, n = 21), urothelial cancer (IMvigor210, anti–PD-L1, n = 298), and a mouse study (GSE117358, anti-CTLΑ4/anti–PD-L1, n = 24 for AB1 mesothelioma model, n = 24 for Renca kidney cancer model). Samples with both survival information and gene expression data were included in the human studies. All samples were included in the mouse study. The datasets for lung cancer and the mouse study were downloaded from GEO database with the assigned accession number. For the urothelial cancer dataset, a fully documented software and data package is freely available under the Creative Commons 3.0 license and can be downloaded from http://research-pub.gene.com/IMvigor210CoreBiologies.

Immunofluorescence staining

Frozen sections of human colorectal adenocarcinoma tissues and adjacent normal tissues were incubated overnight with anti-human ITGB7 (20 μg/mL) purified from hybridoma FIB504.64 (ATCC HB-293) at 4°C, and then incubated with FITC-conjugated goat anti-rat IgG (2 μg/mL; Invitrogen, 31629) for 1 hour at room temperature. For β7+ cell quantification, five random fields of the intratumoral region or the normal tissue region were photographed (20× magnification) using a Leica TCS SP8 confocal microscope with LAS X software (Leica). The number of β7+ cells in each field was then counted, and the average of cell number per field was calculated for each slide. All slides were counted separately by two pathologists blinded to the clinical information.

Tumor-infiltrating immune cell isolation

Tumor and normal tissue samples from patients with colorectal cancer and tumor tissue from mice bearing MC38 orthotopic tumors were dissected, weighed, minced, and incubated in RPMI1640 medium containing collagenase I (100 U/mL; Gibco, 17018029), collagenase IV (100 U/mL; Gibco, 17104019), DNase I (50 U/mL; Roche, 04716728001), and 5% FBS for 1 hour at 37°C. The resulting single-cell suspensions were passed through a 70-μm cell strainer and washed with PBS. Immune cells were enriched with a 40%:80% Percoll (GE Healthcare) gradient centrifugation at 700 × g for 25 minutes at 25°C. Cells in the interphase were washed and resuspended in PBS + 2% BSA (Amresco, 0332) for flow cytometry.

Flow cytometry

Single immune cell suspensions from patients with colorectal cancer and mouse were stained with fluorochrome-conjugated mAbs in PBS supplemented with 2% BSA for 30 minutes at 4°C. Samples were incubated with 7-AAD viability staining solution (Invitrogen, 00-6993-50) for 5 minutes at room temperature at a concentration of 5 μL/test before acquiring using a CytoFLEX LX cytometer (Beckman). The data were analyzed using FlowJo v10. For intracellular IFNγ detection, 1 × 107 isolated tumor-infiltrating immune cells were stimulated with phorbol 12-myristate 13-acetate (20 ng/mL; Sigma), ionomycin (1 μg/mL; Cell Signaling Technology, 9995), and 1× brefeldin A (eBioscience, 00-4506-51) for 4 hours at 37°C. Cells were then washed, fixed by 2% paraformaldehyde, permeabilized (BD Perm/Wash, 554723), and stained with rat anti-mouse IFNγ or isotype control (eBioscience, 12-4714-82) antibody for 30 minutes at 4°C according to the BD intracellular staining protocol. Antibodies used in human study were as following: anti-CD45 (eBioscience, 25-0451-82), anti-CD4 (BD Pharmingen, 562970), anti-CD8 (BD Pharmingen, 555366), anti-CD19 (BD Pharmingen, 566396), anti-CD56 (BD Pharmingen, 555516), anti-CD11c (BD Pharmingen, 562561), anti-CD68 (eBioscience, 11-0689-42), and anti-Integrin αE (BioLegend, 350227). Anti-α4β7 was purified from the culture supernatants of hybridoma alpha-4/beta-7, ACT-1 #39 (ATCC PTA-3663). Antibodies used in mouse study were as following: anti-CD45 (Clone 30-F11, BD Pharmingen, 557659), anti-CD11c (Clone N418, eBioscience, 17-0114-82), anti-MHC II (Clone 2G9, BD OptiBuild, 743871), anti-CD11b (Clone M1/70, eBioscience, 12-0112-81), anti-CD103 (Clone 2E7, BioLegend, 121419), anti-CD45 (Clone 30-F11, BioLegend, 103133), anti-CD4 (Clone RM4–5, BD Pharmingen, 553046), anti-CD8 (Clone 53-6.7, eBioscience, 45-0081-80), anti-NK1.1 (Clone PK136, BioLegend, 108723), anti-IFNγ (Clone 4S.B3, eBioscience, 12-7319-42), anti-α4β7 (Clone DATK32, BD Pharmingen, 553811), anti-CD45.1 (Clone A20, BD Pharmingen, 553776), anti-CD45.2 (Clone 104, BioLegend, 109823). Gating strategies with cell population-defining markers used for flow cytometry data analysis are provided in Supplementary Fig. S1.

IHC

Tissue microarray containing 118 tissue cores were cut into 4-μm slides. The slides were deparaffinized in xylene and rehydrated in graded alcohols and distilled water. Antigen retrieval was achieved by microwaving in antigen unmasking solution (Vector Labs, H-3300) for 20 minutes. Endogenous peroxidase activity was blocked with BLOXALL Blocking Solution (Vector Labs, SP-6000) at room temperature for 10 minutes. The slides were blocked with 10% goat serum for 1 hour at room temperature. Slides were then incubated with primary antibody against β7 (20 μg/mL; FIB504.64) overnight in a humidity chamber at 4°C, and then biotinylated goat anti-rat IgG (Sino Biological, SSA012) for 1 hour at room temperature. Slides were stained using Vector Lab ABC Reagent (Vector Labs, PK-6200) and DAB Substrate Kit (Vector Labs, SK-4100). All slides were counterstained with hematoxylin. Whole slides were scanned on Pannoramic 250/MIDI (3DHISTECH) and viewed using CaseViewer 2.0 software (3DHISTECH). The number of β7+ cells in each tissue core was counted.

Survival analysis

Log-rank statistics were used to identify the optimal cutoff for transforming the continuous variable of ITGB7+ cell numbers or ITGB7 gene expression into categorical high- and low-expression groups using the surv_cutpoint function of R package “survminer” (https://rpkgs.datanovia.com/survminer/index.html), as used in other studies (10, 18). The cutoff point with the highest test score was applied for separating patients into ITGB7hi and ITGB7lo groups with different risks. The optimal cutoff value for ITGB7+ cell number and ITGB7 gene expression was 14 and 1.285, respectively. Overall survival was estimated with the Kaplan–Meier method, and log-rank (Mantel–Cox) test was used to compare the ITGB7hi and ITGB7lo groups. The analysis was conducted using R package “survival” (https://CRAN.R-project.org/package=survival).

Immune cell infiltration analysis

The relative infiltration of 36 immune cell types for each sample was quantified by the single-sample gene set enrichment analysis (ssGSEA) method using the GSVA package in R. Feature gene panels for the 36 immune cell type were collected from two publications (19, 20). The relative abundance of each immune cell type was represented by an enrichment score from the ssGSEA. Finally, Spearman correlation coefficients between the ssGSEA score and ITGB7 RNA expression was calculated using the “cor.test” function in the R package “stats.”

Differentially expressed gene analysis and Gene Ontology enrichment analysis

R package “DESeq2” was used to identify the differentially expressed genes (DEG) between ITGB7hi and ITGB7lo samples. The criterion of statistical significance was FDR < 0.05 and |log2 (FC)| > 1. Gene Ontology (GO) function enrichment analysis of identified DEGs was performed using R package “clusterProfiler” (21). P < 0.05 was considered a statistically significant difference.

Gene set enrichment analysis

The gene set enrichment analysis (GSEA) was performed to illustrate the different enriched biological pathways between ITGB7hi and ITGB7lo groups of TCGA-CRC cohort. GSEA was performed by R package “clusterProfiler” (21). C2.cp.kegg.v7.1.symbols.gmt file was used as the reference gene set file. The threshold was set at P < 0.05.

Cell–cell interaction analysis

scRNA-seq data of colorectal cancer was downloaded from GEO with accession number GSE146771. The expression of ITGB7 and the cytotoxic genes PRF1, IFNG, GZMA, GZMB for B cells, T cells, innate lymphoid cells (ILC), and myeloid cells was analyzed at http://crcleukocyte.cancer-pku.cn/. R package “CellChat” (22) was used to investigate the interactions between immune and nonimmune cells, including malignant cells, fibroblasts, and endothelial cells.

SCNA analysis

The segment file of TCGA-CRC was downloaded from TCGA database (https://tcga-data.nci.nih.gov/) and GISTIC2.0 (23) was used to identify statistically significant broad and focal SCNAs. Focal SCNAs were defined as deletions or amplifications involving a region smaller than 50% of a chromosome arm. The SCNAs for each tumor sample was determined following a modified protocol from a previous study (24), in which the chromosome SCNA level and arm SCNA level were calculated separately; however, in our study, the two were summed up to obtain broad SCNA level. For each patient, each SCNA region was assigned a gistic score G obtained from the GISTIC2.0 analysis. Thus, the broad and focal SCNAs for each tumor sample was calculated as follows:

formula
formula

The BroadL, FocalL value for each tumor sample was further z-score normalized using the “scale” function in R package “base.” The Spearman correlation coefficients between SCNAs and ITGB7 RNA expression was calculated using the “cor.test” function in R package “stats.”

Mutation and neoantigen analysis

The somatic mutations for the TCGA-CRC cohort were downloaded by R package “TCGAmutations,” and the mutation landscape was analyzed by the R package “maftools.” The potential neoantigens of TCGA-CRC cohort, predicted by NetMHCpan 2.8, was downloaded from TSNAdb (ref. 25; http://biopharm.zju.edu.cn/tsnadb/).

Cell lines

The murine colon carcinoma cell line MC38 was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, P.R. China) in 2019 and authenticated by short tandem repeat fingerprinting. MC38 cells were maintained in RPMI1640 medium supplemented with 10% FBS and penicillin-streptomycin (100 U/mL; all from Gibco). Cells that were within two to three passages after thawing and tested negative for Mycoplasma contamination were used in in vivo experiments.

Animals and tumor models

Itgb7−/− C57BL/6J mice were from Jackson Laboratory. C57BL/6J wild-type (WT) and Apcmin/+ mice were from Shanghai Model Organisms Center (Shanghai, P.R. China). Two mouse models of colorectal adenocarcinoma were used in this study. In the MC38 orthotopic tumor studies, 8-week-old female C57BL/6J WT and Itgb7−/− mice were anesthetized by intraperitoneal injection of 2.5% tribromoethanol (Sigma, T48402) in saline with a dose of 0.1 mL/10 g body weight. Then, 10 μL containing 1 × 106 MC38 tumor cells were injected into the colonic subserosa using a 33-gauge micro-injector (Hamilton Company). Mice were euthanized 2 weeks after implantation and tumors were dissected and collected from the colon for volume evaluation, hematoxylin and eosin (H&E) staining, and flow cytometry analysis. In the Apcmin/+ mice studies, male Apcmin/+ mice were crossed with female Itgb7−/− mice to obtain Itgb7−/−Apcmin/+ mice. The mice were euthanized at 4 months old, and intestines and colons harvested. The number of tumors was counted, according to their location along the gastrointestinal tract, and tumor size was also assessed. Tumors were then dissected and collected for H&E staining. All animals were kept under specific pathogen–free conditions with a 12-hour light/dark cycle. All animal studies were approved by the Institutional Animal Care and Use Committee of Tongji University (Shanghai, P.R. China).

Tissue assessment

Volume of MC38 orthotopic tumors was measured with a digital vernier calliper and calculated using the formula: tumor volume = length × width2 × 0.5. Diameter of Apcmin/+ tumors was measured with a digital vernier calliper. Tumors were categorized into three types according to the length of diameter: < 1 mm, 1 to 2 mm, >2 mm. Tumor morphology was examined using routine H&E staining. Images were taken using an Olympus BX51 microscope.

Competitive in vivo homing assay

CD3+ T cells were isolated from WT CD45.1+ (Shanghai Model Organisms Center, Shanghai, P.R. China) and Itgb7−/− CD45.2+ spleens using the EasySep Mouse T-cell Isolation Kit (Stemcell, 19851) and then stimulated with immobilized anti-CD3 (2 μg/mL; BD Pharmingen, 555273), anti-CD28 (2 μg/mL; BD Pharmingen, 553294), and murine IL2 (10 μg/mL; Novoprotein, CK24) for 6 days to let the T cells expand. Equal numbers of WT and Itgb7−/− T cells (2 × 107 each) were mixed and intravenously administered into 12-week-old Rag1−/− recipient mice (Shanghai Model Organisms Center, Shanghai, P.R. China) bearing MC38 orthotopic tumors. Tumors, spleen, mesenteric lymph nodes, and small intestines were harvested 5 days after administration, and flow cytometry was used to analyze the homing of lymphocytes (as described above). The homing index was calculated as the (no. of CD45.2+ CD3+ cells)/(no. of CD45.1+ CD3+ cells) ratio.

Establishment of patient-derived organoids

Fresh tumor tissues dissected from patients with colorectal cancer were gently washed with ice-cold PBS and minced into 1 to 2 mm3 pieces. Tissues were digested with 10 mL gentle cell dissociation reagent (Stemcell, 07174) on a rocking platform for 30 minutes on ice. Dissociated cells were filtered through a 100-μm strainer, centrifuged, and washed with PBS. Cells were then resuspended in 50 μL growth factor reduced (GFR) Matrigel (Corning, 356231), seeded on 48-well flat-bottom cell culture plate (Corning, 3548), and solidified in a cell culture incubator for 30 minutes. A total of 300 μL culture medium [human IntestiCult Organoid Growth Medium (Stemcell, 06010) supplemented with Y27632 (Stemcell, 72304)] were overlaid in the well coated with Matrigel. Culture medium was replaced every 2 days. For passage, organoids were harvested with ice-cold PBS and pipetted with mechanical force through 1 mL pipette. Dissociated organoids were pelleted, washed with ice-cold PBS, and resuspended in GFR matrigel and reseeded on 48-well culture plate. Organoids were isolated from 3 patients with colorectal cancer.

Transwell assays

Organoids isolated from tumor tissue of patients with colorectal cancer (CRC-organoids) were cultured in the lower part of transwell chambers in RPMI1640 medium 24 hours prior to the assay. A total of 1 × 106 CD3+ T cells were isolated from healthy human peripheral blood using the EasySep Human T Cell Isolation Kit (Stemcell, 19661) and expanded with the stimulation of 25 μL ImmunoCult Human CD3/CD28 T Cell Activator (Stemcell, 10971) and 50 ng/mL human IL2 (PeproTech, 200-02) for 14 days. The purity of T cells was examined with flow cytometry, and cells >95% purity were used for the transwell assays. T cells were added into the upper part of the chambers with and without a β7 blockade antibody (10 μg/mL, FIB504.64). T cells were allowed to migrate for 5 hours, and the migrated cells were counted using flow cytometry (BD FACSCanto II) and analyzed by FlowJo v10. CountBright absolute counting beads (Invitrogen, C36950) was used to determine the absolute number of migrated cells according to the manufacturer's protocol.

CIBERSORTx analysis

To determine ITGB7 expression on different immune cell subtypes, gene expression data of TCGA-CRC cohort in TPM form was used as input for CIBERSORTx (https://cibersortx.stanford.edu/; ref. 26) with choosing “High-Resolution” as the expression analysis type and “LM22” as the signature matrix file. ITGB7 expression on CD8+ T cells from normal, ITGB7hi, and ITGB7lo groups were compared using Kruskal–Wallis test.

Statistical analysis

The normality of the variables was tested by the Shapir–Wilk normality test. For comparisons of two groups, statistical significance was determined using Student t test and Mann–Whitney U test (Wilcoxon rank-sum test) for normally distributed variables and nonnormally distributed variables, respectively. For comparisons of three or more groups, one-way ANOVA and Kruskal–Wallis test were used as parametric and nonparametric methods, respectively. Dunn test was used to perform the post hoc analysis for multiple comparisons following Kruskal–Wallis test. Correlation coefficients were calculated by Spearman correlation test. Two-sided Fisher exact tests were used to analyze contingency tables. The Kaplan–Meier method was used to generate survival curves for ITGB7hi and ITGB7lo groups, and the log-rank (Mantel–Cox) test was used to determine the statistical significance of differences. The R package pROC (50) was used to visualize ROC curves to calculate the AUC and confidence intervals to evaluate the diagnostic accuracy of ITGB7 expression. All statistical analyses were conducted using R (https://www.r-project.org/) or GraphPad Prism software (version 7.0), and the P values were two sided. P < 0.05 was considered statistically significant.

ITGB7 expression and its correlation with survival in patients with colorectal cancer

To explore the role of ITGB7 in the pathogenesis of colorectal cancer, we collected 22 pairs of tumor tissue and adjacent normal tissue from patients with colorectal cancer and evaluated the infiltration of β7+ cells. The number of β7+ cells was significantly lower in tumor tissues than in the adjacent normal tissues (Fig. 1A and B). Consistently, the expression of ITGB7 mRNA was significantly lower in tumors than in normal tissues in TCGA cohort (Fig. 1C). Because ITGB7 mediates the homing of immune cells to gut-associated lymphoid tissue (GALT), we next used multiparameter flow cytometry to investigate the infiltration of immune cells, including CD4+ T cells, CD8+ T cells, B cells, dendritic cells (DC), natural killer (NK) cells, and macrophages in tumor and adjacent normal tissues (Supplementary Fig. S1A). Among these immune cell compartments, the number of CD4+ T cells, CD8+ T cells, B cells, and NK cells decreased in tumors compared with normal tissues, whereas the number of DCs and macrophages showed no significant difference between tumor and normal tissues (Supplementary Fig. S1B).

We further examined the expression of α4β7 and αEβ7 in each compartment of infiltrating immune cells in tumor and normal tissues. α4β7 was expressed on nearly all immune cell compartments except for macrophages (Fig. 1D). The expression of α4β7 was comparable between tumor and normal tissues for CD4+ T cells, B cells, DCs, and NK cells; however, decreased α4β7 expression on CD8+ T cells was detected in tumor tissues (Fig. 1E). αEβ7 was expressed on CD8+ T cells and a small subset of DCs and NK cells (Fig. 1D), but expression of αEβ7 was decreased on tumor CD8+ T cells and NK cells; no significant alterations were seen on DCs (Fig. 1E).

To investigate the relationship between ITGB7 expression and survival of patients with colorectal cancer, we first selected 118 eligible patients diagnosed with colorectal cancer from 2008 to 2013 in our hospital (SHSY cohort) and assessed ITGB7+ cells by IHC staining of the tissue microarray block (Supplementary Fig. S2A). We identified the optimal cutoff for determining ITGB7hi and ITGB7lo cells (see Materials and Methods). A statistically significant difference in overall survival was observed between these two groups of patients; the group with the higher number of ITGB7+ cells had a better prognosis than the lower group (log‐rank, P = 0.011; Supplementary Fig. S2B). We next performed survival analysis in TCGA cohort. Similarly, the patients in the cohort were divided into two groups using the optimal cutoff for ITGB7 expression (see Materials and Methods). The ITGB7hi group showed better prognosis than the ITGB7lo group (log‐rank, P = 0.043; Supplementary Fig. S2C). Thus, ITGB7 expression associated with better overall survival in patients with colorectal cancer.

Next, we examined the expression patterns of ITGB7 among patients with colorectal cancer subtyped by different criteria. Among the four consensus molecular subtypes (CMS), CMS1 showed the highest expression of ITGB7, whereas CMS2 had the lowest (Supplementary Fig. S2D), which is consistent with CMS1 having the highest immune cell infiltration and better clinical outcome than the other subtypes (27). Among the four stages, ITGB7 expression was statistically significant between stage I and stage IV, and between stage II and stage IV (Supplementary Fig. S2E). There was no difference in ITGB7 mRNA expression between metastatic and nonmetastatic tissues (Supplementary Fig. S2F).

ITGB7 expression correlates with antitumor immunity

The tumor microenvironment is infiltrated by various kinds of immune cells, both protumor and antitumor. To determine the immune cell types that correlated with ITGB7 expression, we conducted ssGSEA to analyze immune cell infiltration in bulk gene expression data, including the TCGA, GSE14333 (28), GSE17536 (29), GSE21815 (30, 31), GSE37892 (32), GSE39582 (33) cohorts, and determined the correlation with ITGB7 expression. Among the immune cells studied, T cells, activated CD8+ T cells, CD8+ effector memory T cells, activated B cells, and cytotoxic cells showed a positive correlation with ITGB7 expression (coefficiency > 0.3 in all six cohorts and > 0.55 in TCGA cohort). These data indicate that ITGB7 expression associated with the abundance of cytotoxic cells (Fig. 2A). Consistently, ITGB7 expression positively correlated with the cytolytic activity (CYT) score (Fig. 2B), which is calculated by the geometrical mean of perforin 1 (PRF1) and granzyme A (GZMA; ref. 34).

DEG analysis of TCGA cohort identified 2,762 upregulated genes and 367 downregulated genes in the ITGB7hi group compared with the ITGB7lo group (Supplementary Fig. S3A). Upregulated DEGs were significantly enriched in the regulation of leukocyte activation, leukocyte differentiation, leukocyte migration, and T-cell activation, whereas downregulated DEGs were significantly enriched in mRNA processing, ncRNA metabolic process, and RNA splicing (Supplementary Fig. S3B). Further GSEA of TCGA cohort showed that, compared with the ITGB7lo group, the ITGB7hi group was enriched in the T-cell receptor signaling pathway, NK cell–mediated cytotoxicity, JAK–STAT signaling pathway, intestinal immune network for IgA production, cell adhesion molecules, and antigen processing and presentation (Fig. 2C).

We examined the expression of ITGB7 in immune cells in a Smart-seq2 scRNA-seq study GSE146771 (Fig. 2D; ref. 35). Among the four major immune cell clusters (T cells, ILCs, B cells, and myeloid cells), which includes total 36 subclusters, IgA+ B cells, CD103+ NK cells, BATF3+ cDC1, GNLY+CD4+ T cells, IL23R+CD4+ T cells, CD160+CD8+ T cells, LAYN+CD8+ T cells displayed ITGB7 expression (Fig. 2E). These ITGB7 high–expressing cells, except for IgA+ B cells and BATF3+ cDC1, displayed high expression of cytotoxic genes, including IFNγ (IFNG), PRF1, GZMA, and granzyme B (GZMB; Fig. 2F), which is consistent with the finding that ITGB7 expression positively correlated with the CYT score (Fig. 2B).

We then investigated the cell–cell interactions mediated by β7 integrins and their ligands. Two β7 integrin–ligand interactions were found using scRNA-seq data, which were α4β7–VCAM-1 and αEβ7–CDH1. The α4β7–VCAM-1 interaction was mainly detected between endothelium cells, myofibroblasts, cancer-associated fibroblasts (CAF), and various immune cells, including IgA+ B cells, CD16+ NK cells, GZMK+ NK cells, CD103+ NK cells, CD4 subclusters, and CD8 subclusters, except in LEF1+CD8+ T cells and GPR183+CD8+ T cells (Fig. 2G). The αEβ7–CDH1 interaction was found between malignant cells, which had four subclusters (C1–C4) and GZMK+ NK cells, CD103+ NK cells, CXCL13+CD4+ T cells, CTLΑ4+CD4+ T cells, GZMK+CD8+ T cells, CD6+CD8+ T cells, CD160+CD8+ T cells, and LAYN+CD8+ T cells, which might be vital for immune cytotoxicity to tumors (Fig. 2G). Collectively, these data indicate that ITGB7 is essential for antitumor immunity, and it correlates with immune infiltration, especially the cytotoxic cells.

The patterns of SCNAs and mutations related to ITGB7 expression

SCNA was reported to be related to tumor immune evasion (24, 36) and response to immunotherapy. To explore the association between ITGB7 expression and SCNAs, we analyzed SCNAs for the TCGA-CRC cohort. SCNAs of both the broad and focal regions showed a weak negative correlation (the correlation coefficient was −0.3 and −0.35, respectively) with ITGB7 expression (Fig. 3A). Consistently, both the broad and focal SCNAs decreased in the ITGB7hi group (Fig. 3B). The ITGB7lo group displayed more frequent copy-number gains in oncogenes (HNF4A, KLF5, IRS2, CDK8, EGFR, MYC, and CCNA1) and copy-number losses in tumor suppressor genes (SMAD4, MAP2K4, and TP53; Fig. 3C and D).

We next examined somatic mutations related to the high and low groups. Among the top 15 altered genes (Fig. 3E), APC and TP53 showed a higher frequency of mutation in the ITGB7lo group (Fig. 3F). Because neoantigens play an important role in tumor immunity and survival after immunotherapy, we compared the neoantigen distribution in the ITGB7lo and ITGB7hi groups, predicted by NetMHCpan 2.8. The ITGB7hi group showed an increased number of neoantigens per sample (Fig. 3G), indicating a higher immune response.

ITGB7 expression associates with immunotherapeutic response

ITGB7 expression significantly correlated with expression of immunotherapy-related genes, including IDO1, HAVCR2, CTLΑ4, CD274, PDCD1, PDCD1LG2, TIGIT, and LAG3 (Supplementary Fig. S4). Therefore, we investigated the association between ITGB7 expression and immunotherapy responses and evaluated the prognostic value of ITGB7 expression for immune checkpoint therapy. Owing to the lack of gene expression data for colorectal cancer immunotherapy, we used transcriptomic datasets with immunotherapy in lung cancer (GSE136961, anti–PD-1), urothelial cancer (IMvigor210, anti–PD-L1), and a mouse study (GSE117358, anti-CTLΑ4/anti–PD-L1). Patients with high ITGB7 expression had significantly longer progression-free survival and longer overall survival than those with low ITGB7 expression in the GSE136961 (Fig. 4A) and IMvigor210 (Fig. 4E) cohorts, respectively. The proportion of immunotherapy responders in the ITGB7hi group was significantly higher than that in the ITGB7lo group (Fig. 4B, F, and G). The response group showed significantly higher expression of ITGB7 (Fig. 4C and H). ROC curve of GSE136961 showed a predictive value of 0.75 for ITGB7 expression (Fig. 4D). Thus, these data suggest that high ITGB7 expression associated with a good response to different immunotherapy approaches, including anti–PD-1/PD-L1/CTLA4 treatments.

ITGB7 deficiency exaggerates tumorigenesis and progression of colorectal tumors

To confirm the findings of the bioinformatics study and investigate the role of ITGB7 in the pathogenesis of colorectal cancer in vivo, we first implanted MC38 cells into the colonic subserosa of both WT and Itgb7−/− mice. The Itgb7−/− mice displayed shorter survival than WT mice (Fig. 5A) and developed two times greater tumor volume than the WT mice (Fig. 5B and C). The cells in the Itgb7−/− tumors seemed to be more diffuse than the WT tumors, indicating progressive tumor development in the Itgb7−/− mice (Fig. 5D). Next, we crossed the Itgb7−/− mice with the ApcMin/+ mice to generate Itgb7−/−ApcMin/+ mice and compared polyp development in the Apcmin/+and Itgb7−/−ApcMin/+ mice. The Itgb7−/−ApcMin/+ mice developed a higher number of polyps than the ApcMin/+ mice (Fig. 5E). In detail, compared with ApcMin/+ mice, the number of tumors with a diameter >1 mm increased significantly in Itgb7−/−ApcMin/+ mice (Fig. 5F), and the number of tumors increased along the entire gut in Itgb7−/−ApcMin/+ mice, with the distal part of the intestine showing the most significant difference (Fig. 5F). The Itgb7−/−ApcMin/+ tumors displayed a more progressive histology than WT tumors (Fig. 5G). Collectively, these results indicate an inhibitory role of ITGB7 in colorectal cancer tumorigenesis and progression.

Itgb7−/− mice display impaired antitumor immunity

As shown in Fig. 2, ITGB7 was expressed on and associated with the infiltration of several types of immune cells, including T cells, B cells, NK cells, and DCs. We first analyzed surface expression of α4β7 and αEβ7 in tumor-infiltrating immune cells in WT mice bearing orthotopic MC38 tumor. Similar with the expression pattern in human tumor-infiltrating immune cells (Fig. 1D and E), all the subsets of immune cells showed a surface expression of α4β7 except macrophages, while only T cells and a small subset of NK cells and DCs expressed αEβ7 (Supplementary Fig. S5A and S5B).

We then analyzed tumor-infiltrating activated B cells (CD80+ B cells), activated CD8+ T cells (CD69+CD8+ T cells), effector memory (CD44highCD62L) CD8+ T cells, and cytotoxic cells (IFNγ+CD8+ T cells and NK cells), which showed a positive correlation with ITGB7 expression (Fig. 2A) using an orthotopic colorectal cancer model. We also analyzed tumor-infiltrating IgA+ B cells and IFNγ+CD4+ T cells because they showed high expression of ITGB7 (Fig. 2E). Compared with MC38 orthotopic tumors of WT mice, tumors from Itgb7−/− mice showed a decreased infiltration of activated B cells, IgA+ B cells, activated CD8+ T cells, effector memory CD8+ T cells, and IFNγ+CD4+ T cells, CD8+ T cells, and NK cells (Fig. 6A and C). Given DCs express both α4β7 and αEβ7, we also examined tumor-infiltrating DC subsets and found depletion of CD103+ DCs in Itgb7−/− MC38 tumors (Fig. 6B and D), which has been shown to be indispensable for the cross-presentation of antigens to CD8+ T cells (37) and enhancing tumor responses to immunotherapy (38, 39). These data indicate that Itgb7 is indispensable for tumor-infiltrating immune cells required for antitumor immunity.

We next conducted a competitive homing assay to compare the migration of WT and Itgb7−/− T cells into tumors. Splenic CD3+ T cells isolated from WT CD45.1+ and Itgb7−/− CD45.2+ mice were expanded and equal numbers of WT and Itgb7−/− T cells were mixed and intravenously administered into MC38 tumor–bearing Rag1−/− recipient mice, which are devoid of T cells. Tissues were harvested 5 days after administration, and the homing indices were determined (Fig. 6E). As expected, WT and Itgb7−/− T cells homed equally well to spleen, whereas homing of Itgb7−/− T cells to mesenteric lymph nodes (MLN), small intestine lamina propria (SILP), and tumors was decreased (Fig. 6F and G). This indicated that β7 deficiency suppressed T-cell homing into tumors.

To further confirm that T cells required β7 to migrate into tumors, we conducted a transwell assay to compare the migration ability of CD3+ T cells in the presence or absence of β7 blocking antibody using CRC-organoids and in vitro expanded human CD3+ T cells. FIB504.64 treatment decreased the number of migrated T cells toward the CRC-organoids (Supplementary Fig. S6), which is consistent with the in vivo competitive homing assay. These data indicate that T cells require β7 to migrate into tumors.

ITGB7 expression reflects not only tumor immune infiltration

Because ITGB7 is expressed on subsets of immune cells, it is unclear whether the expression of ITGB7 or any other markers expressed on immune cells could reflect the immune cell infiltration into tumor tissue. We first analyzed the expression of integrin β2 (ITGB2), another immune cell–expressing integrins in tumor and normal tissues of patients with colorectal cancer. Similar with CD3E, the expression of ITGB7 decreased in tumors compared with paired normal tissues in TCGA cohort, whereas the expression of ITGB2 was comparable between tumor and normal tissues (Supplementary Fig. S7A). Consistently, the number of ITGB2+ cells in tumors showed no significant alteration with that of normal tissues (Supplementary Fig. S7B and S7C). We next investigated the correlation between ITGB7 expression and lymphocyte infiltration in tumor samples from TCGA cohort. ITGB7 expression positively correlated with tumor-infiltrating total T cells, CD8+ T cells, and B cells but had no significant relation with CD4+ T cells (Supplementary Fig. S7D). Infiltrating total T cells, CD8+ T cells, and B cells decreased gradually from normal to ITGB7hi to ITGB7lo groups (Supplementary Fig. S7E). These data indicate ITGB7 expression could reflect tumor infiltration by CD8+ T cells and B cells.

However, it is unclear whether the ITGB7 expression in tumor tissue simply reflects the level of CD8+ T- and B-cell infiltration. The expression of αEβ7 and α4β7 both decreased in tumor-derived CD8+ T cells compared with normal tissue ones (Fig. 1D and E), which led us to speculate that ITGB7 expression on specific immune cell subsets could also contribute to tissue ITGB7 expression. ITGB7 expression on CD8+ T cells decreased gradually from normal to ITGB7hi to ITGB7lo groups in TCGA cohort. We found that in some tumors with high T-cell infiltration, only a small subset of T cells expressed ITGB7 [Supplementary Fig. S7G, tumor (type 2) image], which might due to the downregulation of ITGB7 on CD8+ T cells and indicate that high T-cell infiltration does not indicate high ITGB7 expression in tumors. Thus, we speculate that ITGB7 expression in tumor tissues may be influenced by both the immune cell infiltration (CD8+ T and B cells in majority) and its expression on CD8+ T cells.

In this study, we applied a series of bioinformatics analyses, together with patient and animal studies, to reveal an inhibitory role of ITGB7 in the pathogenesis of colorectal cancer. Apart from cytotoxic immune cells and DCs, which have been demonstrated to play essential roles in antitumor immunity, IgA+ B cells were also associated with the risk of colorectal cancer. Ludvigsson and colleagues found that IgA deficiency associates with an increased risk of gastrointestinal cancer (40). IgA can have a reciprocal interaction with gut microbiota (41), the alteration of which can be involved in the tumorigenesis of colorectal cancer. Das and colleagues report a promoting role of ITGB7 in tumor growth in the proximal small intestine (42). In their study, the number of tumors decreased in duodenum and jejunum of ApcMin/+Itgb7−/− mice, but were increased in ileum compared to ApcMin/+ mice. Their result for the ileum is similar with our results for the colon. Because the distal ileum is a transition zone consisting of sparse populations of aerobic bacteria of the proximal small intestine and very dense populations of anaerobic bacteria of the large bowel (43, 44), we speculate that ITGB7 plays contrasting roles in colon, ileum, and proximal small intestine, which might be due to the different composition and abundance of microorganisms in different locations of gut.

Our study revealed that ITGB7 deficiency led to increased tumor initiation and development. However, ITGB7 blockade has been used as a therapeutic strategy in treating colitis, such as inflammatory bowel disease (IBD), which is a risk factor for colorectal cancer. Colitis-associated colorectal cancer (CAC) is reported to have a lower number of CD3+, CD8+, Foxp3+, and PD-L1+ cells compared with sporadic colorectal cancer (45), which may be due to the use of immunosuppressive agents in patients with IBD. Similar with sporadic colorectal cancer, higher infiltration of CD3+ and CD8+ T cells in CAC correlates with better survival. Thus, immunosuppressive agents used to treat colitis may have a harmful impact on the outcome of CAC, which might be due to immune cell infiltration having contrasting effects on the outcome of IBD and colorectal cancer.

With both in silico analysis and in vivo experiments, we found that ITGB7 expression associated with a series of immune cell subsets that play essential roles in tumor killing, such as effector memory CD8+ T cells, IFNγ+CD8+ T cells, CD103+ DCs, and IFNγ+ NK cells. ITGB7 might be used as a marker to rapidly identify antitumoral immune cell infiltration. Although high ITGB7 expression is associated with better immunotherapy response, whether it could be used as a biomarker to predict response to immunotherapy needs more careful study.

In conclusion, our study demonstrated that ITGB7 played an inhibitory role in the pathogenesis of colorectal cancer by maintaining antitumor immunity. Impaired antitumor immunity might be due to decreased migration of T cells into tumors. As has been shown in previous studies, the development of GALT, such as Peyer patches and the lamina propria, were impaired in Itgb7−/− mice (15, 46), which could also contribute to increased tumor initiation due to impaired immune surveillance. Thus, the increased tumor development in Itgb7−/− mice may be attributed to both decreased migration of Itgb7−/− T cells into tumors and impaired GALT development. We will investigate the contribution of the two aspects in tumor initiation and growth in Itgb7−/− mice in the future.

No disclosures were reported.

Y. Zhang: Funding acquisition, investigation, visualization, writing–original draft, project administration, writing–review and editing. R. Xie: Formal analysis, validation, investigation, visualization. H. Zhang: Investigation, visualization, methodology, writing–review and editing. Y. Zheng: Validation, investigation, methodology. C. Lin: Validation, investigation, visualization. L. Yang: Validation, investigation, methodology. M. Huang: Validation, investigation, visualization. M. Li: Investigation, methodology. F. Song: Validation, investigation. L. Lu: Formal analysis, validation. M. Yang: Validation, investigation. Y. Liu: Validation, visualization. Q. Wei: Resources, supervision, funding acquisition. J. Li: Resources, supervision, funding acquisition. J. Chen: Conceptualization, funding acquisition, writing–original draft, project administration, writing–review and editing.

The authors thank all the patients with colorectal cancer who participated in this study.

This work was supported by grants from the National Key Research and Development Program of China (2020YFA0509100, to J. Chen), National Natural Science Foundation of China (31525016, 31830112, and 32030024, to J. Chen; 82072647, to J. Li; 82072634, to Q. Wei; and 81702309, to Y. Zhang), Program of Shanghai Academic Research Leader (19XD1404200, to J. Chen), National Ten Thousand Talents Program (to J. Chen), Clinical Research Plan of Shanghai Hospital Development Center (SHDC; SHDC2020CR2069B, to Q. Wei), Shanghai Natural Science Foundation (21ZR1449900, to Y. Zhang), State Key Laboratory of Cell Biology (SKLCB; 2017KF001, to Y. Zhang), and Climbing Talents Plan of Shanghai Tenth People's Hospital (2021SYPDRC012, to Y. Zhang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Garborg
K
,
Holme
O
,
Loberg
M
,
Kalager
M
,
Adami
HO
,
Bretthauer
M
. 
Current status of screening for colorectal cancer
.
Ann Oncol
2013
;
24
:
1963
72
.
2.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2018
.
CA Cancer J Clin
2018
;
68
:
7
30
.
3.
Allemani
C
,
Matsuda
T
,
Di Carlo
V
,
Harewood
R
,
Matz
M
,
Niksic
M
, et al
Global surveillance of trends in cancer survival 2000–14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries
.
Lancet
2018
;
391
:
1023
75
.
4.
Pages
F
,
Berger
A
,
Camus
M
,
Sanchez-Cabo
F
,
Costes
A
,
Molidor
R
, et al
Effector memory T cells, early metastasis, and survival in colorectal cancer
.
N Engl J Med
2005
;
353
:
2654
66
.
5.
Fridman
WH
,
Pages
F
,
Sautes-Fridman
C
,
Galon
J
. 
The immune contexture in human tumours: impact on clinical outcome
.
Nat Rev Cancer
2012
;
12
:
298
306
.
6.
Galon
J
,
Angell
HK
,
Bedognetti
D
,
Marincola
FM
. 
The continuum of cancer immunosurveillance: prognostic, predictive, and mechanistic signatures
.
Immunity
2013
;
39
:
11
26
.
7.
Galon
J
,
Pages
F
,
Marincola
FM
,
Thurin
M
,
Trinchieri
G
,
Fox
BA
, et al
The immune score as a new possible approach for the classification of cancer
.
J Transl Med
2012
;
10
:
1
.
8.
Galon
J
,
Mlecnik
B
,
Bindea
G
,
Angell
HK
,
Berger
A
,
Lagorce
C
, et al
Towards the introduction of the ‘Immunoscore’ in the classification of malignant tumours
.
J Pathol
2014
;
232
:
199
209
.
9.
Angel
H
,
Galon
J
. 
From the immune contexture to the Immunoscore: the role of prognostic and predictive immune markers in cancer
.
Curr Opin Immunol
2013
;
25
:
261
7
.
10.
Fakih
M
,
Ouyang
C
,
Wang
C
,
Tu
TY
,
Gozo
MC
,
Cho
M
, et al
Immune overdrive signature in colorectal tumor subset predicts poor clinical outcome
.
J Clin Invest
2019
;
129
:
4464
76
.
11.
Harburger
DS
,
Calderwood
DA
. 
Integrin signalling at a glance
.
J Cell Sci
2009
;
122
:
159
63
.
12.
Hynes
RO
. 
Integrins: bidirectional, allosteric signaling machines
.
Cell
2002
;
110
:
673
87
.
13.
Berlin
C
,
Berg
EL
,
Briskin
MJ
,
Andrew
DP
,
Kilshaw
PJ
,
Holzmann
B
, et al
Alpha 4 beta 7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1
.
Cell
1993
;
74
:
185
95
.
14.
Cepek
KL
,
Shaw
SK
,
Parker
CM
,
Russell
GJ
,
Morrow
JS
,
Rimm
DL
, et al
Adhesion between epithelial cells and T lymphocytes mediated by E-cadherin and the αEβ7 integrin
.
Nature
1994
;
372
:
190
3
.
15.
Zhang
HL
,
Zheng
YJ
,
Pan
YD
,
Lin
CD
,
Wang
SH
,
Yan
ZJ
, et al
A mutation that blocks integrin α4β7 activation prevents adaptive immune-mediated colitis without increasing susceptibility to innate colitis
.
BMC Biol
2020
;
18
:
64
.
16.
Zhang
HL
,
Zheng
YJ
,
Pan
YD
,
Xie
C
,
Sun
H
,
Zhang
YH
, et al
Regulatory T-cell depletion in the gut caused by integrin beta (7) deficiency exacerbates DSS colitis by evoking aberrant innate immunity
.
Mucosal Immunol
2016
;
9
:
391
400
.
17.
Colaprico
A
,
Silva
TC
,
Olsen
C
,
Garofano
L
,
Cava
C
,
Garolini
D
, et al
TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data
.
Nucleic Acids Res
2016
;
44
:
e71
.
18.
Schrock
AB
,
Ouyang
C
,
Sandhu
J
,
Sokol
E
,
Jin
D
,
Ross
JS
, et al
Tumor mutational burden is predictive of response to immune checkpoint inhibitors in MSI-high metastatic colorectal cancer
.
Ann Oncol
2019
;
30
:
1096
103
.
19.
Charoentong
P
,
Finotello
F
,
Angelova
M
,
Mayer
C
,
Efremova
M
,
Rieder
D
, et al
Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint blockade
.
Cell Rep
2017
;
18
:
248
62
.
20.
Thorsson
V
,
Gibbs
DL
,
Brown
SD
,
Wolf
D
,
Bortone
DS
,
Ou Yang
TH
, et al
The immune landscape of cancer
.
Immunity
2018
;
48
:
812
30
.
21.
Yu
G
,
Wang
LG
,
Han
Y
,
He
QY
. 
clusterProfiler: an R package for comparing biological themes among gene clusters
.
OMICS
2012
;
16
:
284
7
.
22.
Jin
S
,
Guerrero-Juarez
CF
,
Zhang
L
,
Chang
I
,
Ramos
R
,
Kuan
CH
, et al
Inference and analysis of cell-cell communication using CellChat
.
Nat Commun
2021
;
12
:
1088
.
23.
Mermel
CH
,
Schumacher
SE
,
Hill
B
,
Meyerson
ML
,
Beroukhim
R
,
Getz
G
. 
GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers
.
Genome Biol
2011
;
12
:
R41
.
24.
Davoli
T
,
Uno
H
,
Wooten
EC
,
Elledge
SJ
. 
Tumor aneuploidy correlates with markers of immune evasion and with reduced response to immunotherapy
.
Science
2017
;
355
:
eaaf8399
.
25.
Wu
J
,
Zhao
W
,
Zhou
B
,
Su
Z
,
Gu
X
,
Zhou
Z
, et al
TSNAdb: a database for tumor-specific neoantigens from immunogenomics data analysis
.
Genomics Proteomics Bioinformatics
2018
;
16
:
276
82
.
26.
Newman
AM
,
Steen
CB
,
Liu
CL
,
Gentles
AJ
,
Chaudhuri
AA
,
Scherer
F
, et al
Determining cell type abundance and expression from bulk tissues with digital cytometry
.
Nat Biotechnol
2019
;
37
:
773
82
.
27.
Roelands
J
,
Kuppen
PJK
,
Vermeulen
L
,
Maccalli
C
,
Decock
J
,
Wang
E
, et al
Immunogenomic classification of colorectal cancer and therapeutic implications
.
Int J Mol Sci
2017
;
18
:
2229
.
28.
Jorissen
RN
,
Gibbs
P
,
Christie
M
,
Prakash
S
,
Lipton
L
,
Desai
J
, et al
Metastasis-associated gene expression changes predict poor outcomes in patients with Dukes stage B and C colorectal cancer
.
Clin Cancer Res
2009
;
15
:
7642
51
.
29.
Smith
JJ
,
Deane
NG
,
Wu
F
,
Merchant
NB
,
Zhang
B
,
Jiang
A
, et al
Experimentally derived metastasis gene expression profile predicts recurrence and death in patients with colon cancer
.
Gastroenterology
2010
;
138
:
958
68
.
30.
Iwaya
T
,
Yokobori
T
,
Nishida
N
,
Kogo
R
,
Sudo
T
,
Tanaka
F
, et al
Downregulation of miR-144 is associated with colorectal cancer progression via activation of mTOR signaling pathway
.
Carcinogenesis
2012
;
33
:
2391
7
.
31.
Kogo
R
,
Shimamura
T
,
Mimori
K
,
Kawahara
K
,
Imoto
S
,
Sudo
T
, et al
Long noncoding RNA HOTAIR regulates polycomb-dependent chromatin modification and is associated with poor prognosis in colorectal cancers
.
Cancer Res
2011
;
71
:
6320
6
.
32.
Laibe
S
,
Lagarde
A
,
Ferrari
A
,
Monges
G
,
Birnbaum
D
,
Olschwang
S
. 
A seven-gene signature aggregates a subgroup of stage II colon cancers with stage III
.
OMICS
2012
;
16
:
560
5
.
33.
Marisa
L
,
de Reyniès
A
,
Duval
A
,
Selves
J
,
Gaub
MP
,
Vescovo
L
, et al
Gene expression classification of colon cancer into molecular subtypes: characterization, validation, and prognostic value
.
PLoS Med
2013
;
10
:
e1001453
.
34.
Zaravinos
A
,
Roufas
C
,
Nagara
M
,
de
L
,
Moreno
B
,
Oblovatskaya
M
, et al
Cytolytic activity correlates with the mutational burden and deregulated expression of immune checkpoints in colorectal cancer
.
J Exp Clin Cancer Res
2019
;
38
:
364
.
35.
Zhang
L
,
Li
Z
,
Skrzypczynska
KM
,
Fang
Q
,
Zhang
W
,
O'Brien
SA
, et al
Single-cell analyses inform mechanisms of myeloid-targeted therapies in colon cancer
.
Cell
2020
;
181
:
442
59
.
36.
Bhattacharya
A
,
Bense
RD
,
Urzua-Traslavina
CG
,
de Vries
EGE
,
van Vugt
M
,
Fehrmann
RSN
. 
Transcriptional effects of copy number alterations in a large set of human cancers
.
Nat Commun
2020
;
11
:
715
.
37.
Cerovic
V
,
Houston
SA
,
Westlund
J
,
Utriainen
L
,
Davison
ES
,
Scott
CL
, et al
Lymph-borne CD8alpha+ dendritic cells are uniquely able to cross-prime CD8+ T cells with antigen acquired from intestinal epithelial cells
.
Mucosal Immunol
2015
;
8
:
38
48
.
38.
Salmon
H
,
Idoyaga
J
,
Rahman
A
,
Leboeuf
M
,
Remark
R
,
Jordan
S
, et al
Expansion and activation of CD103(+) dendritic cell progenitors at the tumor site enhances tumor responses to therapeutic PD-L1 and BRAF inhibition
.
Immunity
2016
;
44
:
924
38
.
39.
Williford
JM
,
Ishihara
J
,
Ishihara
A
,
Mansurov
A
,
Hosseinchi
P
,
Marchell
TM
, et al
Recruitment of CD103(+) dendritic cells via tumor-targeted chemokine delivery enhances efficacy of checkpoint inhibitor immunotherapy
.
Sci Adv
2019
;
5
:
eaay1357
.
40.
Ludvigsson
JF
,
Neovius
M
,
Ye
W
,
Hammarstrom
L
. 
IgA deficiency and risk of cancer: a population-based matched cohort study
.
J Clin Immunol
2015
;
35
:
182
8
.
41.
Pabst
O
,
Cerovic
V
,
Hornef
M
. 
Secretory IgA in the coordination of establishment and maintenance of the microbiota
.
Trends Immunol
2016
;
37
:
287
96
.
42.
Das
S
,
Donas
C
,
Akeus
P
,
Quiding-Jarbrink
M
,
Mora
JR
,
Villablanca
EJ
. 
Beta7 integrins contribute to intestinal tumor growth in mice
.
PLoS One
2018
;
13
:
e0204181
.
43.
Camp
JG
,
Kanther
M
,
Semova
I
,
Rawls
JF
. 
Patterns and scales in gastrointestinal microbial ecology
.
Gastroenterology
2009
;
136
:
1989
2002
.
44.
Mackie
RI
,
Sghir
A
,
Gaskins
HR
. 
Developmental microbial ecology of the neonatal gastrointestinal tract
.
Am J Clin Nutr
1999
;
69
:
1035S
45S
.
45.
Soh
JS
,
Jo
SI
,
Lee
H
,
Do
EJ
,
Hwang
SW
,
Park
SH
, et al
Immunoprofiling of Colitis-associated and sporadic colorectal cancer and its clinical significance
.
Sci Rep
2019
;
9
:
6833
.
46.
Wagner
N
,
Löhler
J
,
Kunkel
EJ
,
Ley
K
,
Leung
E
,
Krissansen
G
, et al
Critical role for beta7 integrins in formation of the gut-associated lymphoid tissue
.
Nature
1996
;
382
:
366
70
.