Patients
The study of archived human CRC was approved by the Institutional Research Board (IRB) of the University Hospitals Case Medical Center, Cleveland Clinic Foundation, and Fudan University Shanghai Cancer Center. Informed consent was obtained from patients who agreed to donate tissues for the purpose of research according to the regulation by the IRB. All cases included in the study were confirmed as colorectal adenocarcinoma by two experienced pathologists. All methods used in this study were carried out in accordance with IRB guidelines and regulations. Demographic and clinicopathological data were collected from the medical records. Mutational status was determined by the ColonCore next-generation sequencing (NGS) panel (Burning Rock Biotech, Guangzhou, China) which is designed for simultaneous detection of microsatellite instability (MSI) status and mutations in 38 CRC related genes. Cases included in this study were microsatellite stable (MSS) CRCs that also had KRAS and APC mutation (Table 1). Another cohort of MSS and KRAS wild type CRCs were included. These cases were part of an IRB-approved annotated biobank. Biobank tumors had been previously evaluated for microsatellite instability and KRAS mutation status as previously described (15). Cases that carry other frequently mutated genes (TP53, BRAF, NRAS) were excluded from the study.
Gene expression analysis
Ten cases, 5 HES1 (-) and 5 HES1 (+), were subjected to RNA sequencing (cohort 1). RNA of these cases was isolated from the fresh frozen tissue followed by mRNA library preparation using Illumina’s TruSeq RNA Sample Prep Kit v2 (Illumina, RS-122-2001, San Diego, CA, USA). Sequencing was performed using Illumina HiSeq 2500 System (Illumina, San Diego, CA, USA). Another cohort of 9 cases, 4 HES1 (-) and 5 HES1 (+) were subjected to Nanostring RNA gene expression array analysis (cohort 2). RNA was isolated from formalin-fixed, paraffin-embedded (FFPE) tissue and assessed by the nCounter® PanCancer IO 360TM Panel (NanoString technologies, Seattle, WA, USA) (16) .
TCGA data acquisition
TCGA data of colorectal cancer (n=431) level 3 gene-expression (counts) and somatic mutation were obtained from Genomic Data Commons (GDC) (https://portal.gdc.cancer.gov/). KRAS mutation was identified based on the “maftool” R package. The original counts data were transformed into transcript per kilobase million (TPM). Patients who lacked follow-up and somatic mutation information were excluded. A total of 155 colorectal cancer patients with KRAS mutation and 222 colorectal cancer patients with wild type KRAS were enrolled in this study. The high and low expression of HES1 in TCGA data was determined by the median expression (10.799) as a cutoff. Survival analysis was performed with Kaplan-Meier analysis in all colorectal cancer patients.
Differentially expressed gene (DEG) and GSEA analysis
To identify genes associated with HES1 expression, DEGs was determined by using limma R package. The significant criteria were selected using P value < 0.05 and absolute fold-change (FC) >1. The Venn Diagram was generated by the package of “venn”. Gene Set Enrichment Analysis (GSEA) analysis was performed by “ClusterProfiler” package.
Immune cell infiltration analysis
Immune cells signature was determined by previously published method (17). Briefly, Gene Set Variation Analysis (GSVA) was used to calculate the scale of value of each immune cells.
Immunohistochemical staining and evaluation
Paraffin blocks of 25 cases from cohort 1 & 2 were selected for the construction of the tissue microarray (TMA). For each block, three cores with a diameter of 2 mm were obtained from the tumor. Immunohistochemical staining (IHC) was performed using the automated immunostainer (Ventana, Tucson, AZ, USA). Primary antibodies used in this study include HES1 (Clone: EPR4226, Cat. No. ab108937, Abcam), Ki67 (Clone: 30-9, Cat. No. 790-4286, Ventana), TP53 (Clone: DO-7, Cat. No. M7001, Dako), RB1 (Clone: 4H1, Cat. No. 9309, Cell Signaling Technology), Cyclin D1 (Clone: SP4-R, Cat. No. 790-4508, Ventana), E-cadherin (Clone: NCH-38, Cat. No. M3612, Dako), Vimentin (Clone: V9, Cat. No. IR630, Dako), CD44 (Clone: DF1485, Cat. No. M7082, Dako), CD8 (Clone: SP57, Cat. No. 790-4460, Ventana), CD163 (Clone: MX081, Cat. No. MAB-0869, Fuzhou Maixin Biotechnology), CD68 (Clone: KP1, Cat. No. M-0160-1.0, Shanghai Changdao Biotechnology), phospho-STAT3 (Tyr705, Clone: D3A7, Cat. No. 9145, Cell Signaling Technology), IL10 (Clone: 2472A, Cat. No. MAB91842, R&D Systems). Expression of HES1 was evaluated as previously described (12). The presence of HES1 nuclear expression was considered HES1 (+), while loss of HES1 nuclear expression was classified as HES1 (-). Histoscores (H-scores) were calculated by multiplying the staining intensity (0=negative, 1=weak, 2=moderate, 3=strong) and the percentage of positive cells (number of positive tumor cells/ number of total tumor cells, range 0-100). All cases were scored by two experienced pathologists. The expression status of TP53, RB1 and Cyclin D1, IL10 were evaluated using H-scores. Tumor cells showed homogeneously strong membrane expression of E-cadherin were considered positive, while weak or loss expression of E-cadherin of tumor cells was classified as abnormal. The percentage of positive Ki67 staining in tumor cells was evaluated. Densities of CD8, CD163 and CD68 were calculated (area of positive immune cells/total area of tissue).
Statistical Analysis
Data were analyzed using R software (version4.2.1). Comparisons of ≥ 2 groups were conducted using a parametric test (Student t-test or ANOVA test) or a nonparametric test (Wilcoxon rank-sum test or Kruskal-Wallis test, Pearson Chi-Square test or Fisher’s Exact Test). ns, *, **, and *** represent not significant (p ≥ 0.05), p < 0.05, p ≤ 0.01, and p ≤ 0.001, respectively.