Introduction

Oncology Letters

1792-1074 1792-1082

D.A. Spandidos

10.3892/ol.2023.13663

OL-25-2-13663

Articles

Collagen type VII α1 chain: A promising prognostic and immune infiltration biomarker of pancreatic cancer

Ding

Cheng

1 * Yu

Zhangping

2 * Li

Xianliang

1 Zhu

Jiqiao

1 Dai

Menghua

2 He

Qiang

1Department of Hepatobiliary Surgery, Beijing Chao-Yang Hospital Affiliated to Capital Medical University, Beijing 100020, P.R. China 2Department of General Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing 100730, P.R. China

Correspondence to: Dr Qiang He, Department of Hepatobiliary Surgery, Beijing Chao-Yang Hospital Affiliated to Capital Medical University, 8 Gongti South Road, Chaoyang, Beijing 100020, P.R. China, E-mail: heqiang349@sina.com Dr Menghua Dai, Department of General Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, 1 Shuaifu Garden, Wangfujing Street, Beijing 100730, P.R. China, E-mail: daimh@pumch.cn *

Contributed equally

02 2023

05 01 2023

25 2

12032022 25112022

2023

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Pancreatic cancer (PC) is a stubborn malignancy with high lethality and a low 5-year overall survival (OS) rate. Collagen type VII α1 chain (COL7A1), a major component of the extracellular matrix, serves important roles in numerous physiological processes and various illnesses. COL7A1 protein acts as an anchoring fibril between the external epithelial cells and the underlying stroma, and mutation of COL7A1 could cause recessive dystrophic epidermolysis bullosa. Raw data for PC were acquired from The Cancer Genome Atlas and the Gene Expression Omnibus database, and raw data for the normal pancreas were obtained from the Genotype-Tissue Expression database. COL7A1 mRNA expression in PC tissues was compared with that in either paired (GSE15471 dataset) or unpaired (all other data) normal pancreas tissues. The association between COL7A1 mRNA expression and clinicopathological factors was assessed using logistic regression analysis. Cox analysis and Kaplan-Meier analysis were used to evaluate the role of COL7A1 mRNA expression in prognosis and nomograms were constructed. Gene Ontology analysis, Kyoto Encyclopedia of Genes and Genomes analysis, Gene Set Enrichment Analysis (GSEA) and single-sample GSEA (ssGSEA) were performed to evaluate the relevant functions of COL7A1 and correlation with immune cell infiltration. Furthermore, reverse transcription-quantitative PCR was used to assess the mRNA expression levels of COL7A1 in PC. The present study demonstrated that COL7A1 mRNA expression was higher in PC tissues compared with in normal pancreas tissues. The Kaplan-Meier survival analysis indicated that patients with PC with high COL7A1 mRNA expression had shorter overall survival (OS), disease-specific survival (DSS) and progression-free interval (PFI) times compared with patients with PC with low COL7A1 mRNA expression. Multivariate analysis demonstrated that COL7A1 mRNA expression was an independent risk factor for OS, DSS and PFI. Nomogram and calibration plots were constructed to predict the prognosis of patients with PC. GSEA demonstrated that high mRNA expression levels of COL7A1 were associated with multiple cancer-related pathways. ssGSEA analysis indicated that COL7A1 expression was positively associated with natural killer CD56bright cells and T helper (Th)2 cells, and negatively associated with Th17 cells and eosinophils. The results of the present study suggested that COL7A1 could be an independent biomarker and an influential moderator of immune infiltration in PC.

collagen type VII α1 chain pancreatic cancer immune infiltration biomarker prognosis

Cheng Ding

The costs of the present study were supported by Cheng Ding.

Introduction

Pancreatic cancer (PC), which has a 5-year overall survival (OS) rate of only ~10% (1), remains one of the most malignant cancer types worldwide. PC is notorious for its high morbidity and mortality rates (2). It has been reported that close to 460,000 individuals were diagnosed with PC and that there were >430,000 PC-related deaths worldwide in 2018 (2). Unlike other malignancies, the incidence of PC has been slowly growing for decades, which brings a tremendous economic burden (3).

The extracellular matrix (ECM) is an important non-cellular component of the tumor microenvironment. The ECM not only provides a physical scaffold for cells but also a depot for cytokines that promote tumor development (4). The crosstalk between the ECM and tumor-infiltrating immune cells (TIICs) serves a crucial role in the progression of tumors (5). Through remodeling of the ECM, cancer-associated fibroblasts can facilitate the occurrence and development of cancer (6). Mast cells can serve as tumor contributors to simulate angiogenesis and degradation of the ECM (7,8). Furthermore, the ECM and immune-associated mechanisms are involved in the development and progression of PC (9,10). Due to the low radical resection rate, chemotherapy and immune-therapy resistance, little progress has been made in the management of PC during the past decades (11). Therefore, investigation of effective and reliable biomarkers for the diagnosis and treatment of PC is urgently required to improve survival rates.

Type VII collagen, which is encoded by collagen type VII α1 chain (COL7A1), is distributed to the basal area beneath the squamous epithelium (12,13). It is composed of three α collagen chains and acts as an anchoring fiber between the external epithelium and underlying substrate (12). Mutations of COL7A1 can result in recessive dystrophic epidermolysis bullosa (RDEB), which is an incurable autoimmune disease and associated with increased risk of skin carcinoma (13). In squamous cell carcinoma, the loss of type VII collagen can enhance tumor cell invasive behavior and promotes epithelial-mesenchymal transition (14). Aberrant COL7A1 expression has been reported in esophageal cancer and COL7A1 expression has been reported to be positively associated with depth of invasion and lymph node metastasis, and negatively associated with survival (15,16). In gastric carcinoma, COL7A1 expression has been reported to be upregulated in cancer tissues compared with normal tissues (17). High intracellular COL7A1 expression has been reported to indicate a poor 5-year OS and a high immunohistochemistry score is associated with distant metastasis (17). However, little is known regarding the prognostic role of COL7A1 and its association with TME in PC.

To better evaluate the value of COL7A1 in the assessment of the progression of PC, RNA sequencing (RNA-seq) data from The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO) and Genotype-Tissue Expression (GTEx) databases were downloaded. Cox and logistic regression analyses, Kaplan-Meier survival analysis and nomograms were used to evaluate the prognostic role of COL7A1. Gene Ontology (GO) analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis and Gene Set Enrichment Analysis (GSEA) were used to evaluate the underlying mechanisms of COL7A1. Single-sample GSEA (ssGSEA) was used to assess the relationship between COL7A1 expression and TIICs. Furthermore, reverse transcription-quantitative PCR (RT-qPCR) was used to validate COL7A1 mRNA expression in PC cell lines.

Materials and methods <sec> <title>RNA-seq data acquisition and analysis

All transcriptome RNA-seq data with clinical information of patients with PC were accessed from TCGA [https://portal.gdc.cancer.gov; pancreatic adenocarcinoma (PAAD) dataset] and normal pancreas expression data were retrieved from the GTEx database (https://gtexportal.org), the aforementioned data was downloaded using the University of California, Santa Cruz Xena browser (https://xenabrowser.net/datapages/). RNA expression data were retrieved from GEO (https://www.ncbi.nlm.nih.gov/geo/) (GSE15471 and GSE101448 datasets).

COL7A1 mRNA expression in PC samples and normal tissues

Boxplots and scatter plots were used to compare the mRNA expression levels of COL7A1 in tumor and normal samples, the ggplot2 package (v3.3.3) (18) was used to generate the visualization. The expression levels of COL7A1 in TCGA and GTEx databases were compared using a Wilcoxon rank sum test, COL7A1 expression was compared between normal and pancreatic cancer tissues in GSE15471 using a Wilcoxon signed rank test, and COL7A1 expression was compared between pancreatic cancer and normal samples in GSE101448 using an unpaired t-test. In the present study, the samples in GSE15471 dataset were paired tissues, while in other instances, the normal tissues referred to unpaired tissues. The diagnostic value of COL7A1 in patients with PC was estimated using receiver operating characteristic (ROC) curves, the pROC (v1.17.0.1) (19) and ggplot2 (v3.3.3) (18) packages were used to analyze and visualize these data. The patients were divided into two groups, COL7A1-high and COL7A1-low, based on the median expression level.

Identification of differentially expressed genes (DEGs)

The DESeq2 package (v1.26.0) (20) was used to identify DEGs between COL7A1-high and COL7A1-low groups from TCGA. The criteria used were |log (fold change)|>1 and adjusted P<0.05. The results were presented as volcano plots and heat maps using the ggplot2 package (v3.3.3) (18).

Enrichment analysis and immune cell infiltration

To evaluate the biological effects of DEGs, GO (geneontology.org) and KEGG (www.kegg.jp) enrichment analyses were performed. The criteria for both GO and KEGG analysis were as follows: A minimum count of 5, a maximum count of 5,000, P<0.05 and false discovery rate (FDR)<0.25 were considered statistically significant. Furthermore, GSEA was performed to assess potential biological functions and pathways in the COL7A1-high and COL7A1-low groups. Gene sets with absolute normalized enrichment score >1, nominal P<0.05 and FDR<0.25 were considered to be statistically significant. The c2.cp.kegg.v7.4.symbols.gmt and h.all.v7.4.symbols.gmt were downloaded from the Molecular Signature Database (MSigDB) (www.gsea-msigdb.org). All of the enrichment analyses were performed using the clusterProfiler package (v3.14.3) (21). Furthermore, ssGSEA was used to perform immune infiltration analysis using the GSVA package (v1.34.0) (22) and based on gene expression profiles, the infiltration levels of 24 immune cell types (23) were quantified. To further evaluate the association between COL7A1 mRNA expression and immune cell infiltration levels, the data were assessed using Spearman correlation and Wilcoxon rank sum tests.

Association of COL7A1 with prognosis and model construction

OS, disease-specific survival (DSS) and progression-free interval (PFI) were used to evaluate the relationship between COL7A1 expression and prognosis in PC. The Kaplan-Meier Plotter (kmplot.com) was used to evaluate survival according to mRNA expression levels using the log-rank test and the parameters selected were as follows: mRNA(RNA-seq) for pan-cancer, split patient by median (24). To analyze the prognostic value of COL7A1 mRNA expression in PC, univariate and multivariate Cox analysis of TCGA-PAAD dataset was performed. The median COL7A1 mRNA expression level was defined as the cut-off value. Multivariate Cox analysis was then applied to validate the independent prognostic value of COL7A1 mRNA expression levels. Nomograms were constructed to evaluate the prognosis for 1-, 2- and 3-year OS for patients with PC a using the rms (v6.2-0) (https://cran.r-project.org/web/packages/rms) and survival (v3.2-10) (https://cran.r-project.org/web/packages/survival) packages for analysis and visualization respectively.

Cell culture

The MIA PaCa-2 (CRM-CRL-1420), BxPC-3 (CRL-1687), Capan-1 (HTB-79) and PATU-8988 (ACC 162) PC cell lines, and the hTERT-HPNE (CRL-4023) human normal pancreatic duct cell line were purchased from American Type Culture Collection. PANC-1, MIA PaCa-2, PATU-8988 and HPNE cells were cultured in Dulbecco's modified Eagle's medium (HyClone; Cytiva) containing 10% FBS (HyClone; Cytiva), BxPC-3 cells were cultured in RPMI-1640 medium (HyClone; Cytiva) containing 10% FBS (HyClone; Cytiva) and Capan-1 cells were cultured in Iscove's Modified Dulbecco's Medium (HyClone; Cytiva) containing 20% FBS (HyClone; Cytiva). All cell lines were cultured at 37°C in a 5% carbon dioxide cell culture incubator.

RT-qPCR

Total RNA was extracted from cells using TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.). GoScript™ Reverse Transcription Mix (cat. no. A25742; Promega Corporation) and PowerUP™ SYBR™ Green Master Mix (Thermo Fisher Scientific, Inc.) were used for RT-qPCR assays according to the manufacturers' protocols. The thermocycling conditions for qPCR were as follows: Initial denaturation and polymerase activation at 95°C for 2 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. The 2^−∆∆Cq (25) method was used to calculate relative mRNA expression levels and GAPDH was used as the reference. The primer sequences used in the present study were as follows: COL7A1 forward, 5′-GTTGGAGAGAAAGGTGACGAGG-3′ and reverse, 3′-TGGTCTCCCTTTTCACCCACAG-5′; and GAPDH forward, 5′-GTCTCCTCTGACTTCAACAGCG-3′ and reverse, 3′-ACCACCCTGTTGCTGTAGCCAA-5′.

Statistical analysis

RStudio software(version 1.4.171; http://www.rstudio.com) and R software (version 3.6.3; http://www.r-project.org)were used to perform statistical analyses. Two-tailed, unpaired Student's t-test was used to compare two groups. The RT-qPCR data were analyzed using one-way ANOVA and Dunnett's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

Results <sec> <title>COL7A1 mRNA expression in pan-cancerous and PC samples

Using data from TCGA and GTEx databases the mRNA expression levels of COL7A1 in human cancer types and normal tissues were assessed. Compared with those in normal samples, COL7A1 mRNA expression levels were significantly higher in bladder urothelial carcinoma, cervical squamous cell carcinoma and endocervical adenocarcinoma, cholangiocarcinoma, colon adenocarcinoma, lymphoid neoplasm diffuse large B-cell lymphoma, esophageal carcinoma, head and neck squamous cell carcinoma, liver hepatocellular carcinoma, PAAD, pheochromocytoma and paraganglioma, stomach adenocarcinoma and thymoma; however, COL7A1 was significantly downregulated in adrenocortical carcinoma, breast invasive carcinoma, glioblastoma multiforme, kidney chromophobe, kidney renal clear cell carcinoma and kidney renal papillary cell carcinoma (P<0.05; Fig. 1A). Additionally, in PC, COL7A1 mRNA expression levels were significantly higher compared with those in normal pancreas tissues (Fig. 1B). Furthermore, the mRNA expression levels of COL7A1 in PC and normal tissues were assessed in the GSE15471 and GSE101448 datasets. In the GSE15471 dataset, COL7A1 mRNA expression was significantly higher (P<0.001) in PC tissues (6.84±0.71) than in normal tissues (6.20±0.24). Similarly, in the GSE101448 dataset, COL7A1 mRNA expression was significantly higher (P<0.001) in PC tissues (10.08±1.29) than in normal tissues (8.32±0.93) (Fig. 1C and D). Furthermore, a ROC curve was used to evaluate the diagnostic value of COL7A1 in PC. The area under the curve of COL7A1 was 0.963 (95% CI, 0.942-0.983; Fig. 1E), which suggested that COL7A1 could serve as an effective marker for the diagnosis of PC. To identify DEGs in PC, 89 COL7A1-high samples were compared with 89 COL7A1-low PC samples. A total of 1,288 DEGs were identified, which included 1,007 downregulated genes and 281 upregulated genes (Fig. 1F).

GO and KEGG enrichment analyses and GSEA

To evaluate the relative biofunctions and pathways associated with COL7A1 in PC, the clusterProfiler package was used to perform GO and KEGG analysis. The results demonstrated that COL7A1-associated genes participated in multiple biological processes, cellular components and molecular functions, including ‘Regulation of biological quality’, ‘Cell-cell signaling’, ‘Plasma membrane part’, ‘Neuron part’, ‘Transporter activity’ and ‘Transmembrane transporter activity’ (Fig. 2A-C). Furthermore, KEGG enrichment analysis demonstrated that COL7A1-associated DEGs were involved in ‘Neuroactive ligand-receptor interaction’, ‘cAMP signaling pathway’, ‘cGMP-PKG signaling pathway’, ‘ECM-receptor interaction’ and numerous other signaling pathways (Fig. 2D). To better evaluate COL7A1-related signaling pathways, GSEA was performed. The GSEA results showed that COL7A1-associated genes were mainly involved in ‘ECM-receptor interaction’, ‘cell cycle’, ‘G2M checkpoint’ and ‘epithelial-mesenchymal-transition (EMT)’ pathways (Fig. 3A and B).

Correlation between COL7A1 expression and immune cell infiltration

ssGSEA and spearman correlation analysis were used to evaluate the correlation between COL7A1 mRNA expression levels and immune cell infiltration levels. As shown in Fig. 4, high COL7A1 mRNA expression in PC was significantly positively associated with the abundance of T helper (Th)2 cells (R=0.32; P<0.001) (Fig. 4A, B and F), natural killer (NK) CD56bright cells (R=0.25; P<0.001) (Fig. 4A, C and G), NK cells (R=0.19; P<0.05), Th1 cells (R=0.18; P<0.05) and regulatory T cells (R=0.15; P<0.05), and negatively related to infiltration levels of Th17 cells (R=−0.24; P<0.01)(Fig. 4A, E and I), eosinophils (R=−0.22; P<0.01) (Fig. 4A, D and H), T follicular helper cells (R=−0.18; P<0.05) and plasmacytoid dendritic cells (R=−0.15; P<0.05).

Association of COL7A1 mRNA expression with clinicopathological variables in PC

Based on the median mRNA expression levels of COL7A1, the patients were divided into COL7A1-high and COL7A1-low groups, and the clinical features and the expression levels of COL7A1 were presented in Table I. Logistic regression analysis was used to evaluate the correlation between COL7A1 expression and clinical features, the results indicated that COL7A1 mRNA expression levels were not significantly associated with any of the 10 clinical characteristics (gender, age, T stage, N stage, M stage, pathologic stage, histologic stage, history of diabetes, history of chronic pancreatitis and primary therapy outcome) analyzed (P>0.05; Table II).

High COL7A1 expression is associated with poor survival in PC

The Kaplan-Meier survival analysis demonstrated that high mRNA expression levels of COL7A1 were significantly associated with poor OS, DSS and PFI (Fig. 5A-C), in line with the results of the Kaplan-Meier Plotter analysis (Fig. 5D). Univariate analyses demonstrated that COL7A1 mRNA expression was a prognostic factor for OS, DSS and PFI. Furthermore, multivariate analysis demonstrated that COL7A1 mRNA expression was an independent prognostic indicator for PC (Tables III, SI and SII). Furthermore, the nomogram utilized T stage, N stage, age, histologic grade and COL7A1 mRNA expression to predict the 1-, 2- and 3-year OS, DSS and PFI in PC (Figs. 6A, S1A and S2A). The calibration curve was constructed to evaluate the efficiency of the nomogram (Fig. 6B, S1B and S2B). The 1, 2 and 3-year OS, DSS and PFI lines were close to ideal line, which indicated that this nomogram model demonstrated a high level of accuracy. Furthermore, ROC analysis was performed to predict 1, 2 and 3-year OS, DSS and PFI (Figs. 6C, S1C and S2C).

Upregulation of COL7A1 expression in PC cell lines assessed using RT-qPCR

To evaluate the accuracy of the bioinformatics analysis results, RT-qPCR was performed in PC cell lines (MIA PaCa-2, BxPC-3, Capan-1 and PATU-8988) and hTERT-HPNE cells. The results demonstrated that the mRNA expression levels of COL7A1 in the BxPC-3, Capan-1 and PATU-8988 cell lines were significantly higher than that in the hTERT-HPNE cell line. However, there was no significant difference in the COL7A1 mRNA expression levels between the MIA PaCa-2 and hTERT-HPNE cell lines. The RT-qPCR results indicated that COL7A1 was commonly overexpressed in PC cell lines, which was in accordance with the results bioinformatics analysis of public datasets (Fig. 7).

Discussion

According to a previous study, the ECM serves crucial roles in the progression of cancer (26) and the basement membrane is one type of ECM structure. Type VII collagen, a type of basement membrane collagen, is an important component for basement membrane functions as it can form anchoring fibrils which mediate dermal-epidermal adhesion (27). Therefore, loss of function of type VII collagen can lead to a RDEB and increases the risk of skin cancer (7). Martins et al (28) reported that type VII collagen suppressed TGF-β signaling and angiogenesis in cutaneous SCC. However, certain studies have reported that high COL7A1 expression was associated with poor prognosis in gastric cancer and esophageal squamous cell carcinoma (15,17). However, to the best of our knowledge, the mechanism is still unclear and it may serve different roles in different cancer types.

Despite vigorous research focusing on PC, the improvement in its prognosis remains poor due to its aggressive nature and insensitivity to treatment (11). Therefore, it is imperative to identify efficient and compelling diagnostic and prognostic biomarkers for patients with PC. As such, the present study evaluated COL7A1 expression and the prognostic value of COL7A1 mRNA expression in PC using RT-qPCR and public databases. The results demonstrated that COL7A1 mRNA expression was significantly higher in PC samples and cell lines compared with in normal pancreas tissues. The results of the present study demonstrated that COL7A1 was expressed differentially in human cancer types. For example, it was expressed at significantly higher levels in digestive carcinomas (esophageal cancer, gastric cancer, hepatocellular carcinoma, colorectal cancer and PC) (Fig. 1A) compared with normal tissues, which indicated that COL7A1 might serve as an oncogene in digestive system cancers. Furthermore, the diagnostic value of COL7A1 in PC was evaluated and the ROC curve indicated that the mRNA expression levels of COL7A1 could be an effective biomarker for PC diagnosis.

To further understand the biological functions and mechanism of COL7A1 in PC, GO and KEGG analyses and GSEA were performed. The results of GO and KEGG enrichment analyses indicated that the DEGs were involved in ‘Ion transport’, ‘Ion channel activity’, ‘cAMP signaling pathway’, ‘Calcium signaling pathway’ and ‘ECM-receptor interaction’. Recently, numerous studies have reported that ion transport and calcium channels serve an important role in tumorigenesis and progression (29–31). These findings supported further investigation of the function of COL7A1 in PC. GSEA results suggested that COL7A1 was associated with ‘ECM receptor interaction’, ‘cell cycle’, ‘DNA replication’, ‘G2M checkpoint’, ‘epithelial mesenchymal transition’, ‘E2F targets’ and ‘mitotic spindle’. These pathways have all been reported to participate in the proliferation, invasion and metastasis of PC (32–34).

TIICs exhibit a dual role in the development of PC, as it can inhibit tumor progression as well as promote cancer cells' escape of immune surveillance (35–37). ssGSEA was used to evaluate the association between COL7A1 mRNA expression and TIICs in PC. The results demonstrated that COL7A1 mRNA expression was significantly correlated with Th2 and NK cell levels. Yang et al (38) reported that an elevated Th2/Th1 ratio could result in an immune-suppressive microenvironment and promote tumor growth. Marcon et al (39) reported that, in PC, the cytotoxicity of NK cells was inhibited and favored tumor escape of immune surveillance. These results partially indicated why COL7A1 mRNA expression exhibited a negative association with the survival of patients with PC. However, it must be acknowledged that the association between TIIC and COL7A1 mRNA expression was only based on the results of TCGA database analysis and the correlation coefficient of TIICs was not high. The complex interactions between TIICs and the PC tumor immune microenvironment still require further study.

The association between COL7A1 mRNA expression and the prognosis of patients with PC was also evaluated. In the present study, Kaplan-Meier survival analysis indicated that low COL7A1 mRNA expression was significantly associated with longer OS, DSS and PFI times. Univariate and multivariate Cox regression analysis demonstrated that COL7A1 mRNA expression was a powerful and independent prognostic biomarker for patients with PC. Furthermore, a prognostic nomogram involving clinicopathological factors and COL7A1 mRNA expression was constructed to predict the survival (OS, DSS and PFI) of patients with PC. This predictive model provides a novel option for assessing the prognosis of patients with PC.

Although a comprehensive analysis of the association between COL7A1 mRNA expression and PC was performed, the present study still included certain limitations. Firstly, the association between COL7A1 mRNA expression and the prognosis of patients with PC should be validated using more clinical samples. Secondly, the raw data, excluding the PCR data, were acquired from public databases and bias due to confounding factors may be unavoidable. Furthermore, the functional and molecular mechanisms associated with COL7A1 need to be further elucidated. Additionally, the clinical role of COL7A1 in PC requires further assessment in the near future.

To the best our knowledge, the present study was the first to evaluate the prognostic and immunologic values of COL7A1 in PC. However, it is limited by the aforementioned limitations and a deeper mechanistic investigation of COL7A1 is required.

Overall, the present study demonstrated that COL7A1 expression was upregulated in PC tissues and high mRNA expression levels of COL7A1 were an independent risk factor in patients with PC. Furthermore, its expression may be associated with tumor immune infiltration cells. Based on these findings, it was proposed that COL7A1 could be a novel biomarker for the diagnosis and prognosis of PC.

Supplementary Material

Supporting Data

Acknowledgments

Not applicable.

Availability of data and materials

The data used to support the bioinformatic results are available at the TCGA [pancreatic adenocarcinoma (PAAD) dataset; https://portal.gdc.cancer.gov], GTEx (https://xenabrowser.net/datapages/; TOIL RSEM tpm dataset) and GEO databases (accession no. GSE15471 and GSE101448; http://www.ncbi.nlm.nih.gov/geo/). The other datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

MD and QH designed the study. CD and ZY collected the data, performed the analysis and wrote the manuscript. ZY performed the RT-qPCR experiments. JZ and XL revised the statistical findings and edited and revised the manuscript. MD and QH confirm the authenticity of the raw data. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

All authors declare that they have no competing interests.

References 1

Siegel

Miller

Fuchs

Jemal

Cancer statistics, 2021

CA Cancer J Clin717332021

10.3322/caac.21654

33433946

Huang

Lok

Ngai

Zhang

Yuan

Lao

Chong

Zheng

Wong

MCS

Worldwide burden of, risk factors for, and trends in pancreatic cancer

Gastroenterology1607447542021

10.1053/j.gastro.2020.10.007

33058868

Khalaf

El-Serag

Abrams

Thrift

Burden of pancreatic cancer: From epidemiology to practice

Clin Gastroenterol Hepatol198768842021

10.1016/j.cgh.2020.02.054

32147593

Anderson

Simon

The tumor microenvironment

Curr Biol30R921R9252020

10.1016/j.cub.2020.06.081

32810447

Mao

Wang

Liang

Hua

Liu

Zhang

Meng

Shi

Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: New findings and future perspectives

Mol Cancer201312021

10.1186/s12943-021-01428-1

34635121

Ziani

Chouaib

Thiery

Alteration of the antitumor immune response by cancer-associated fibroblasts

Front Immunol94142018

10.3389/fimmu.2018.00414

29545811

Beer

Murray

Mast cells have prognostic value in Merkel cell carcinoma

Am J Dermatopathol3027302008

10.1097/DAD.0b013e31815c932a

18212540

Kolset

Pejler

Serglycin: A structural and functional chameleon with wide impact on immune cells

J Immunol187492749332011

10.4049/jimmunol.1100806

22049227

Weniger

Honselmann

Liss

The extracellular matrix and pancreatic cancer: A complex relationship

Cancers (Basel)103162018

10.3390/cancers10090316

30200666

Leinwand

Miller

Regulation and modulation of antitumor immunity in pancreatic cancer

Nat Immunol21115211592020

10.1038/s41590-020-0761-y

32807942

Mizrahi

Surana

Valle

Shroff

Pancreatic cancer

Lancet395200820202020

10.1016/S0140-6736(20)30974-0

32593337

Gelse

Pöschl

Aigner

Collagens-structure, function, and biosynthesis

Adv Drug Deliv Rev55153115462003

10.1016/j.addr.2003.08.002

14623400

Klingberg

Mortimore

Parkes

Chick

Clague

Murrell

Weedon

Glass

Prenatal diagnosis of dominant dystrophic epidermolysis bullosa, by COL7A1 molecular analysis

Prenat Diagn206186222000

10.1002/1097-0223(200008)20:8<618::AID-PD864>3.0.CO;2-8

10951471

Martins

Vyas

Chen

Purdie

Mein

South

Storey

McGrath

O'Toole

Increased invasive behaviour in cutaneous squamous cell carcinoma with loss of basement-membrane type VII collagen

J Cell Sci122178817992009

10.1242/jcs.042895

19435799

Kita

Mimori

Tanaka

Matsumoto

Haraguchi

Ishikawa

Matsuzaki

Fukuyoshi

Inoue

Natsugoe

Clinical significance of LAMB3 and COL7A1 mRNA in esophageal squamous cell carcinoma

Eur J Surg Oncol3552582009

10.1016/j.ejso.2008.01.025

18331784

Baba

Iyama

Honda

Ishikawa

Miyanari

Baba

Cytoplasmic expression of type VII collagen is related to prognosis in patients with esophageal squamous cell carcinoma

Oncology712212282006

10.1159/000106426

17652943

Lee

Sohn

Bae

Choi

Kim

Prognostic value of highly expressed type VII collagen (COL7A1) in patients with gastric cancer

Pathol Oncol Res2716098602021

10.3389/pore.2021.1609860

34512204

Wickham

(2016) ggplot2: Elegant graphics for data analysis

Springer-Verlag

New York

ISBN 978-3-319-24277-4

Robin

Turck

Hainard

Tiberti

Lisacek

Sanchez

Müller

pROC: An open-source package for R and S+ to analyze and compare ROC curves

BMC Bioinformatics12772011

10.1186/1471-2105-12-77

21414208

Love

Huber

Anders

Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2

Genome Biol155502014

10.1186/s13059-014-0550-8

25516281

Wang

Han

clusterProfiler: An R package for comparing biological themes among gene clusters

OMICS162842872012

10.1089/omi.2011.0118

22455463

Hänzelmann

Castelo

Guinney

GSVA: Gene set variation analysis for microarray and RNA-seq data

BMC Bioinformatics1472013

10.1186/1471-2105-14-7

23323831

Bindea

Mlecnik

Tosolini

Kirilovsky

Waldner

Obenauf

Angell

Fredriksen

Lafontaine

Berger

Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer

Immunity397827952013

10.1016/j.immuni.2013.10.003

24138885

Nagy

Munkácsy

Győrffy

Pancancer survival analysis of cancer hallmark genes

Sci Rep1160472021

10.1038/s41598-021-84787-5

33723286

Livak

Schmittgen

Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) method

Methods254024082001

10.1006/meth.2001.1262

11846609

Walker

Mojares

Del Río Hernández

Role of extracellular matrix in development and cancer progression

Int J Mol Sci1930282018

10.3390/ijms19103028

30287763

Gatseva

Sin

Brezzo

Van Agtmael

Basement membrane collagens and disease mechanisms

Essays Biochem632973122019

10.1042/EBC20180071

31387942

Martins

Caley

Moore

Szentpetery

Marsh

Murrell

Kim

Avari

McGrath

Cerio

Suppression of TGFβ and angiogenesis by type VII collagen in cutaneous SCC

J Natl Cancer Inst108djv2932015

10.1093/jnci/djv293

26476432

Pedersen

Flinck

Pardo

The interplay between dysregulated ion transport and mitochondrial architecture as a dangerous liaison in cancer

Int J Mol Sci2252092021

10.3390/ijms22105209

34069047

Marchi

Giorgi

Galluzzi

Pinton

Ca2+ fluxes and cancer

Mol Cell78105510692020

10.1016/j.molcel.2020.04.017

32559424

Han

Liu

Zhang

Men

Huo

Geng

Wang

Gao

Zhang

Jia

Mechanosensitive ion channel Piezo1 promotes prostate cancer development through the activation of the Akt/mTOR pathway and acceleration of cell cycle

Int J Oncol556296442019

31322184

Oshi

Newman

Tokumaru

Yan

Matsuyama

Endo

Katz

MHG

Takabe

High G2M pathway score pancreatic cancer is associated with worse survival, particularly after margin-positive (R1 or R2) resection

Cancers (Basel)1228712020

10.3390/cancers12102871

33036243

Krebs

Mitschke

Losada

Schmalhofer

Boerries

Busch

Boettcher

Mougiakakos

Reichardt

Bronsert

The EMT-activator Zeb1 is a key factor for cell plasticity and promotes metastasis in pancreatic cancer

Nat Cell Biol195185292017

10.1038/ncb3513

28414315

Zhou

Liu

Zhang

Wang

Liu

Yao

The epithelial to mesenchymal transition (EMT) and cancer stem cells: Implication for treatment resistance in pancreatic cancer

Mol Cancer16522017

10.1186/s12943-017-0624-9

28245823

Fang

Saiyin

Zhao

Han

Lou

IL-8-positive tumor-infiltrating inflammatory cells are a novel prognostic marker in pancreatic ductal adenocarcinoma patients

Pancreas456716782016

10.1097/MPA.0000000000000520

26495785

Zhang

Velez-Delgado

Mathew

Mendez

Flannagan

Rhim

Simeone

Beatty

di Magliano

Myeloid cells are required for PD-1/PD-L1 checkpoint activation and the establishment of an immunosuppressive environment in pancreatic cancer

Gut661241362017

10.1136/gutjnl-2016-312078

27402485

Protti

De Monte

Immune infiltrates as predictive markers of survival in pancreatic cancer patients

Front Physiol42102013

10.3389/fphys.2013.00210

23950747

Yang

Tao

Jiang

Yang

Huo

Liu

Lin

Perineural invasion reprograms the immune microenvironment through cholinergic signaling in pancreatic ductal adenocarcinoma

Cancer Res80199120032020

10.1158/0008-5472.CAN-19-2689

32098780

Marcon

Zuo

Pearce

Nicol

Margielewska-Davies

Farhat

Mahon

Middleton

Brown

Roberts

Moss

NK cells in pancreatic cancer demonstrate impaired cytotoxicity and a regulatory IL-10 phenotype

Oncoimmunology918454242020

10.1080/2162402X.2020.1845424

33299656

Figure 1.

mRNA expression levels of COL7A1 in pan-cancerous and PC samples. (A) mRNA expression levels of COL7A1 in tumor and normal (unpaired) samples based on TCGA and GTEx datasets. (B) mRNA expression levels of COL7A1 in PC and normal (unpaired) pancreas tissues based on TCGA and GTEx datasets. (C) COL7A1 mRNA expression between tumor and normal (paired) samples in the GSE15471 dataset. (D) COL7A1 mRNA expression between tumor and normal (unpaired) samples in the GSE101448 dataset. (E) Receiver operating characteristic curve analysis was used to evaluate the diagnostic value of COL7A1 mRNA expression in PC. (F) Volcano plot of the differentially expressed genes. *P<0.05, **P<0.01 and ***P<0.001. AUC, area under the curve; COL7A1, collagen type VII α1 chain; GTEx, Genotype-Tissue Expression; ns, not significant; P.adj, adjusted P-value; PC, pancreatic cancer; TCGA, The Cancer Genome Atlas; TPM, transcripts per million; TPR, true positive rate; FPR, false positive rate.

Figure 2.

GO and KEGG enrichment analysis of DEGs. GO enrichment analysis of DEGs for (A) biological processes, (B) cellular components and (C) molecular functions. (D) KEGG pathway enrichment analysis of DEGs. DEGs, differentially expressed genes; ECM, extracellular matrix; GABA, γ-aminobutyric acid; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; PKG, protein kinase G.

Figure 3.

Enrichment plots from Gene Set Enrichment Analysis. (A) COL7A1-related signaling pathways in the c2.cp.kegg.v7.4.symbols.gmt dataset. (B) COL7A1-related signaling pathways in h.all.v7.4.symbols.gmt. COL7A1, collagen type VII α1 chain; E2F, E2 factor; ECM, extracellular matrix; ES, enrichment score; FDR, false discovery rate.

Figure 4.

Association between immune cell infiltration and COL7A1 expression in the tumor microenvironment. (A) Correlation between COL7A1 mRNA expression and immune cell levels. (B-E) Comparison of the abundances of (B) Th2 cells, (C) NK CD56bright cells, (D) eosinophils and (E) Th17 cells between the COL7A1-low and COL7A1-high groups. (F-I) Scatter plots demonstrated the correlation between the abundances of (F) Th2 cells, (G) NK CD56 bright cells, (H) eosinophils and (I) Th17 cells and COL7A1 mRNA expression in pancreatic cancer. aDC, activated dendritic cells. COL7A1, collagen type VII α1 chain; DC, dendritic cells; iDC, immature dendritic cells; NK, natural killer; pDC, plasmacytoid dendritic cells; Tcm, central memory T cells; TFH, T follicular helper cells; TGD, γδ T cells; Th, T helper; TPM, transcripts per million; TReg, regulatory T cells.

Figure 5.

Kaplan-Meier survival plots comparing the COL7A1-high and COL7A1-low groups in pancreatic cancer. Analysis of (A) overall survival, (B) disease-specific survival and (C) progression-free interval using The Cancer Genome Atlas-pancreatic adenocarcinoma dataset. (D) Relapse-free survival analysis using Kaplan-Meier plotter. COL7A1, collagen type VII α1 chain; HR, hazard ratio.

Figure 6.

Prognostic model for the prediction of 1-, 2- and 3-year OS in PC. (A) A nomogram for the prediction of the probability of 1-, 2- and 3-year OS for patients with PC. (B) Calibration plots of the nomogram for the estimation of the probability of OS at 1, 2 and 3 years. (C) Receiver operating characteristic curve of COL7A1 mRNA expression for the prediction of 1-, 2- and 3-year OS for patients with PC. AUC, area under the curve; COL7A1, collagen type VII α1 chain; OS, overall survival; PC, pancreatic cancer; TPR, true positive rate; FPR, false positive rate.

Figure 7.

Reverse transcription-quantitative PCR assessment of COL7A1 mRNA expression in pancreatic cancer cell lines and a normal pancreatic duct cell line. *P<0.05 and ****P<0.0001. COL7A1, collagen type VII α1 chain; ns, not significant.

Table I.

Association between clinicopathological variables and COL7A1 expression in patients with pancreatic cancer.

Characteristic	Low COL7A1 expression, n (%) (n=89)	High COL7A1 expression, n (%) (n=89)
Sex
Female	38 (47.5)	42 (52.5)
Male	51 (52.0)	47 (48.0)
Age, years
≤65	48 (51.6)	45 (48.4)
>65	41 (48.2)	44 (51.8)
T stage
T1	6 (85.7)	1 (14.3)
T2	13 (54.2)	11 (45.8)
T3	65 (45.8)	77 (54.2)
T4	3 (100.0)	0 (0.0)
N stage
N0	28 (56.0)	22 (44.0)
N1	58 (47.2)	65 (52.8)
M stage
M0	39 (49.4)	40 (50.6)
M1	1 (20.0)	4 (80.0)
Pathologic stage
Stage I	14 (66.7)	7 (33.3)
Stage II	69 (47.3)	77 (52.7)
Stage III	3 (100.0)	0 (0.0)
Stage IV	1 (20.0)	4 (80.0)
Histologic grade
G1	23 (74.2)	8 (25.8)
G2	43 (45.3)	52 (54.7)
G3	21 (43.8)	27 (56.2)
G4	1 (50.0)	1 (50.0)
History of chronic pancreatitis
No	69 (53.9)	59 (46.1)
Yes	5 (38.5)	8 (61.5)
History of diabetes
No	52 (48.1)	56 (51.9)
Yes	22 (57.9)	16 (42.1)
Primary therapy outcome
PD	20 (40.8)	29 (59.2)
SD	6 (66.7)	3 (33.3)
PR	5 (50.0)	5 (50.0)
CR	41 (57.7)	30 (42.3)

The total patient number in each group does not equal to 89 for all variables owing to a lack of patient information in certain cases. COL7A1, collagen type VII α1 chain; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease.

Table II.

Association between type VII collagen α1 chain expression and clinicopathological characteristics (logistic regression).

Characteristic	Total, n	Odds ratio (95% CI)	P-value
Sex (male vs. female)	177	1.176 (0.650-2.132)	0.592
Age (>65 vs. ≤65 years)	177	0.852 (0.471-1.538)	0.596
T stage (T3+T4 vs. T1+T2)	175	0.550 (0.243-1.202)	0.139
N stage (N1 vs. N0)	172	0.727 (0.371-1.411)	0.348
M stage (M1 vs. M0)	83	0.342 (0.017-2.800)	0.362
Pathologic stage (stage II–IV vs. stage I)	174	0.444 (0.161-1.130)	0.098
Histologic grade (G3+G4 vs. G1+G2)	175	0.725 (0.372-1.399)	0.340
History of diabetes (yes vs. no)	146	1.330 (0.634-2.822)	0.451
History of chronic pancreatitis (yes vs. no)	141	0.551 (0.159-1.743)	0.319
Primary therapy outcome (SD+PR+CR vs. PD)	139	1.896 (0.941-3.881)	0.076

The total patient number is not consistent for all variables owing to a lack of patient information in certain cases. CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease

Table III.

Univariate and multivariate analysis (overall survival) for prognostic factors in pancreatic cancer.

		Univariate analysis		Multivariate analysis

Characteristic	Total, n	Hazard ratio (95% CI)	P-value	Hazard ratio (95% CI)	P-value
Sex	178
Female	80	-
Male	98	0.809 (0.537-1.219)	0.311
Age, years	178
≤65	93	-
>65	85	1.290 (0.854-1.948)	0.227
T stage	176
T1+T2	31	-
T3+T4	145	2.023 (1.072-3.816)	0.030	1.367 (0.691-2.707)	0.369
N stage	173
N0	50	-
N1	123	2.154 (1.282-3.618)	0.004	1.933 (1.115-3.352)	0.019
M stage	84
M0	79	-
M1	5	0.756 (0.181-3.157)	0.701
Pathologic stage	175
Stage I+II	167	-
Stage III+IV	8	0.673 (0.212-2.135)	0.501
Histologic grade	176
G1+G2	126	-
G3+G4	50	1.538 (0.996-2.376)	0.052	1.301 (0.836-2.024)	0.243
History of chronic pancreatitis	141
No	128	-
Yes	13	1.177 (0.562-2.464)	0.666
History of diabetes	146
No	108	-
Yes	38	0.927 (0.532-1.615)	0.790
COL7A1	178
Low	89	-
High	89	1.908 (1.254-2.904)	0.003	1.543 (1.010-2.358)	0.045

The total patient number does not equal 178 for all variables owing to a lack of patient information in certain cases. COL7A1, collagen type VII α1 chain.