This article is mentioned in:

Introduction

Pancreatic adenocarcinoma (PAAD) is a gastrointestinal malignancy with high mortality. The 5-year survival rate for patients with pancreatic cancer is <5% (1,2). It is expected that pancreatic cancer will rank second in terms of mortality rate among cancers worldwide in the next 20 years (3). For patients with pancreatic cancer, the only cure is surgical resection, but the majority of patients are diagnosed with local inoperable tumors or distant metastasis; thus, patients with pancreatic cancer have a poor prognosis (4,5). Therefore, further research is urgently required to develop effective prevention measures and early diagnostic methods. In the past few decades, efforts to study the molecular mechanisms of pancreatic cancer have provided hope for molecular diagnostics and molecular targeted therapy for various diseases.

Genes with significantly high expression levels in pancreatic cancer include KRAS, BRAF serine/threonine kinase (BRAF) and AKT serine/threonine kinase 2 (AKT2), as previously reported (6). These genes may be used as biomarkers for early diagnosis of pancreatic cancer. Gene expression profile analysis is a high-throughput method for detecting mRNA expression in tissues or cell samples. For instance, based on data from Gene Expression Omnibus (GEO), Long et al (7) screened the differentially expressed genes (DEGs) in pancreatic cancer and analyzed the copy number variation in DEGs. Their study indicated that transforming growth factor β receptor 1 (TGFBR1) and transforming growth factor β 1 (TGFB1) have an important role in the development of pancreatic cancer. The Wnt (8) and hedgehog (9) signaling pathways have been identified as being of marked significance in pancreatic cancer. microRNAs (miRNAs or miRs) have attracted widespread attention in recent years (10). A previous study indicated that miRNA-27a promotes the proliferation of pancreatic cancer cells by activating the Wnt/β-catenin signaling pathway (11). miR-132 has recently been demonstrated to promote pancreatic cancer cell proliferation and inhibit apoptosis through the hedgehog signaling pathway (12). Despite these tremendous advances, the underlying key mechanisms of pancreatic cancer require to be further elucidated to screen promising prognostic biomarkers and potential targets for diagnosis and treatment of pancreatic cancer. In the present study, genes associated with pancreatic cancer were determined from datasets obtained from the online database GEO. GEO2R analysis was performed to identify the DEGs associated with pancreatic cancer (13). Further enrichment analysis of DEGs was applied to explore the molecular mechanisms associated with pancreatic cancer. The core genes in the development of pancreatic cancer were then explored through the analysis of differential gene-protein networks and sub-network modules. Overexpression of integrin subunit α 2 (ITGA2) and ITGB6 was determined to be associated with poor prognosis. Silencing of ITGB6 inhibited cell proliferation in pancreatic cancer and produced cell cycle arrest at G2/M phase.

Materials and methods

Screening of DEGs from GEO datasets

The GEO database is an international public database of datasets, including data from single- and dual-channel determination of mRNA expression and experimental data for genomic DNA and proteins (14). In the present study, three expression profiling datasets [GSE28735(15), GSE16515(16) and GSE15471(17)] were downloaded from GEO (https://www.ncbi.nlm.nih.gov/geo/). GSE15471 contains 39 tumor and paired adjacent normal tissues; GSE16515 contains 36 tumor and 16 adjacent normal tissues; and GSE28735 contains 45 tumor and paired adjacent normal tissues. GEO2R (13) was used to screen for DEGs, and those DEGs shared by the three sets of expression profiles were further selected by the Venn mapping tool. A log2 fold change >1.5 and adjusted P<0.05 were considered to indicate a statistically significant difference.

Database for Annotation, Visualization and Integrated Discovery database enrichment analysis

Cluster Profiler is an ontology-based R package that automates the process of biological-term classification and the enrichment analysis of gene clusters, and provides a visualization module for displaying the analysis results (18). DEGs were subjected to Gene Ontology (GO) enrichment analysis [molecular function (MF), biological process (BP) and cellular component (19,20)] and signaling pathway Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis using Cluster Profiler V3.6.0. The enrichment analysis and function annotation data were obtained and displayed in the form of a bubble chart. P<0.05 was considered to indicate a statistically significant difference.

Protein interaction network construction and sub-network module analysis

The Search Tool for the Retrieval of Interacting Genes/proteins (STRING; https://string-db.org/) is an online tool for searching for gene interactions and protein interactions (21). Selected DEGs were inputted into the online database STRING to generate a protein network diagram, which was visualized by Cytoscape v3.7.0 software (https://cytoscape.org/). Degree was used as the criterion for screening key target genes (where the degree of nodes indicates the number of proteins that the nodes are able to interact with). The sub-network modules related to pancreatic cancer development were analyzed using MCODE plug-in.

Gene Expression Profiling Interactive Analysis (GEPIA) online survival analysis

GEPIA is a newly developed interactive web server for analyzing the RNA sequencing expression data of 9,736 tumors and 8,587 normal samples from The Cancer Genome Atlas (TCGA) and the GTEx projects using a standard processing pipeline. It is able to perform survival and correlation analyses of DEGs (22). In the present study, candidate key genes were incorporated into the GEPIA database to further verify their expression in normal pancreatic and pancreatic cancer tissues. To produce the survival curves for key genes, the genes to be analyzed were searched in the main interface of the GEPIA database (http://gepia.cancer-pku.cn/index.html). Subsequently, ‘Survival Plots’ was selected in the analysis toolbar, the tumor type was set to PAAD and the confidence interval was set to 95%. For the other parameters, the database's default settings were used.

Screening of cell lines for expression of ITGb6 by reverse transcription-quantitative (RT-q)PCR

A total of six pancreatic cancer cell lines (BXPC-3, CFPAC-1, MIA PaCa-2, ASPC-1, PANC-1 and SW1990) were provided by Shanghai GeneChem Co., Ltd. and cultured in RPMI-1640 basic medium (Corning, Inc.). All cells were routinely subcultured at 37˚C in the presence of 5% CO2 in an incubator with saturated humidity. Total RNA was extracted from the six cell lines using TRIzol reagent according to the manufacturer's instructions. RNA was reverse transcribed to complementary DNA using Promega M-MLV at 42˚C (Promega Corp.). The mRNA expression levels of the ITGB6 gene in different cell lines of interest were detected by quantitative PCR using a LightCycler 480 II (Roche Molecular Systems, Inc.). The composition of the reaction mixture was SYBR premix ex taq 6.0 µl, primer mix 0.3 µl, reverse transcription product 0.6 µl and RNase-free H2O 5.1 µl. The reaction conditions were as follows pre-denaturation at 95˚C for 30 sec, followed by denaturation for 5 sec at 95˚C and annealing for 30 sec at 60˚Cfor a total of 40 cycles. The primer sequences were as follows: ITGB6 forward, 5'-TGATCTTCGCTGTAACCC-3' and reverse, 5'-CAGACCGCAGTTCTTCATA-3'; GAPDH forward, 5'-TGACTTCAACAGCGACACCCA-3' and reverse, 5'-CACCCTGTTGCTGTAGCCAAA-3'. The experimental results were analyzed by the 2-∆∆Cq method (23) for relative quantitative analysis.

Lentivirus (LV) transfection to silence ITGB6 expression in the ASPC-1 cell line

Negative control (NC) virus CON077 and LV-ITGB6-RNA interference (RNAi) LV were constructed by Shanghai Jikai. The ITGB6-small interfering (si)RNA target sequence designed for the ITGB6 gene sequence was 5'-gcCTCCAAACATTCCCATGAT-3'. The CON077 sequence was 5'-TTCTCCGAACGTGTCACGT-3'. ASPC-1 cells with relatively high expression were selected for transfection and the following experimental groups were established: i) Mock group, normal ITGB6 cells; ii) Short hairpin (sh)NC group, ASPC-1 cells transfected with recombinant LV carrying NC-siRNA; iii) shITGB6 group, ASPC-1 cells transfected with recombinant LV carrying ITGB6-siRNA. Total RNA was extracted from cells in the shNC and shITGB6 groups using Total RNA extraction reagent (Shanghai Pufei) according to the manufacturer's instructions. and RT-qPCR was used to detect the expression levels of ITGB6 mRNA in the cells.

Cell proliferation assays

Cells transfected for 24 h were inoculated into 96-well plates and cultured for a further 24 h before MTT (Shanghai Dingguo Biotechnology Co., Ltd.) solution (20 µl at 5 g/l) was added to each well. After incubation at 37˚C for 4 h, the supernatant was discarded and dimethyl sulfoxide was added to each well (100 µl). The absorbance of each well was determined at a wavelength of 490 nm.

Cell cycle detection

The cells from the different experimental groups were digested with trypsin and centrifuged to collect the cells. After washing with D-Hanks' solution (pH 7.2) pre-cooled at 4˚C, cells were fixed with 75% ethanol at 4˚C for 1 h. The cell cycle was detected by flow cytometry, as previously described (24).

Statistical analysis

Values are expressed as the mean ± standard deviation and all experiments were repeated independently three times. Statistical analyses were performed with SPSS 24.0 statistical software (IBM Corp.). Graphs and curves were constructed by GraphPad Prism 7 (GraphPad Software, Inc.). An independent-samples t-test was used to assess differences between paired samples. One-way ANOVA was used for comparison between groups with the least-significant difference method used for pairwise comparisons. P<0.05 was considered to indicate a statistically significant difference.

Results

Identification of DEGs in PAAD

In the present study, the GEO2R analysis platform was used to preprocess and filter the original data of three selected datasets (GSE28735, GSE16515 and GSE15471). The Venn diagram (Fig. 1) contained a total of 72 genes with an intersection in the three datasets, including 20 downregulated and 52 upregulated genes (Table I).

Figure 1 Analysis of the common differentially expressed genes in the microarray datasets of GSE28735, GSE16515 and GSE15471. (A) Upregulated genes. (B) Downregulated genes.

Table I Differentially expressed genes (n=72).

Table I

Differentially expressed genes (n=72).

Item	Gene names
Downregulated genes	ALB, ANPEP, AQP8, CELA2B, EGF, ERO1B, ERP27, FGL1, GP2, KIAA1324, KLK1, PAIP2B, PDIA2, PDK4, PNLIPRP1, PNLIPRP2, RBPJL, SERPINI2, TMED6, TRHDE
Upregulated genes	AGR2, AHNAK2, ANLN, ANTXR1, ANXA10, CEACAM5, CEACAM6, CEMIP, CLDN18, COL10A1, COL11A1, COL12A1, COL1A1, CP, CST1, CTRL, CTSE, CXCL5, DPCR1, EDIL3, FERMT1, FN1, FXYD3, GABRP, GATM, INHBA, ITGA2, ITGB6, KRT17, KRT19, KRT7, LAMB3, LAMC2, MMP11, MMP12, NOX4, NR5A2, PLAC8, POSTN, SDR16C5, SERPINB5, SLC6A14, SLPI, SULF1, TCN1, TFF1, THBS2, TMC5, TMPRSS4, TSPAN1, VCAN, VSIG1

Differential gene enrichment analysis

In order to further study the role of DEGs in the development of pancreatic cancer, Cluster Profiler was used to subject the DEGs to GO functional enrichment and KEGG pathway analysis. GO analysis indicated that the DEGs mainly act on the extracellular matrix (ECM); are involved in BPs such as cell-substrate adhesion, extracellular structure organization and ECM organization; and participate in the composition of the collagen-containing ECM and endoplasmic reticulum lumen. Other MFs such as ECM structural constituent, endopeptidase activity and glycosaminoglycan binding were also determined (Table II; Fig. 2A-C). Through KEGG signaling pathway analysis, it was indicated that the DEGs are mainly involved in human papillomavirus infection, ECM-receptor interaction, focal adhesion, PI3K-Akt signaling pathway, protein digestion and absorption, and pathways in pancreatic secretion (Table III; Fig. 2D).

Figure 2 Functional enrichment and pathway analysis. (A-C) Gene ontology terms in the categories (A) biological process, (B) cellular component and (C) molecular function. (D) Kyoto Encyclopedia of Genes and Genomes pathways. ECM, extracellular matrix.

Table II GO enrichment analysis of the differentially expressed genes.

Table II

GO enrichment analysis of the differentially expressed genes.

Category	Term	Count	P-value	Exemplary genes
BP	GO:0030198~extracellular matrix organization	12	3.55x10-10	LAMB3, ERO1B, ITGB6
BP	GO:0007155~cell adhesion	12	2.12x10-6	LAMB3, ITGB6, FERMT1
BP	GO:0035987~endodermal cell differentiation	5	4.02x10-6	INHBA, LAMB3, COL12A1
BP	GO:0030574~collagen catabolic process	6	5.94x10-6	COL1A1,MMP12, COL11A1
BP	GO:0022617~extracellular matrix disassembly	5	2.53x10-4	LAMB3, LAMC2, FN1
CC	GO:0005615~extracellular space	24	4.85x10-10	CXCL5, CST1, POSTN
CC	GO:0005576~extracellular region	22	3.63x10-7	PNLIPRP1, PNLIPRP2, CXCL5
CC	GO:0070062~extracellular exosome	27	7.40x10-6	FXYD3, TSPAN1, KIAA1324
CC	GO:0031012~extracellular matrix	9	1.74x10-5	COL12A1, SLPI, POSTN
MF	GO:0005509~calciumion binding	10	0.001453	PNLIPRP1, SULF1, ANXA10
MF	GO:0004252~serine-type endopeptidase activity	6	0.002791	CELA2B, KLK1, MMP12
MF	GO:0003756~protein disulfide isomerase activity	3	0.003112	ERO1B, ERP27, PDIA2
MF	GO:0005178~integrin binding	4	0.007401	ITGB6, ITGA2, EDIL3, FN1

[i] GO, gene ontology; BP, biological process; CC, cellular component; MF, molecular function.

Table III Kyoto Encyclopedia of Genes and Genomes pathways significantly enriched by the differentially expressed genes.

Table III

Kyoto Encyclopedia of Genes and Genomes pathways significantly enriched by the differentially expressed genes.

Term	Count	P-value	Exemplary genes
hsa04512:ECM-receptor interaction	8	1.06x10-7	LAMB3, ITGB6, ITGA2
hsa04510:Focal adhesion	9	3.20x10-6	LAMB3, ITGB6, ITGA2
hsa04151:PI3K-Akt signaling pathway	6	1.34x10-4	LAMB3, ITGB6, ITGA2
hsa04974:Protein digestion and absorption	9	4.69x10-5	COL12A1, CELA2B, COL1A1
hsa04972:Pancreatic secretion	4	0.00893	PNLIPRP2, CELA2B, CTRL

[i] ECM, extracellular matrix; Hsa, Homo sapiens.

Protein network and sub-network module analysis

A protein-protein interaction (PPI) network for DEGs was constructed using the STRING database (Fig. 3A). The first 10 nodes [fibronectin (FN)1, EGF, albumin (ALB), collagen α1 chain 1 (COL1A1), integrin subunit α2(ITGA2), keratin 19(KRT19), collagen type XI α 1 chain (COL11A1), thrombospondin 2 (THBS2), integrin subunit β6 (ITGB6) and matrix metallopeptidase 12 (MMP12)] with the highest degrees were screened as hub genes. In order to further explore the association in the PPI network, the first three modules in the PPI network were extracted and certain key genes [periostin (POSTN), matrix metallopeptidase 11 (MMP11) and KRT19] were indicated to have local regulatory roles in the development and progression of pancreatic cancer (Fig. 3B). This provides additional grounds for studying the molecular mechanisms of pancreatic cancer.

Figure 3 PPI network complex of DEGs in pancreatic adenocarcinoma. (A) PPI networks of the DEGs. In the figure, a larger sphere area indicates a greater degree value of the target protein. Lines represent the protein interaction between the genes. (B) Three sub-modules in a PPI network. POSTN, MMP11 and KRT19 are located in three sub-modules. PPI, protein-protein interaction; DEG, differentially expressed gene; POSTN, periostin; MMP11, matrix metallopeptidase 11; KRT19, keratin 19.

Online survival analysis through GEPIA

To explore the association between the expression of key genes and the prognosis of patients with pancreatic cancer, the key genes were inputted into the GEPIA database for survival analysis. It was indicated that the expression levels of the ITGB6 and ITGA2 genes are closely associated with the patients' survival rate (Fig. 4A). This means that these genes have a negative impact on the survival time of patients with pancreatic cancer. Finally, using the GEPIA online database, it was verified that these genes were highly expressed in pancreatic cancer (Fig. 4B).

Figure 4 Prognostic values of hub genes in patients with PAAD from the GEPIA database. (A) Survival analysis of hub genes in protein-protein interaction networks (overall survival and disease-free survival for hub genes). Dotted lines indicate the 95% confidence interval range. Red represents high gene expression, blue represents low gene expression. (B) Verification results of key gene expression in pancreatic cancer tissues. Red color indicates tumor tissues, while grey color indicates normal tissues. Each dot representing a distinct tumor or normal sample. *P<0.05. PAAD, pancreatic adenocarcinoma; HR, hazard ratio; ITGA2, integrin subunit α 2; T, tumor sample; N, normal sample; TPM, Transcripts Per Kilobase of exon model per Million mapped reads.

Expression of ITGB6 in various cell lines

As indicated by the RT-qPCR results, the expression levels of the gene ITGB6 were low in MIA PaCa-2 and PANC-1 cells, but high in BXPC-3, CFPAC-1, ASPC-1 and SW1990 cells (Fig. 5A). Therefore, ASPC-1 cells were selected for the subsequent experiments.

Figure 5 (A) Expression abundance of the ITGB6 gene in different cell lines. When the ∆Cq value was ≤12, the gene expression abundance in the cell was high; when the ∆Cq value was ≥16, the gene expression abundance in the cell was low. (B) Recombinant lentivirus infection image in ASPC-1 cells (upper panels, bright field; lower panels, green fluorescence field; magnification, x100). (C) Results of the reverse transcription-quantitative PCR analysis of ITGB6. *P<0.05. ITGA6, integrin subunit β 6; shNC, negative control shRNA; shITGB6, shRNA targeting ITGB6; shRNA, short hairpin RNA; Cq, quantification cycle.

Results of LV-ITGB6-RNAi lentivirus infection in AsPC-1 cells

Observation of the green fluorescent protein encoded in the plasmid with ITGB6 under a fluorescence microscope confirmed that 24 h after AsPC-1 cells had been infected with the recombinant virus LV-ITGB6-RNAi, the expression of green fluorescent protein was observed in both groups of cells (Fig. 5B). The transfection efficiency detected by flow cytometry was >80%. RT-qPCR analysis demonstrated that the ITGB6 silencing effect of the plasmid was evident 72 h after infection (P<0.05; Fig. 5C).

Effect of ITGB6 on the proliferation of pancreatic cancer ASPC-1 cells

According to the results of the growth curve analysis, compared with that of the shNC group, the growth curve of the shITGB6 group was flatter and the overall growth rate of cells was significantly reduced. From the fourth day onwards, the proliferation rate of cells in the experimental group was significantly lower than that of cells in the control group and the difference was statistically significant (P<0.01). It was indicated that knockdown of ITGB6 significantly inhibited the proliferation of ASPC-1 cells (Fig. 6A).

Figure 6 (A) Effect of ITGB6 gene silencing on the proliferation of pancreatic cancer ASPC-1 cells. The OD490 value indicates the number of viable cells, while the OD490/fold indicates the multiplication ratio on each day. (B and C) Effect of ITGB6 gene silencing on the cell cycle of pancreatic cancer ASPC-1 cells. Compared with the number of cells in each phase of the cell cycle in the control group, the results of the statistical analysis of the number of cells in G1, S and G2 phases in shITGB6 group was P<0.05. *P<0.05, **P<0.01. OD490, optical density at 490 nm; ITGB6, integrin subunit β 6; shNC, negative control shRNA; shITGB6, shRNA targeting ITGB6; shRNA, short hairpin RNA.

Effect of ITGB6 on the cell cycle of pancreatic cancer ASPC-1 cells

Flow cytometric analysis of the cell cycle with PI indicated that, compared with that of the shNC group, the percentage of cells in G1- and S-phase in the shITGB6 group was significantly reduced (P<0.05), while the percentage of cells in G2/M-phase was significantly increased (P<0.01). This suggested that silencing of ITGB6 gene expression causes cell-cycle arrest in the G2/M phase, thus significantly inhibiting the cell cycle (Fig. 6B and C).

Discussion

Pancreatic cancer is a highly malignant tumor type of the digestive tract (25). Due to the lack of specific clinical manifestations in the early stage of the disease, >80% of patients are diagnosed at an advanced stage, where treatment becomes markedly difficult (26). Revealing the molecular mechanisms involved in the development of pancreatic cancer will help to discover novel tumor markers that may allow for an early diagnosis of pancreatic cancer, develop novel effective treatment strategies, evaluate prognosis and improve patient survival. In the present study, datasets of pancreatic cancer-related chips from the GEO database were analyzed and 72 significant DEGs were mined.

GO functional analysis indicated that the DEGs are mainly involved in BPs such as cell adhesion and ECM decomposition. KEGG signaling pathway analysis revealed the involvement of focal adhesion, ECM-receptor interaction, PI3K-Akt signaling pathway and protein digestion and absorption. Of these, the PI3K-Akt signaling pathway is particularly noteworthy, as it is known to be an important oncogenic signaling pathway involved in tumorigenesis and resistance to targeted anticancer therapies for various tumor types (27). The PI3K-Akt signaling pathway is also an important cause of pancreatic cancer and is associated with poor prognosis of pancreatic cancer. Abnormal activation of this pathway involves cellular metabolism and survival and cell cycle progression. Several inhibitors targeting Akt, PI3K and mTOR have recently been developed for clinical research (28).

In the present study, the STRING online tool was used to analyze the PPIs encoded by the DEGs associated with pancreatic cancer. It was determined that the interactions between the proteins encoded by these genes were mainly concentrated in key node genes such as FN1, EGF, ALB, ITGA2 and ITGB6. FN1 exhibited the highest connectivity in the PPI network, suggesting its important role in pancreatic cancer. A recent study (29) revealed that FN1 was abundantly expressed in the tumor microenvironment of PAAD, which is consistent with the present study, demonstrating that FN1 is highly expressed in PAAD, while its expression in normal pancreatic tissue was low or marginal. Expression of FN1 matrix was associated with aggressive tumor characteristics, including greater tumor size and advanced T and N stages.

ALB and EGF were the only downregulated genes among the hub genes. The ALB gene, which encodes the most abundant protein in human blood, was a downregulated gene in pancreatic cancer. The protein regulates plasma colloid osmotic pressure and serves as a carrier protein for a variety of endogenous molecules and exogenous drugs. A recent study indicated that the combined detection of the derived neutrophil-to-lymphocyte ratio and ALB are able to improve the diagnostic efficiency for pancreatic cancer (30). Another study demonstrated the C-reactive protein (CRP)/ALB ratio may serve as a significant and promising inflammatory prognostic score in pancreatic cancer, since high CRP/Alb indicates a poor prognosis (31). Combined with the results of the present study, this suggests that the detection of ALB levels in patients with pancreatic cancer may improve the sensitivity of pancreatic cancer diagnosis. EGF was indicated to be significantly downregulated in the present study, which is inconsistent with the results of Hao et al (32). A study with a larger sample of patients is required for further validation.

Furthermore, the present study detected novel genes involved in the local regulation of the development of pancreatic cancer, including POSTN, MMP11 and KRT19. PPI module analysis indicated the important role of these central genes, which are involved in key pathways and BPs associated with pancreatic cancer.

COL1A1 is a key structural component of the ECM. It occurs in the majority of connective and embryonic tissues, and is an important member of the collagen family (33). Typically, type I collagen consists of COL1A1 and COL1A2(34). Abnormal expression of COL1A1 has been reported in kidney cancer, hepatocellular carcinoma and melanoma (35-37). Li et al (38) indicated that COL1A1 and COL1A2 were overexpressed in gastric cancer and high COL1A1 may be a monitoring factor for early gastric cancer or a prognostic factor for predicting overall survival. COL1A1 secreted by pancreatic stellate cells (PSCs) promotes invasion and migration of pancreatic cancer cells (39). A previous study (40) has indicated that inhibition of COL1A1 with the novel hydrophilic agent palmatine (PMT) is able to inhibit the growth of PSCs. The role of collagen family members in pancreatic cancer deserves further investigation.

Improving the prognosis of patients with pancreatic cancer is an urgent problem to be solved. In the present study, the key genes selected from pancreatic cancer samples in TCGA database were analyzed and the results suggested that high expression of the ITGA2 and ITGB6 genes is a high-risk factor for poor prognosis in patients with pancreatic cancer. This evaluation of prognosis has obvious clinical significance and future studies focusing on these genes may contribute to the treatment of pancreatic cancer. ITGA2 is an essential collagen receptor on epithelial cells. Gong et al (41) suggested that changes in UCA1 (urothelial cancer associated 1) Expression may affect the expression of ITGA2, further interfering with the progression of cancer. The adhesion spot pathway was identified as the regulatory mechanism of ITGA2. The expression profile of the long non-coding RNA UCA1 was associated with the migration and apoptosis of SW-1990 cells (42). Specifically, upregulated co-expression of UCA1-ITGA2 promoted the migration and invasion of pancreatic cancer cells. Thus, ITGA2 may become a novel potential target for gene therapy.

The mechanisms of ITGB6 in cancer have remained largely elusive and studying the TGF-β/ITGB6 signaling pathway may be worthwhile. ITGB6 has a role in signal transmission from the ECM to cells. TGF-β is an important inflammatory factor produced by macrophages, stromal cells and tumor cells in the tumor microenvironment and is involved in the occurrence, development and metastasis of tumors. A previous study (43) indicated that the TGF-β/ITGB6 signaling pathway has an important role in the invasion and metastasis of esophageal squamous cell carcinoma and the present study suggested that this pathway may be inhibited by miR-17/20a. A recent bioinformatics study on pancreatic cancer used ITGB6 as an independent risk factor for the prognosis of pancreatic cancer, which also supports the present results (44). However, the function and mechanism of ITGB6 in pancreatic cancer are unclear and further studies should be performed to detect the levels of this gene and confirm its role in PAAD. The present study suggested that the expression of ITGB6 mRNA was high in the majority of pancreatic cancer cell lines evaluated, including BXPC-3, CFPAC-1, ASPC-1 and SW1990. On the contrary, according to the RT-qPCR results, the expression of ITGB6 in the MIA PaCa-2 and PANC-1 cell lines was lower than that in the other four cell lines. Infinite cell proliferation caused by imbalances in various stages of the cell cycle is closely associated with formation of tumors (45). In the cellular functional experiments of the present study, the role of ITGB6 as an oncogene was confirmed. Through functional analyses with silencing of ITGB6 and determination of the cell proliferation and cell cycle distribution in ASPC-1 cells, it was indicated that inhibition of ITGB6 decreased cell proliferation and induced G2/M arrest, which supports the results of the present bioinformatics analysis. The effect of silencing the ITGB6 gene on the cell cycle may be a potential mechanism for inhibition of further progression of pancreatic tumors. The above results strongly supported the possibility of ITGB6 as an optimal target for therapeutic intervention. Detection of the cell cycle, particularly the expression of G2/M phase-related regulatory proteins, will be the focus of future studies by our group. At the same time, the study still has certain limitations, such as the lack of western blot data. Further research on the impact of ITGB6 on pancreatic cancer cell invasion and migration, and the potential molecular mechanism of ITGB6 in pancreatic cancer, require to be further explored.

In conclusion, the present study employed a series of bioinformatics methods to identify key genes in pancreatic cancer. ITGA2 and ITGB6 may be used as potential biomarkers for the diagnosis and treatment of patients with pancreatic cancer. The DEGs and metabolic pathways revealed in the present study may help us understand the mechanisms of the molecular development of pancreatic cancer and provide a theoretical basis for future research on clinical targeted therapies. Analytical data mining through bioinformatics analysis is a feasible method to systematically identify potential biomarkers. However, the molecular mechanisms of pancreatic cancer require to be further investigated through biological experiments.

Acknowledgements

Not applicable.

Funding

The present study was supported by Zhenjiang Science and Technology Committee (grant. no. SH 2019061).

Availability of data and materials

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Authors' contributions

XX and TR were responsible for designing the study, performing the experiment, collecting the data and writing the manuscript. XW, XZ and SD were responsible for designing the study, performing the experiment, and collecting the data and reviewing the manuscript. SD was responsible for providing experimental ideas and reviewing the manuscript. XX and TR were responsible for the confirming the authenticity of all the raw data. All authors read and approved the final manuscript.

Ethics approval and consent to participate

All animal experimentation protocols were approved by the Institutional Animal Care and Use Committees of Jiangsu University and were conducted according to the Regulation on Animal Experimentation at Jiangsu University.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References