Abstract

Background

   Less than 10% of people who have pancreatic ductal adenocarcinoma
   (PDAC) will survive the malignancy for five years. The ion channel
   genes-related biomarker and predictive model were needed for
   exploitation.

Methods

   Differentially expressed ion channel genes (DEICGs) were detected in
   PDAC patients. GO and KEGG enrichment analysis was conducted on DEICGs.
   The prognostic genes were found using Cox regression analysis. After
   that, a risk model was created and examined. A nomogram was created
   based on independent predictive analysis. The molecular functions of
   two risk groups were explored. Immune checkpoint molecule expression
   was compared in two risk groups. We evaluated the possible cancer
   immunotherapy response in two risk groups using the TIDE method. We
   further examined how TRPV2 functions in PDAC as a potent oncogene and
   regulates the activity of macrophages by in vitro validation, including
   CCK8, EdU, and Transwell assays.

Results

   A total of twenty-four DEICGs were found. Next, we discovered that two
   DEICGs (TRPV2 and GJB3) were connected to PDAC patients' overall
   survival (OS). The risk model was created and validated, and a nomogram
   was used to forecast the overall survival of PDAC patients. The
   high-risk group considerably accumulated oncogenic pathways.
   Furthermore, we discovered a correlation between the expression of
   critical immunological checkpoints and the risk score. Furthermore,
   patients in the high-risk category had a lower chance of benefiting
   from immune therapy. The HPA database confirmed that TRPV2 is expressed
   as a protein. Lastly, TRPV2 controls macrophage activity and acts as a
   potent oncogene in PDAC.

Conclusion

   Altogether, this study suggested that two ion channel genes, TRPV2 and
   GJB3, were potential biomarkers for the prognosis of PDAC and
   immunotherapy targets, and the research will be crucial for creating
   novel PDAC treatment targets and predictive molecular indicators.

   Keywords: Pancreatic ductal adenocarcinoma, Ion channel genes,
   Predictive model, Nomogram, Immunotherapy

1. Introduction

   By 2030, pancreatic ductal adenocarcinoma (PDAC) is expected to rank
   second in the United States among cancer-related deaths. Currently
   ranked seventh globally, PDAC is expected to claim 75,000 lives from
   cancer. Even worse, in certain nations, the five-year survival rate
   from this illness is only 9%, despite advancements in surgical methods,
   chemotherapeutic regimens, and the use of neo-adjuvant
   chemoradiotherapy [[27]1]. Thus, to provide a foundation for clinical
   prognosis and treatment, it is imperative to identify prognostic
   indicators for PDAC and investigate the molecular targets of
   immunotherapy.

   Membrane proteins called ion channels facilitate ions' passage across
   biological membranes [[28]2]. According to specific research, ion
   channels are essential for regulating cell volume, immunological
   response, muscular contraction, hormone production, gene expression,
   and cell proliferation. Ion channels are involved in many different
   biological processes, which explains why the number of human disorders
   linked to ion channel failure has increased throughout the past few
   years. Specifically, there is mounting evidence that both voltage- and
   ligand-gated ion channels play a role in the development and
   pathophysiology of various human malignancies [[29]3]. Angiogenesis,
   tumor development, metastasis, and vascular permeability are all
   associated with transient receptor potential (TRP) channels [[30]4].
   Numerous ligand-gated channels, including nicotinic acetylcholine
   receptors, influence the growth of tumor cells, their programmed cell
   death, and the formation of new blood vessels [[31]5]. Ion channel
   genes are involved with pancreatic cancer in earlier research. For
   example, a high level of MCOLN1 was linked to poor
   clinical-pathological characteristics and a low survival rate in PDAC
   patients, indicating a potential function for MCOLN1 in PDAC tumor
   growth [[32]6]. According to specific research, overexpressing TRPM2 in
   the PDAC cell line PANC-1 enhanced the potential of the cells to
   proliferate, invade, and spread [[33]7].

   An approach to cancer treatment known as “cancer immunotherapy” targets
   the patient's immune system [[34]8]. It functions by inducing the
   immune system to identify and combat cancerous cells [[35]9]. Cancer
   immunotherapy comes in various forms, such as immune checkpoint
   inhibitors, CAR-T cell treatment, and cancer vaccines [[36][10],
   [37][11], [38][12]]. These therapies are actively being studied and
   developed because they have demonstrated encouraging outcomes in
   treating various cancer types [[39]13]. However, the interconnection
   between ion channels and immunotherapy in PDAC has yet to be explored.

   The prognostic biomarker genes TRPV2 and GJB3, which are based on the
   ion channel gene, were screened, and a novel prognostic model was
   created to accurately predict the prognosis of PDAC patients. TRPV2 and
   GJB3 were further found as potent cancer immunotherapy determinants.

2. Materials and methods

2.1. Data source

   The Cancer Genome Atlas (TCGA) database extracted the transcriptome
   data of 172 pancreatic ductal adenocarcinoma (PDAC) tissues and 4
   normal tissues with complete clinical information. The [40]GSE78229
   dataset's transcriptome data and complete clinical information were
   obtained from the Gene Expression Omnibus (GEO) database. The dataset
   GSE782229 contained 49 PDAC samples with survival data. The HUGO Gene
   Nomenclature Committee (HGNC) database provided the list of 328 ion
   channel genes.

2.2. Identification of differentially expressed ion channel genes (DEICGs)

   The ‘limma' R package was utilized to extract the differentially
   expressed ion channel genes (DEICGs) (P-value <0.05, |log2 Fold Change
   (FC)| > 0.5) in the TCGA-PDAC dataset between normal and PDAC tissues
   [[41]14,[42]15].

2.3. Functional enrichment analysis

   GO (BP, CC, MF) and KEGG enrichment analyses were implemented on the
   Metascape website, and the functions of DEICGs were annotated.

2.4. Construction and validation of the prognostic model

   We randomly split 172 samples from the TCGA-PDAC dataset into 69
   testing and 103 training sets, with a 6:4 ratio. DEICGs that were
   substantially (P-value <0.2) correlated with the OS of the PDAC
   patients in the training sample were found using the univariate Cox
   regression model [[43]16]. The potential DEICGs were then put through a
   stepwise multivariate regression analysis to assess how well they
   predicted patient survival [[44]17,[45]18]. The formula used to
   calculate the risk score was: Riskscore = β1X1 + β2X2+ … + βnXn. The
   gene expression level is represented by X1 in this formula, and the
   regression coefficient is denoted by β. Using the R software's
   “survminer” package, the median risk score was calculated to divide
   patients into high- and low-risk groups. The OS between two risk groups
   was then assessed using the K-M survival curve and the log-rank test.
   Using the R software's “survival ROC” package, the receiver operating
   characteristic (ROC)'s area under the curve (AUC) was determined. The R
   software's “pheatmap” package also depicted the risk plot. The
   [46]GSE78229 dataset's patients (49 samples) served as the validation
   set, and the testing and validation sets underwent the same process to
   verify the risk model.

2.5. Independent prognostic analysis and construction of a nomogram

   Using the TCGA-PDAC dataset, univariate and multivariate Cox regression
   analyses were used to find independent prognostic predictors. Next,
   using the R packages “rms” and “nomogramEx,” a nomogram with the risk
   score and clinicopathological parameters was created. Using the R
   package “regplot,” we created calibration curves (1, 3, and 5 years) to
   confirm the nomogram's correctness.

2.6. Gene set enrichment analysis

   Using the R software's “clusterProfiler” package, KEGG-based GSEA was
   conducted to examine the various signaling pathways that separate the
   low- and high-risk groups in the TCGA-PDAC dataset [[47]19,[48]20].

2.7. Evaluation of immunotherapy efficiency

   There was a comparison of immune checkpoint molecule expression in two
   risk groups. The Tumor Immune Dysfunction and Exclusion (TIDE) is a
   computational technique to predict possible responsiveness to cancer
   immunotherapy [[49]21].

2.8. Verification of gene expression

   The prognostic gene's protein expression was compared using the Human
   Protein Atlas (HPA) database.

2.9. Experimental validation

   The SW1990 and THP-1 cell lines were purchased from iCell (Shanghai,
   China, [50]http://www.icellbioscience.com/search) as previously
   described [[51]22]. SiRNA sequences of TRPV2: Forward
   GCCGGATCCAAACCGATTTGA; Forward GCTGGAGATCATTGCCTTTCA; Forward
   GCTGGCTGAACCTGCTTTACT. Please see the supplementary materials for the
   detailed methods.

2.10. Statistical analysis

   The R project was used for all analyses, and the Wilcoxon and
   Kruskal-Wallis tests were used to compare the data from various groups.
   P-values of less than 0.05 were deemed statistically significant in all
   analyses.

3. Results

3.1. The DEICGs expression profile in PDAC

   We delineated the expression pattern of 328 ion channel genes in normal
   and PDAC samples in the TCGA-PDAC dataset. 24 DEICGs, including 17
   up-regulated and seven down-regulated genes, were screened and listed
   in [52]Supplementary Table 1. The DEICGs expression profile was shown
   in a volcano map and a heatmap ([53]Fig. 1A-B).

Fig. 1.

   [54]Fig. 1
   [55]Open in a new tab

   Identification of DEICGs in PDAC. (A)Volcano plot of all DEICGs. The
   orange dot represents up-regulated genes, the green dot represents
   down-regulated genes, and the grey dot represents unchanged genes.
   DEICGs: differentially expressed ion channel genes. (B) Heat maps of
   DEICGs. Functional and pathway enrichment analysis of DEICGs. (C–D)
   Contains 10 GO terms and 2 KEGG pathways. (For interpretation of the
   references to colour in this figure legend, the reader is referred to