Gene expression data and associated clinical data for PAAD patients were downloaded from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/) database and Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) database (GSE15471, GSE62165, GSE16515 and GSE62452). Additional normal sample data were obtained from Genotype-Tissue Expression (GTEx). We utilized the Gene Expression Profiling Interactive Analysis 2.0 (GEPIA2) platform (http://gepia2.cancer-pku.cn/) to investigate the expression patterns of MVP across multiple cancer types.

The TMA was purchased from Shanghai Outdo Biotech, including 10 primary PAAD tumor tissues and 10 matched adjacent normal tissues for the analysis of MVP expression patterns.

Based on the TCGA-Clinical Data Resource, the correlation between the expression of MVP and the survival rate for different clinical characteristics was estimated using the survival package and the survminer package. The survival analyses were assessed by the Kaplan-Meier curves and Cox proportional hazard regression models. The TCGAplot package (12) was utilized to plot receiver operating characteristic (ROC) curve.

A heatmap of the most differentially expressed genes based on MVP expression and gene set enrichment analysis (GSEA) was generated by TCGAplot (12). Gene Ontology (GO) and Kyoto Encyclopedia of Gene and Genomes (KEGG) pathway analysis were conducted to analyze the biological processes and signal pathways in which genes with the strongest association with MVP in PAAD may participate. STRING (https://cn.string-db.org/) and GeneMANIA (https://genemania.org/) were utilized to analyze the protein–protein interaction network.

Correlation data between the expression of MVP and 28 kinds of tumor-infiltrating immune cells (TIICs) were downloaded from the TIMER2.0 (http://timer.cistrome.org/) (13). The Tumor Immune Dysfunction and Exclusion (TIDE) scores and Microsatellite Instability (MSI) status of PAAD samples (GSE15471 and GSE62542) were calculated by the TIDE database (http://tide.dfci.harvard.edu/). TISIDB database (http://cis.hku.hk/TISIDB/) was applied to investigate the association between MVP expression and immunomodulators (14). Furthermore, TISCH database (http://tisch.comp-genomics.org/home/) (15) was performed to investigate the relationship between MVP expression levels and immune cell at the single-cell level.

The cluster analysis of individual CpGs of MVP in PAAD patients and the assessment of survival differences based on methylation levels at different CpG sites of MVP in patients were conducted using MethSurv (https://biit.cs.ut.ee/methsurv/) (16). Differential analysis of MVP phosphorylation levels and protein abundance between PAAD and adjacent tissues were performed using the cProSite function in CPTAC (https://pdc.cancer.gov/pdc/). The promoter methylation levels of MVP in PAAD were analyzed by UALCAN (https://ualcan.path.uab.edu/index.html). Additionally, the alternation frequency of MVP across various cancer types was calculated from the cBioPortal (https://www.cbioportal.org/) based on the study of Pan-cancer analysis of whole genomes (ICGC/TCGA).

The CellMiner (https://discover.nci.nih.gov/cellminer/home.do) online database was utilized to investigate the correlation between drug treatment effects and gene expression levels (17). Additionally, with the Enrichr’s Drug Signature Database (https://maayanlab.cloud/Enrichr/), we predicted potential drug candidates targeting MVP and ranked these candidates based on their p-values, with lower p-values indicating higher priority. The structure of these predicted drugs was downloaded from Pubchem (https://pubchem.ncbi.nlm.nih.gov/).

Miapaca-2 and sw1990 cells obtained from ATCC were cultured in DMEM medium (WISENT, 319–005-CL) supplemented with 10% FBS and 1% Pen/Strep in a 5% CO2 incubator at 37°C. These two cell lines were transfected with lipofectamine 2000 (Thermo Fisher Scientifc, Inc.) mixed with siRNA to knock down the expression of MVP, respectively. The antisense sequence of the siRNA (siMVP) is: 5’- CCUACAUGCUGACCCAGGATT-3’.

Total RNA was isolated using RNA isolater Total RNA Extraction Reagent (Vazyme, China) following the manufacturer’s protocol. cDNA was synthesized using the PrimeScript RT reagent Kit with gDNA Eraser kit (Takara, Japan). Quantitative PCR (qPCR) was carried out using ChamQ Universal SYBR qPCR Master Mix (Vazyme, China). The specific primer sequences utilized are provided below.

24 hours after transfection, PAAD cells (3 × 10³ cells per well) were inoculated in 96-well plates and cell viability was estimated with CCK8 (Beyotime, China) at 24-, 48- and 72-hours post-transfection, respectively.

After a 72-hour transfection with siMVP or siNC, miapaca-2 and sw1990 cells were subjected to flow cytometry analysis to assess the impact of MVP knockdown on the cell cycle. The cells were trypsinized, centrifuged at 1200 rpm for 5 minutes, and washed three times with pre-chilled PBS. The cells were fixed with 70% ethanol at -20°C overnight and then the cell cycle distribution was determined using a cell cycle kit (Beyotime, China) and a flow cytometry (BECKMAN COULTER, USA).

Cells were collected for protein extraction with RIPA lysis buffer (Beyotime, China). Western blot was performed as previously reported (18) The following primary antibodies were used: anti-MVP (proteintech, China), anti-GAPDH, anti-Caspase-3, anti-Survivin, anti-AKT, anti-Phospho-Akt, anti-p44/42 MAPK (Erk1/2), anti-Phospho-p44/42 MAPK (Erk1/2) were purchased from Cell Signaling Technology.

Transfected PAAD cells were seeded into 6-well plates at a density of 1000 cells/well for 7 days. The colonies were then fixed with 4% polyformaldehyde for 15 minutes and then stained with crystal violet for 30 minutes. The number of colonies was evaluated by ImageJ software.

PAAD cell lines (Miapaca-2 and sw1990) were seeded into 6-well plates, respectively. When the cell confluency reached almost 80%, the monolayer cells were scratched into a straight line with a sterile micropipette tip. The wound healing within the scrape line was observed at 0- and -24 hours. Corresponding images were captured and subsequently analyzed using ImageJ software.

Transwell assay was conducted to assess cell migration capacity. After transfected with MVP siRNA, PAAD cells were harvested and seeded at a density of 1 × 10⁵ cells/well in medium containing 200 μL of serum-free medium above an 8 μm chamber (14341, LABSELECT, China). 500 μL of medium containing 10% FBS was added to lower chamber. After 24 hours of incubation, cells that had migrated into the lower chamber were fixed with 4% polyformaldehyde for 15 minutes and subsequently stained with crystal violet for 30 minutes. The number of stained cells was quantified by ImageJ software.

The Annexin V-FITC/PI Apoptosis Detection Kit (CX006S) was purchased from Epizyme Biomedical. Adherent cells and suspension cells in culture medium were collected 72 hours after transfection, cells were rinsed with PBS, centrifuged, resuspended in 100 μl binding buffer, and then 5 µl of Annexin V-FITC and 10 µl of PI were added. The mixture was incubated for 15 minutes under dark conditions. Flow cytometry (BD FACSCanto™II system, USA) was utilized for detecting the proportion of apoptotic cells.

Transfected PAAD cells were seeded in 96-well plates at a density of 3 × 10³ cells/well. Following overnight incubation, cells were treated with varying concentrations of Gemcitabine (a chemotherapy agent widely used in the clinical treatment of PAAD) ranging from 0.004 μM to 3 μM in threefold increments for 72 hours. The cell viability was detected as the protocol provided by manufacturer for CellTilter Glo Luminescent Cell Viability Assay (Promega).

All statistical analyses were visualized via GraphPad Prism 9.0 or R software (version 3.1.6 and 4.3.2). The established R packages were utilized as mentioned above. In addition, various statistical tests, including the chi-squared test, Student’s t-test, two-way ANOVA, Wilcoxon test, and Kruskal–Wallis test, were employed to determine the significance of observed differences between distinct groups. Statistical significance is described as follows: ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Genes exhibiting dysregulated expression in tumor samples are considered to be potentially involved in tumor occurrence and progression. Thus, we analyzed the expression of MVP across 15 diverse human cancers from the TCGA database through GEPIA2, and found that the expression of MVP was significantly increased in PAAD compared to normal tissues ( Figures 1A, B ). This aberrant expression pattern was further validated in the GSE15471, GSE62165 and GSE16515 of GEO database ( Figures 1C–F ). Subsequently, to corroborate these observations, we examined the protein levels of MVP in 10 primary PAAD tumor tissues and 10 matched adjacent normal tissues. As expected, immunohistochemistry (IHC) analysis further confirmed that MVP was expressed at higher levels in PAAD than in normal tissues ( Figures 1G, H ).

In order to comprehensively assess the clinical significance of MVP, we evaluated the correlation between distinct clinical parameters (race, tumor grade, pancreatitis status, and TP53 mutation status) of PAAD and the expression levels of MVP using the UALCAN database. The results revealed the notably elevated expression of MVP in Grade 3 tumors compared to other grades, the obviously higher expression in patients with pancreatitis than those without pancreatitis or normal individuals, and the significantly upregulated expression in patients with TP53-mutant PAAD compared to those with the non-mutant type ( Supplementary Figures 1A–D ). Previous research has shown a large proportion of TP53 mutations (64%-83%) in PAAD cases (19), which is consistent with our findings, thus suggesting the potential of MVP as a promising biomarker in clinical diagnosis of PAAD.

To evaluate the prognostic value of MVP in PAAD, we stratified the TCGA-PAAD cohort into MVP high- and low- expression groups, and performed Kaplan-Meier (KM) survival curve analysis of overall survival (OS), which revealed that the elevated expression of MVP was a prognostic indicator associated with a markedly poor prognosis. ( Figure 2A ). Notably, this prognostic value remained consistent across specific clinical parameters, including gender, age and tumor stage ( Figures 2B–F ; Supplementary Figures 1E–H ). In addition, the receiver operating characteristic (ROC) analyses were employed to assess the diagnostic value of MVP, and the area under the curve (AUC) was 0.567, indicating its potential as a diagnostic indicator for PAAD ( Figure 2G ). We further explored the MVP-related OS via multivariate Cox regression analyses, which confirmed the high MVP expression as a significant risk factor for PAAD ( Figure 2H ).

To elucidate the genes correlated with MVP and to further explore the biological functions of MVP in the development of PAAD, we performed functional enrichment analysis of the differential gene expression profiles between MVP high- and low- expression groups. The volcano plot displays the landscape of differentially expressed genes (DEGs) between these groups ( Supplementary Figure 2A ), and the heatmap exhibits the top 10 up- and down-regulated DEGs ( Figure 3A ). After the screening of GO enrichment analysis with those DEGs related to MVP, we focused on the top 30 GO terms containing biological process (BP), cellular component (CC) and molecular function (MF). BP is mainly enriched in “digestion” and “O-glycan processing”; CC is primarily related with “apical part of cell” and “apical plasma membrane”; MF shows a relatively high relationship with “serine-type endopeptidase activity” and “serine-type peptidase activity” ( Figure 3B ; Supplementary Figures 2B–D ). KEGG analysis indicates that “pancreatic secretion”, “fat digestion and absorption”, and “glycosphingolipid biosynthesis” might be involved in the role of MVP in tumor pathogenesis of PAAD ( Figures 3C, D ; Supplementary Figure 2E ). To explore the proteins that potentially interact with MVP, we constructed a protein-protein interaction (PPI) network with STRING database and geneMANIA database, which revealed potential interactions between multiple proteins and MVP, such as PTEN, PARP1, and PARP4 ( Figure 3E ; Supplementary Figure 2F ).

Additionally, we utilized Gene Set Enrichment Analysis (GSEA) to elucidate biological functions and signaling pathways in which MVP may be involved in PAAD. GSEA indicated that the most enriched GO terms were associated with “epidermal cell differentiation” and “epidermis development” ( Figure 3F ), and that the enrichment in KEGG pathways were primarily related to the “chemokine signaling pathway” and “cytokine-cytokine receptor interaction” ( Figure 3G ), suggesting potential functions of MVP in immunomodulation and related signaling events.

As we have known that immune regulation plays a key role in tumor initiation, growth, and treatment, herein, we carried out a multifaceted investigation to assess the relationship between MVP and immunity.

Firstly, the GSEA_KEGG enrichment analysis suggests that MVP is associated with immune regulation ( Figure 3G ). Then we clarified the correlation between the expression of MVP and various kinds of tumor-infiltrating immune cells (TIICs) in PAAD. We downloaded the correlation data between the expression of MVP and 28 kinds of TIICs in PAAD via Timer2.0, and the lollipop plot depicted a predominantly significant correlations between MVP and a cohort of TIICs, particularly prominent for CD56dim natural killer cell, central memory CD4 T cell and Type 17 T helper (Th17) cell ( Figures 4A–D ).

Subsequently, we sought to elucidate the relationship between MVP expression and ICI (immune checkpoint inhibitor) therapy, an emerging paradigm in pancreatic carcinoma treatment (20). Tumor immune dysfunction and exclusion (TIDE), a computational method that mimics the two major mechanisms of tumor immune evasion, has the ability to predict the prognosis of patients with tumors receiving first-line anti-PD1 or anti-CTLA4 therapy (21). It is commonly assumed that elevated TIDE scores and dysfunction scores are correlated with an increased probability of immune evasion, suggesting diminished efficacy of ICI therapy for patients. Contrarily, microsatellite instability (MSI) scores exhibit an inverse relationship. Therefore, we accessed TIDE scores, dysfunction scores and MSI scores of GSE15471 and GSE62542 datasets ( Figures 4E–J ). The results revealed an increased tendency for better benefit from ICI therapy with decreased expression level of MVP, which is consistent with the better prognosis observed in the MVP low expression group shown in Figure 2 .

Immunomodulators and chemokines play momentous roles within the intricate landscape of the tumor microenvironment (TME), promoting tumor development and metastasis (22). To investigate the correlation between MVP expression levels in PAAD and key immune-related factors, we analyzed data from the TISIDB website. Our findings demonstrate a predominantly negative correlation between MVP expression and immunostimulatory factors ( Figures 5A, B ), yet positive correlations with immunoinhibitors ( Figures 5C, D ), immune checkpoints ( Figures 5E, F ), chemokines ( Supplementary Figures 3A, B ), and chemokine receptors ( Supplementary Figures 3C, D ).

To explore the correlation of immune cell distribution with MVP expression levels at the single-cell level, we obtained 7 independent datasets of PAAD (PAAD_CRA001160, PAAD_GSE111672, PAAD_GSE141017, PAAD_GSE148673, PAAD_GSE154778, PAAD_GSE158356, and PAAD_GSE162708) from the scRNA-seq TISCH database ( Figure 6A ). The proportion of malignant cells in PAAD tumor tissues exceeded that in normal tissues, as illustrated in Figure 6B . Moreover, we examined the PAAD_CRA001160 and PAAD_GSE111672 cohorts for deeper investigation. In the PAAD_CRA001160 cohort, a UMAP plot of immune cells within tumor tissues ( Figure 6C ) revealed a widespread distribution of MVP among various immune cell types ( Figure 6D ), with the highest percentage observed in malignant cells ( Figure 6E ). This observation was consistent in the PAAD_GSE111672 cohort, as depicted in Figures 6F–H .

Taken together, these results suggest that MVP may promote tumorigenesis by modulating the immune system, and that PAAD patients with heightened MVP expression may be less responsive to immunotherapeutic response such as ICI therapy.

To comprehensively unravel the aberrant alterations of MVP in PAAD initiation and progression, we conducted an in-depth investigation of the genetic alterations, promoter methylation level, and phosphorylation status of MVP in PAAD. The landscape of genetic alterations of MVP in pan-cancer were detected via the cBioPortal database, which unveiling a prevalence pattern of genetic variation in MVP within the cancer spectrum ( Figure 7A ). Generally, the main type of its genetic alterations was “amplification”, “mutation”, and “deep deletion”. In the case of PAAD, we analyzed 308 PAAD patients with a frequency of MVP gene alterations up to 1.62%, containing a “mutation” rate of 0.65% and an “amplification” rate of 0.97% ( Figure 7B ).

Upon scrutiny of the UALCAN database, we found a marked reduction in the promoter methylation level of MVP in primary tumors compared to normal tissues ( Figure 7C ). To further investigate the methylation pattern of MVP in PAAD, we utilized the MethSurv database and identified 17 methylation sites that have a significant impact on MVP expression ( Figure 7D ). Notably, the methylation levels of most sites were positively correlated with survival, such as cg04633683 and cg09309024, whose KM analysis demonstrated consistency with the MVP related survival analysis in PAAD patients ( Figure 7E ).

Moreover, we explored phosphorylation mutations of MVP in PAAD using the Clinical Proteomic Tumor Analysis Consortium (CPTAC) database. Our research findings indicated that in PAAD, phosphorylation mutation sites of the MVP protein were located at the s807 and s445 sites ( Figure 7F ), with the most significant increase in phosphorylation level observed at the s445 site in PAAD compared to normal tissues ( Figure 7G ). Concurrently, in line with our findings shown in Figure 1 , the abundance of MVP protein was markedly higher in PAAD than in normal tissues ( Figure 7H ).

Collectively, these results propose that the variations in promoter methylation level and protein phosphorylation status of MVP in PAAD may be a contributing factor to its tumorigenic potential.

In this study, we explored the relationship between MVP expression in PAAD and the pharmacological efficacy of both (Food and Drug Administration) FDA-approved and clinically evaluated drugs using the data of RNA-seq and compound activity accessed from CellMiner. By calculating the Pearson correlation coefficient, we identified a significant association between the half-maximal inhibitory concentration (IC50) of a subset of compounds and MVP expression levels, the most notable of which were displayed in Figure 8A . Specifically, our analysis revealed a positive correlation between MVP expression levels and the IC50 values of several drugs, including H-89 (Cor=0.479), GSK-2126458 (Cor=0.454), Deforolimius (Cor=0.449), JNJ-38877605 (Cor=0.443), Staurosporine (Cor=0.433), AZD-8055 (Cor=0.414), XAV-939 (Cor=0.402) and Simvastatin (Cor=0.391). Conversely, we observed a negative correlation between MVP expression and the IC50 values of certain drugs like Oxaliplatin (Cor=-0.465), TAK Plk inhibitor (Cor=-0.453), and Volasertib (Cor=-0.450). These results suggest that MVP may serve as a biomarker for tailoring personalized therapeutic strategies for patients with PAAD, helping to optimize precision treatments for patients with elevated MVP levels.

Next, based on the molecular conformations of candidate compounds and their binding abilities with MVP, we further predicted small-molecular drugs targeting MVP for the treatment of PAAD. Utilizing the Drug Signatures Database (DSigDB) accessed through the Enrichr website, we identified a set of pharmacological compounds with specificity for MVP ( Figure 8B ). The chemical structural features of these compounds were subsequently obtained from PubChem, and they were prioritized according to their statistical significance (P-values) as illustrated in Table 1 . Additionally, with AutoDock software, we forecasted the binding affinities and conformation of MVP with Suloctidil ( Figure 8C ) and Tetradioxin ( Figure 8D ), depicting their potential as MVP-targeted therapeutic drugs through the quantification of binding free energy and elucidation of interaction modalities.

Based on the multifaceted findings mentioned above, we have elucidated the oncogenic potential of MVP in driving tumorigenesis of PAAD and its feasibility as a biomarker for clinical diagnosis and prognosis of PAAD. This oncogene leads to a range of adverse prognostic effects by promoting cell proliferation and triggering immune dysfunction. To this end, we conducted a series of experiments to validate the potential inhibitory effects on malignant progression of tumors through targeted suppression of MVP at the cellular level.

At first, we verified the significant knockdown efficiency of siMVP in two PAAD cell lines (miapaca-2 and sw1990) at the transcriptional level and protein level via qRT-PCR and Western blot techniques, respectively ( Figures 9A, B ).

Next, Cell Counting Kit-8 (CCK-8) assay ( Figure 9C ) and colony formation assay ( Figure 9D ) indicated that knockdown of MVP inhibited the cell viability, showing significantly decreased proliferation of PAAD cell lines compared to negative controls (si-NC). The cell cycle was performed by flow cytometry analysis, which indicated a significant increase in the number of cells in the G0/G1 phase and a marked decrease in the number of cells in the S phase in the MVP knockdown (si-MVP) groups compared to the negative control (si-NC) groups, suggesting that blocking the expression of MVP in PAAD cell lines would partially hinder their proliferative capacity ( Figures 9E, F ).

Additionally, flow cytometry analysis detected elevated levels of apoptosis in PAAD cell lines following MVP downregulation ( Figure 9G ), with Figure 9H providing the quantified rates of total apoptosis for these cell lines.

The migration ability of tumor cells is a fundamental element in the process of tumor metastasis. To explore the effects of MVP on the migration ability of PAAD cells, we performed wound healing assays and found that PAAD cells in the absence of MVP exhibited a severe decline in the migratory capacity ( Figures 10A, B ). This finding was corroborated by transwell assays, which also demonstrated a significant impairment in migrative ability of PAAD cells upon MVP knockdown ( Figures 10C, D ).

Previous studies have reported that MVP suppresses the progression of papillary thyroid cancer by interfering with the ERK and AKT pathways. Hence, we wondered whether MVP exerts a similar inhibitory effect on these pathways in PAAD. In this regard, through Western blot analysis, we observed a decrease in the expression levels of phosphorylated ERK (p-ERK) and phosphorylated AKT (p-AKT) proteins after MVP knockdown ( Figure 10E ).

Moreover, we performed the cell viability assay to evaluate the chemo-sensitivity of PAAD cells following MVP knockdown. The results indicated that the IC50 values were comparable to those of the negative controls (si-NC), suggesting no significant change in chemo-resistance to Gemcitabine, the standard chemotherapeutic agent ( Figure 10F ).

In summary, knockdown of MVP in PAAD cell lines impairs cell proliferation and migration abilities, accelerates cell apoptosis, and thereby inhibits the tumor progression, which strongly supports the results of our bioinformatics analyses.