The TCGA and the Genotype-Tissue Expression (GTEx) pan-cancer datasets, which include normalized expression profiles, clinical information, somatic mutation data, and survival data, were downloaded from UCSC (https://xenabrowser.net/datapages/), cBioPortal (https://www.cbioportal.org/), and the R package “TCGAbiolinks”. Protein levels across various cancer types were assessed using cProSite (https://cprosite.ccr.cancer.gov/). The relationships between ADGRG6 and clinical characteristics, as well as survival times, were analyzed using data from GEO (https://www.ncbi.nlm.nih.gov/geo/), which includes datasets GSE71729, GSE15471, GSE183795, GSE62452, GSE28735, GSE57495, GSE21501, and GSE85916. PAAD cohort data related to ICGC-AU was retrieved from International Cancer Genome Consortium (ICGC, https://dcc.icgc.org/). Kaplan-Meier survival curves were generated using the Kaplan-Meier Plotter (KM-plot, https://kmplot.com/analysis/index.php?p=service) to evaluate the potential of ADGRG6 as a survival marker for pancreatic cancer. Additionally, PAAD single-cell datasets were downloaded from GEO (https://www.ncbi.nlm.nih.gov/geo/).

For the TCGA pan-cancer cohort, Cox regression analysis on the four survival indicators—overall survival (OS), progression-free interval (PFI), disease-specific survival (DSS), and disease-free interval (DFI)—was conducted using the R package “survival”. The R packages “survival” and “survminer” were employed to determine the optimal cutoff values and generate Kaplan-Meier curves to assess the prognostic value of ADGRG6.

Gene functional enrichment analysis was performed using collections downloaded from the Molecular Signatures Database (MSigDB, https://www.gsea-msigdb.org/gsea/msigdb) and the R package “clusterProfiler”. Gene set scoring was conducted with the R package “GSVA”. To explore the relationship between ADGRG6 expression and tumor immune infiltration, we utilized the “cibersort” algorithm (via the R package “CIBERSORT”), “ssGSEA” (via the R package “GSVA”), and “xCell” (via the R package “xCell”). Additionally, the Tracking Tumor Immunophenotype (TIP, http://biocc.hrbmu.edu.cn/TIP/) tool was employed to predict potential correlations between ADGRG6 expression and the tumor immune cycle.

Single-cell sequencing data were downloaded from GEO and processed using the Seurat R package. For quality control, cells expressing fewer than 200 genes or more than 20% mitochondrial reads were removed. The “doubletFinder” function from the R package “doubletFinder” was employed to identify and exclude potential doublets in each sample. The filtered data were then scaled and normalized prior to advanced analysis. Batch effects between samples were corrected using the R package “Harmony”. Cell clustering and dimensionality reduction were performed with the “FindClusters” and “RunUMAP” functions. To annotate cell types, “FindAllMarkers” was used to identify specific genes within each cell cluster. Copy number variation (CNV) was assessed with the R package “CopyKAT” to differentiate malignant cells from normal cells. Cell-cell interactions among different cell types were evaluated using CellChat with the default parameters of the database, focusing exclusively on receptor-ligand interactions as annotated in the database.

Anti-ADGRG6 (Cat. #TA315653, OriGene), anti-p53 (Cat. #3H2821, Santa Cruz Biotechnology), β-actin (Cat. #A2066, Sigma-Aldrich), anti-N-cadherin (Cat. #393933, Santa Cruz Biotechnology), anti-PCNA (Cat. #2586, Cell Signaling Technology), anti-EGFR (Cat. #4267, Cell Signaling Technology), anti-p-EGFR (Cat. #2234, Cell Signaling Technology), anti-p65 (Cat. #8242, Cell Signaling Technology), anti-AMPK (Cat. #2532, Cell Signaling Technology), anti-p-AMPK (Cat. #2531, Cell Signaling Technology) antibodies were used. HRP-conjugated affinipure donkey anti-rabbit IgG(H+L) (Cat.SA00001-9) and goat anti-mouse IgG(H+L) (Cat. SA00001-1) antibodies were purchased from Proteintech Group (Wuhan, China).

Transwell chamber apparatus (Cat. #3422) and Matrigel (Cat. #354234) were purchased from Corning (New York, NY, USA). MTT powder were purchased from Sigma (Louis, MO, USA).

Human pancreatic cancer cell lines (AsPC-1, CFPAC-1, MIAPaca-2, BxPC-3, and PANC-1) and HEK293T cells were purchased from the Cell Bank of the Chinese Academy of Sciences, Type Culture Collection, Shanghai, China. Human normal pancreatic ductal cells (hTERT-HPNE) were acquired from the American Type Culture Collection (ATCC, Manassas, VA, USA). All cell lines were cultured in media supplemented with 10% fetal bovine serum (FBS, Sigma) at 37 °C in a 5% CO₂ incubator. AsPC-1 cells were grown in RPMI 1640 medium (Gibco), CFPAC-1 cells in Iscove’s Modified Dulbecco’s Medium (IMDM, Gibco), and HEK293T, BxPC-3, PANC-1, and MIAPaca-2 cells in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco). The complete medium consisted of base culture media, 10% fetal bovine serum, 100 U·mL^-1 penicillin, and 100 μg·mL^-1 streptomycin (Life Technologies, Carlsbad, CA, USA).

ShRNA oligos targeting ADGRG6 were annealed and subcloned into the pLKO.1-puro plasmid. For overexpression, the ADGRG6 or mutated p53 gene was subcloned into the pLVX-3Flag-Puro plasmid. HEK293T cells were transfected with PEI40000 and packaging plasmids (pMDLg-pRRE, pRSV-REV, and pCMV-VSV-G) to produce lentivirus. After 48 hours, the lentiviral supernatant was collected, centrifuged at 1600 rpm for 10 minutes, and filtered through 0.45 μm filters (Millipore, Billerica, MA, USA). The filtered lentiviral supernatant was added to the culture medium of the human pancreatic cancer cell lines. Following a 48-hour infection period, puromycin (2 μg·mL^-¹) was used to select for stable knockdown/overexpression cell lines.

The shRNA sequences used were as follows:

Total RNA was extracted from cells using Trizol (Takara, Dalian, China) and reverse-transcribed into cDNA using a Synthesis SuperMix kit (Yeasen, Shanghai, China). Real-time PCR was conducted using the qPCR SYBR Green Master Mix kit (Yeasen, Shanghai, China) according to the manufacturer’s instructions. The ACTB gene served as an internal control.

The primers used were as follows:

Cells were lysed using ELB lysis buffer (150 mM NaCl, 100 mM NaF, 25 mM Tris-HCl [pH 7.6], 1% Nonidet P-40 [NP-40], 1 mM PMSF, and protease inhibitors) for 10 minutes on ice. The lysate was collected with a scraper, subjected to ultrasonic treatment, and then centrifuged at 13,000 rpm for 15 minutes at 4°C. The supernatant was collected and protein concentration was determined. Proteins were separated by SDS-PAGE, transferred to a PVDF membrane, and blocked with 5% skim milk. The membrane was incubated overnight at 4°C with specific primary antibodies. The following day, it was incubated with appropriate fluorescent secondary antibodies. Immunoreactive bands were detected using an Enhanced Chemiluminescence (ECL) system (Bio-Rad, Hercules, CA, USA).

The Transwell chamber apparatus was used for the assay. Before the invasion experiment, a 10% matrix gel was prepared and 50 μL of this gel was added to the upper chamber, allowing the gel to solidify. For both the migration and the invasion assays, a specified number of cells were seeded in the upper chamber (cells plated in each Transwell chamber: CFPAC-1: 6×10⁴, Panc-1: 5×10⁴, MIAPaca-2: 10×10⁴). After 24 hours, non-invading cells in the upper chamber were removed with cotton swabs. The invading cells on the lower side of the membrane were fixed with 4% paraformaldehyde for 5-10 minutes, dried, and stained with 0.1% crystal violet for 5-10 minutes. Three fields of view from each well were captured for statistical analysis.

In the MTT assay, cells were seeded in a 96-well plate. After 12 hours, 20 μL of MTT solution was added to each well, and the plate was incubated for an additional 4 hours. Following incubation, the medium was replaced with 200 μL of DMSO, and the plate was shaken for 15 minutes to dissolve the formazan crystals produced by MTT. The absorbance was measured at 570 nm using a microplate reader. This procedure was repeated for 5 consecutive days (initial number of cells plated: CFPAC-1: 3000, Panc-1: 3000, MIAPaca-2: 2500).

In the colony formation assay, a small number of cells were seeded in a six-well plate. Once colonies formed (approximately 14 days), the culture medium was discarded, and the cells were fixed with 4% paraformaldehyde. The colonies were then stained with 0.1% crystal violet. Three fields of view from each well were captured for statistical analysis (initial number of cells plated: CFPAC-1: 1000, Panc-1: 1000, MIAPaca-2: 500).

CFPAC-1 cells (1.5 × 10⁶ per mouse) were suspended in 200 μL of base DMEM medium (without serum) and injected subcutaneously into the right flank of each nude mouse. Seven days later, when tumors became visible, measurements were taken daily, and tumor volume (Tv) was calculated using the formula: Tv = L (length) × W² (width)/2. At the end of the observation period, mice were euthanized, weighed, and tumor samples were collected for Western blot analysis.

Animal experiments were conducted according to protocols approved by the Animal Care and Use Committee of Xiamen University. Female nude mice (BALB/c, 15–20 g, 5–6 weeks old) were obtained from the SLAC Laboratory Animal Center (Shanghai, China). The mice were housed and handled under specific pathogen-free conditions at the Laboratory Animal Center of Xiamen University (Xiamen, China).

For bioinformatics analysis, statistical computations were performed using R software (version). Specific tests such as the Wilcoxon rank-sum test and Log-rank test were applied. In biological experiments, data were presented as means ± SEM. Statistical comparisons between groups were made using Student’s t-test or one-way ANOVA, with significance determined by GraphPad Prism 8 (GraphPad Software) and Image J. A p-value of < 0.05 was considered statistically significant.

To investigate the expression and potential tumor-regulatory role of ADGRG6, we analyzed expression data from the TCGA Pan-Cancer cohort and the GTEx database. The results revealed that ADGRG6 mRNA levels were elevated in several cancers when compared to the corresponding normal tissues, including CESC (cervical squamous cell carcinoma and endocervical adenocarcinoma), ESCA (Esophageal Cancer), GBM (Glioblastoma), KIRC (kidney renal clear cell carcinoma), KIRP (kidney renal papillary cell carcinoma), LGG (brain Lower Grade Glioma), OV (ovarian serous cystadenocarcinoma), PAAD (pancreatic adenocarcinoma), STAD (stomach adenocarcinoma), TGCT (testicular germ cell tumors), THYM (thymoma), and UCS (uterine carcinosarcoma) ( Figure 1A ). Additionally, CPTAC data showed that ADGRG6 protein expression was higher in tumors compared to normal tissues in KC (kidney cancer) and PDAC (pancreatic ductal adenocarcinomas) ( Figure 1B ).

Survival time is a crucial indicator for assessing tumor prognosis. To evaluate the prognostic value of ADGRG6, we performed univariate Cox analysis to examine its relationship with survival time across four metrics: overall survival (OS), progression-free interval (PFI), disease-specific survival (DSS), and disease-free interval (DFI). ADGRG6 was consistently associated with poor prognosis in PAAD, rather than other cancers, across all four evaluation methods ( Figures 1C–F , Supplementary Table S1–S4 ). Kaplan-Meier curves further confirmed that ADGRG6 expression correlated with worse prognosis in the TCGA PAAD cohort ( Figures 1G–J ). Therefore, these results suggested that ADGRG6 could potentially serve as a relatively specific biomarker for PAAD.

To further substantiate our findings regarding the potential oncogenic role of ADGRG6 based on the TCGA PAAD cohort, several independent PAAD cohorts from the GEO database were then utilized to validate the observation. Analysis of the GSE71729, GSE15471, GSE183795, GSE62452, and GSE28735 datasets revealed that ADGRG6 expression was significantly higher in tumor samples compared to normal pancreatic tissues ( Figures 2A–E ).

Further examination of the relationship between ADGRG6 mRNA expression and clinical characteristics showed that, in the TCGA PAAD cohort, ADGRG6 was more highly expressed in younger patients and those with higher T-stage and grade ( Figures 2F–H ), although no significant association was observed with N-stage or chronic pancreatitis status ( Supplementary Figure S1A, B ). In the GSE71729 cohort from the GEO database, ADGRG6 expression was notably higher in metastatic PAAD compared to primary tumors ( Figure 2I ). Tumor molecular typing, crucial for guiding molecular-targeted therapies, classified PAAD into “classical” and “basal-like” subtypes, with the latter generally associated with poorer prognosis (13). Despite this, no significant difference in ADGRG6 expression was found between these two subtypes ( Figure 2J ). Additionally, analysis based on stromal molecular signatures revealed that ADGRG6 expression was higher in the “activated” subtype, which is associated with macrophages, inflammation, and an activated fibroblast state, compared to the “normal” subtype ( Figure 2K ). Kaplan-Meier survival analysis across all validation cohorts (GSE71729, GSE183795, GSE57495, GSE21501, GSE85916, ICGC-AU, and KM-plot) demonstrated that high ADGRG6 expression was correlated with shorter survival ( Figures 2L–R ). These results collectively suggested that ADGRG6 held significant potential as a prognostic marker for PAAD.

Gene mutation is one of the main characteristics of tumors. We analyzed the relationship between ADGRG6 and common mutations in PAAD. The ADGRG6-high group notably exhibited a higher proportion of KRAS mutations (a common mutation in PAAD) ( Figure 3A ), suggesting a potential association between ADGRG6 and increased malignancy. Furthermore, ADGRG6 expression was positively correlated with tumor mutational burden (TMB) and microsatellite instability (MSI) across PAAD ( Figure 3B , Supplementary Figure S2A, B ). Lollipop charts illustrating ADGRG6 protein mutations in the TCGA pan-cancer cohort indicated that missense mutations were the predominant mutation type ( Figure 3C ). To further elucidate the potential role of ADGRG6 in PAAD, we performed gene set enrichment analysis (GSEA) by dividing the TCGA PAAD cohort into ADGRG6-high and -low groups based on median expression levels. ADGRG6 was associated with several extracellular matrix and migration-related pathways, including “cell-cell junction organization,” “actin filament organization,” “focal adhesion,” and “adherens junction” ( Figures 3D, E ). Gene set variation analysis (GSVA) scores indicated that the ADGRG6-high group had elevated scores for the p53 signaling pathway, WNT signaling pathway, MAPK signaling pathway, TGF-β signaling pathway, PI3K-AKT signaling pathway, and G2/M checkpoint pathway compared to the ADGRG6-low group ( Figures 3F, G ), suggesting a strong association with cancer growth and metastasis.

Given the immunosuppressive nature of PAAD’s tumor microenvironment (27–29), we performed immune infiltration analysis using three independent PAAD datasets (TCGA, GSE71729, and GSE183795). CIBERSORT analysis showed that high ADGRG6 expression was associated with lower levels of naive B cells in all three groups ( Supplementary Figure S3A–C ). Although the role of naive B cells in the pancreatic cancer microenvironment has not been fully elucidated, some reports have indicated that high infiltration of B cells is associated with anti-tumor immunity and better prognosis (30–32). CD8⁺ T cells are essential for tumor immunity, as they can recognize and kill tumor cells (33). CIBERSORT analysis revealed a trend towards lower CD8+ T cell infiltration in the ADGRG6-high group across the three datasets, although this did not reach statistical significance ( Supplementary Figure S3A–C ). The xCell algorithm further demonstrated a significant decrease in CD8+ T cell infiltration in the ADGRG6-high group across all the three datasets ( Figures 4A–C ). Moreover, ssGSEA analysis showed that the ADGRG6-high expression group had a lower activated CD8⁺ T cell score but a higher Th17 score ( Figures 4D–F ). High Th17 infiltration may be related to tumor growth, pancreatic intraepithelial neoplasia (PanIN) formation, stem cell properties, and migration (34–36). Further exploration of the relationship between ADGRG6 expression and various stages of the tumor immune cycle revealed distinct immune activity levels between the two groups, particularly during the priming and activation stages. Notably, the high ADGRG6 expression group exhibited reduced T cell and APC activation ( Figure 4G ). Cancer cells can exploit immune checkpoints to evade immune responses, and the ADGRG6-high group displayed increased levels of CD274/PDL1 and SIGLEC15 expression ( Figure 4H ). These findings suggested that ADGRG6 might facilitate PAAD development by regulating immune infiltration, ultimately leading to immune escape.

We obtained single-cell sequencing data from PAAD and corresponding normal pancreatic tissue samples from 11 patients, sourced from the GEO database ( Figure 5A ). After rigorous quality control, dimensionality reduction, and clustering, we identified 13 distinct cell subsets ( Figure 5B ). Further cell annotation led to the classification of 12 cell types, including T cells, ductal cells, macrophages, fibroblasts, natural killer (NK) cells, acinar cells, monocytes, B cells, endothelial cells, mast cells, plasma cells, and endocrine cells, which were visualized using UMAP ( Figure 5C ). The marker genes of each cell type following single-cell clustering, displayed in Figure 5D , served as the foundation for annotating the cell types. Additionally, we examined the cellular composition across samples. As indicated in Figure 5E and Supplementary Figure S4 , GSM5910786, derived from normal pancreatic tissue, was predominantly composed of acinar and ductal cells; in contrast, PAAD tissues were rich in diverse immune cells and fibroblasts, highlighting the complexity of their tumor microenvironment.

For the gene ADGRG6, its expression was most pronounced in ductal cells ( Figure 5F ). A comparative analysis of normal and tumor tissues revealed significantly elevated ADGRG6 expression in the ductal cells of tumor samples compared to those in normal tissue ( Figures 5G, H ). Employing the COPYKAT algorithm, we detected aneuploidy in ductal cells within tumor samples, which we classified as malignant cells ( Figure 5I ). Intriguingly, ADGRG6 expression was significantly higher in these malignant cells compared to non-malignant cells within tumor samples ( Figure 5J ).

We designated malignant cells expressing ADGRG6 as ADGRG6⁺ malignant cells and those without expression as ADGRG6^- malignant cells. The ADGRG6⁺ group exhibited higher scores in the P53 pathway, G2/M checkpoint, TGF-β signaling, PI3K-AKT signaling, and oxidative phosphorylation pathways ( Figure 6A ), corroborating our previous analysis of the TCGA PAAD dataset ( Figures 4C, D ). Moreover, differential gene enrichment analysis between the two groups indicated that ADGRG6 was associated with pathways such as Toll-like receptor signaling, mineral absorption, necroptosis, ferroptosis, and ECM-receptor interaction ( Figure 6B ).

Utilizing the “CellChat” platform, we visualized ligand-receptor (L-R) mediated cell-cell interactions, revealing a higher frequency of interactions between the ADGRG6⁺ group and other cells, notably fibroblasts and macrophages ( Figures 6C, D ). We conducted a detailed analysis of the L-R interaction strength between ADGRG6^- and ADGRG6⁺ malignant cells and their counterparts among T cells, macrophages, and fibroblasts, both in terms of input and output signals ( Figures 6E, F ). These interactions were depicted in bubble plots for clarity. Secreted phosphoprotein 1 (SPP1), an integrin-binding protein, is released by various cell types, including macrophages, endothelial cells, and osteoclasts (37). We observed that the interaction strength of the SPP1-(ITGAV+ITGB5/ITGB6) axis between macrophages and ADGRG6⁺ malignant cells surpassed that of the ADGRG6^- group; likewise, the frequency of LGALS9-CD44/CD45 interactions was higher in the ADGRG6⁺ group ( Figures 6E, F ). LGALS9 is known to enhance the function and stability of induced T regulatory cells (iTregs), promote the expression of Foxp3 in iTregs, and suppress the activation of CD8⁺ T cells, thereby contributing to an immunosuppressive state (38). Further analysis of the input and output signaling patterns among different cell types revealed significant differences in the transmission of SEMA3, GDF, WNT, GAS, TRAIL, PARs, IL1, and other signals between the two groups ( Figure 6G ). To highlight the differential L-R interactions, we presented the expression levels of these genes in each cell type. Consistent with our findings, the genes showed higher expression levels in the ADGRG6⁺ group as compared to the ADGRG6^- group ( Figure 6H ).

To investigate the role of ADGRG6 in PAAD, we first analyzed its mRNA and protein levels in several PAAD cell lines. Compared to the human pancreatic duct epithelial cell line hTERT-HPNE, both the mRNA and protein expression levels of ADGRG6 were significantly upregulated in all the five PAAD cell lines, including AsPC-1, BxPC-3, MIAPaca-2, CFPAC-1, and PANC-1 ( Figures 7A, B ). We then knocked down ADGRG6 in ADGRG6-high CFPAC-1 and PANC-1 cells, while overexpressed it in ADGRG6-low MIAPaca-2 cells ( Figures 7C, F ). MTT and colony formation assays demonstrated that ADGRG6 knockdown reduced the proliferation of CFPAC-1 and PANC-1 cells ( Figures 7G–J ). Conversely, ADGRG6 overexpression enhanced the growth of MIAPaca-2 cells ( Figures 7K–L ). We also investigated the impact of ADGRG6 on the metastatic potential of PAAD cells. Transwell assays were conducted to assess the migration and invasion capabilities of PAAD cells following ADGRG6 knockdown and overexpression. The results showed that ADGRG6 knockdown significantly reduced the migration and invasion abilities of CFPAC-1 and PANC-1 cells ( Figures 7M, N ), while ADGRG6 overexpression enhanced those of MIAPaca-2 cells ( Figure 7O ). These results indicated that ADGRG6 promotes the growth, migration and invasion of PAAD cells.

To assess the impact of ADGRG6 knockdown on the growth of subcutaneously transplanted pancreatic tumors in a mouse model, we injected BALB/c nude mice with either control or ADGRG6 knockdown CFPAC-1 cells (1 × 10⁶ cells per mouse). We began monitoring tumor size from day 7 post-injection, and the tumors were harvested at the end of observation. Our results demonstrated a noticeable reduction in both the volume and weight of the subcutaneous tumors in the mice that received ADGRG6 knockdown cells as compared to the control group ( Figures 8A-C ).

Together, these findings provide evidence supporting the role of ADGRG6 in promoting PAAD progression.

We then investigated how ADGRG6 regulated the progression of PAAD. Utilizing real-time PCR, we observed a reduction in the expression of EMT-related marker genes following ADGRG6 knockdown in CFPAC-1 and PANC-1 cells, with a converse effect observed upon its overexpression in MIAPaca-2 cells, most notably for CDH2 (N-Cadherin) as depicted in Figures 9A–C . Similarly, the knockdown of ADGRG6 in CFPAC-1 and PANC-1 cells led to a decrease in the expression of various immune-related genes, including PDL1 (CD274), CD44, ITGAV, ITGB5, ITGB6, SIGLEC15, and LGASL9, while overexpression of ADGRG6 increased the expression of these genes in MIAPaca-2 cells ( Figures 9D–F ).

It is well established that wild-type p53 functions as a tumor suppressor, whereas most of the mutated p53 exhibit gain-of-function (GOF) activities, such as promoting cell proliferation, metastasis and evading immune surveillance (39–41). Notably, the p53 gene mutation frequency in PAAD exceeds 70% (9, 10). Using the CCLE database, we identified specific mutations of p53 in the three cell lines—CFPAC-1: C242R, PANC-1: R273H, MIAPaCa-2: R248W—that were associated with gain-of-function activities (42), as presented in Supplementary Figure S5A . Previous research has shown that p53 knockdown can reduce the migration and invasion capabilities of CFPAC-1 and PANC-1 cells (43). Interestingly, our western blot analysis demonstrated that knockdown of ADGRG6 resulted in a decrease in p53 protein levels in CFPAC-1 cells, while ADGRG6 overexpression had the opposite effect in MIAPaCa-2 ( Figures 9G, H ). Similarly, in vivo experiments showed similar results, the knockdown effect of ADGRG6 and the consequent decrease in p53 protein levels in the transplanted tumors were further validated by western blot analysis ( Figure 9I , Supplementary Figure S5B ). Furthermore, in these cell lines, we uncovered that ADGRG6 activated EGFR, AMPK and NF-κB signaling pathways ( Figure 9G, H ), which are believed to be modulated by mutated p53 protein to exert its GOF and work as the upstream molecules for the regulation of the proliferation-, EMT- and immune-related genes mentioned above (39, 44–47).

To further explore the potential role of the ADGRG6-mutated p53 signaling axis in pancreatic cancer, by constructing a CFPAC-1 cell model in which mutated p53 (C242R) was reintroduced following ADGRG6 knockdown, we observed that the inhibition of proliferation, migration, and invasion caused by ADGRG6 knockdown was substantially reversed by the overexpression of exogenous p53(C242R) ( Figures 10A–C ). Conversely, when we overexpressed ADGRG6 in MIAPaca-2 cells with endogenous p53 (mutp53) knockdown beforehand, the regulatory effects of ADGRG6 overexpression on proliferation, migration, invasion and the expression of the proliferation-, EMT- and immune-related markers of PAAD cells were greatly impaired when endogenous p53 was knocked down, suggesting that ADGRG6 mediated these effects through mutated p53 in PAAD cells ( Figures 10D–H ). These results demonstrated the pivotal role of the ADGRG6-mutated p53 signaling axis in facilitating the progression of PAAD.

Interestingly, we also found that ADGRG6 mRNA expression was elevated in samples harboring mutated p53 compared to those with wild-type p53 as well as its mRNA levels showed a positive correlation with p53 in the TCGA PAAD cohort ( Figure 10I , Supplementary Figure S6 ). Furthermore, we found that ADGRG6 expression was significantly downregulated after knockdown of p53 in MIAPaca2 cells ( Figure 10J ), indicating a positively feedback loop of ADGRG6 and mutp53.