Skip to main content
NCBI home page
BMC Gastroenterology logoLink to BMC Gastroenterology
. 2026 May 13;26:416. doi: 10.1186/s12876-026-04913-3

ADGRF4 and ADGRL4 as novel prognostic biomarkers and potential therapeutic implications in stomach adenocarcinoma

Jia-Hui Wang 1, Ming-Jun Li 2, Dan-Dan Zang 3, Xiao-Wei Wang 2, Peng Huang 2,✉
PMCID: PMC13339262  PMID: 42129668

Abstract

Background

This study explored the differential expression of adhesion G protein-coupled receptor (ADGR) genes in stomach adenocarcinoma (STAD), evaluated their association with STAD prognosis, and explored their potential therapeutic implications.

Methods

The gene expression profiles of STAD and normal gastric tissues were obtained from TCGA and GTEx databases and analyzed using the DEseq2 R package. External validation was performed using three GEO datasets. Co-expression patterns of ADGRF4 and ADGRL4 with immune-related molecules were analyzed using Spearman correlation. The prognostic values of ADGRs were evaluated using multivariate Cox risk models and Kaplan–Meier survival analyses. The tumor microenvironment composition and immune infiltration were analyzed using the ESTIMATE R package and CIBERSORT algorithm. GO and KEGG enrichment analyses were performed using DAVID. Immunohistochemistry was conducted to validate gene expression in 50 pairs of STAD and adjacent normal tissues. Drug sensitivity was analyzed using the oncoPredict R package.

Results

ADGRF4 and ADGRL4 were significantly upregulated in STAD tissues and independently associated with poor prognosis and the tumor microenvironment composition. Functional enrichment analysis revealed that ADGRF4 was primarily associated with epithelial development and neuroactive ligand–receptor interaction, whereas ADGRL4 was linked to extracellular matrix organization and calcium signaling pathways. Immunohistochemistry confirmed the significantly higher expression of both proteins in gastric cancer tissues than in normal tissues. External validation using GEO datasets confirmed the upregulation and prognostic significance of ADGRL4. Drug sensitivity analysis indicated that ADGRL4 expression was significantly correlated with sensitivity to the Chk1/Chk2 inhibitor AZD7762.

Conclusion

ADGRF4 and ADGRL4 were significantly overexpressed in STAD and independently associated with poor prognosis. Expression of these genes was correlated with changes in tumor microenvironment and immune cell infiltration, indicating their potential association in STAD progression. These findings suggest that ADGRF4 and ADGRL4 could serve as novel prognostic biomarkers with potential therapeutic significance in gastric cancer.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-026-04913-3.

Keywords: Stomach adenocarcinoma, Adhesion G protein-coupled receptors, ADGRF4, ADGRL4, Tumor microenvironment

Background

Stomach adenocarcinoma (STAD), accounting for more than 1 million new cases and approximately 769,000 deaths annually, is one of the most common malignant tumors globally [1]. Despite advances in diagnosis and therapeutic approaches, the 5-year survival rate for patients with STAD remains lower than 30%, largely because of its delayed diagnosis, high recurrence rates, and treatment resistance [2]. Thus, novel biomarkers for early diagnosis and potential therapeutic implications for STAD are urgently needed.

G protein-coupled receptors (GPCRs) comprise the largest family of membrane proteins in humans, with over 800 members that participate in various cellular activities [3]. According to a recent study, approximately 34% of FDA-approved drugs target GPCRs, highlighting the promising therapeutic utility of these proteins [4]. The GPCR superfamily consists of five major families: rhodopsin, adhesion, glutamate, secretin, and frizzled [5]. Among these families, adhesion GPCRs (ADGRs) are characterized by their long N-terminal extracellular domains containing multiple adhesion domains, which enable them to participate in cellular activities involving cell–cell and cell–extracellular matrix interactions, such as tumorigenesis [6]. The ADGR family consists of 33 members, which are classified into nine subfamilies (ADGRA, ADGRB, ADGRC, ADGRD, ADGRE, ADGRF, ADGRG, ADGRL, and ADGRV) based on their phylogenetic relationships and structural characteristics [7]. However, apart from their structural characterization, the functional importances of the majority of ADGRs remain largely unexplored.

Emerging evidence suggests that ADGRs play important roles in cancer progression [8–10]. Several ADGRs have been implicated in tumorigenesis across various cancer types. For instance, ADGRD1 is correlated with overall survival in glioblastoma, and tumor cell proliferation was inhibited upon ADGRD1 silencing [8]. Moreover, ADGRE5 promotes tumor cell proliferation and invasion [11]. Additionally, ADGRG1 was detected in breast, pancreas, colorectal, and lung cancers [9]. Furthermore, ADGRL4 has been demonstrated to promote tumor metastasis and angiogenesis [12, 13]. Although different ADGR members exhibited cancer-promoting roles in various types of malignancies, the roles of ADGRF4 and ADGRL4 in STAD have not been evaluated. An increasing body of research has highlighted the clinical significance of ADGR family members in gastric cancer and other malignancies in gastrointestinal tract. For instance, pan-cancer analyses have identified gene signatures associated with immune cell infiltration, immune checkpoint expression, and poor survival in STAD [14]. In addition, GPCR-mediated Rho-GTPase signaling and ammonia-related metabolic reprogramming within the tumor microenvironment (TME) have been recognized as mechanisms through which stromal and immune cells establish immunosuppression and promote tumor development [15, 16]. Moreover, cancer-associated fibroblast heterogeneity has been increasingly characterized across multiple tumor types and plays important roles in shaping the immunosuppressive microenvironment in malignancies, including gastric cancer [17, 18]. These previous investigations suggest that ADGR family proteins are closely associated with tumor immune microenvironment regulation, patient prognosis, and GPCR-mediated oncogenic signaling.

It is well documented that the tumor microenvironment (TME) plays important roles in tumor initiation, progression, and treatment resistance [19]. Meanwhile, immune cell infiltration in the TME significantly affects the treatment outcomes of patients with cancer [20]. Accumulating evidence suggests that ADGRs play significant roles in the modulation of the TME and immune cell infiltration. For instance, ADGRE1 was found to promote M2 macrophage polarization in colon cancer [21].

In this study, we used various bioinformatic tools to investigate the expression of ADGRF4 and ADGRL4 in STAD and explore their associations with STAD prognosis. We also evaluated their potential roles in modulating the TME and immune cell infiltration. In addition, we examined their correlations with drug sensitivity to explore potential translational implications.

Methods

Data analysis

The data were queried and analyzed following the process presented in Fig. 1.

Fig. 1.

Fig. 1

Open in a new tab

Study schematic

Sample collection

Tumor tissue and adjacent normal gastric tissue specimens were collected from 50 patients with STAD who underwent surgical treatment at the General Surgery Department of Panjin Central Hospital between May 2023 and October 2024. The samples collected were fixed in formalin and embedded in paraffin for further investigations. This research was approved by the Ethics Committee of Panjin Central Hospital (Approval No.: 20250120LWLL-KS-02) and was performed in compliance with the Declaration of Helsinki. Written informed consent was obtained from all participants.

The inclusion criteria for patient selection were as follows: pathologically confirmed primary STAD with no other concurrent malignant tumors; treatment-naïve status (no chemotherapy, radiotherapy, targeted therapy, or immunotherapy) before surgery; and the availability of complete clinical records. Patients were excluded for the following reasons: younger than 18 years; prior receipt of anti-cancer treatments (chemotherapy, radiotherapy, targeted therapy, or immunotherapy) before surgery; development of severe perioperative complications or mortality; and incomplete medical records.

Bioinformatic analysis

RNA-seq data acquisition and processing

The gene expression data (raw counts) of 446 STAD and 503 normal gastric tissue samples were downloaded from The Cancer Genome Atlas (TCGA) database and Genotype-Tissue Expression (GTEx) database, respectively. Data were analyzed using R (version 4.4). The reference genome was hg19 (GRCh37), and quality control was performed using the “DEseq2” R package (version 1.44.0) to remove low-quality genes (raw counts < 10 in more than 80% of samples). After that, the gene expression profiles obtained from the two databases were integrated, and the transcriptomic data were extracted. Batch effects between TCGA and GTEx datasets were removed using the ComBat_Seq function from the SVA package (version 3.52.0), which adjusts raw count data directly for known batch sources without requiring surrogate variable estimation. The effectiveness of batch correction was evaluated by principal component analysis (PCA) using the prcomp function in R (version 4.4) before and after ComBat correction (Supplementary Fig. S1). The DEseq2 package was used to analyze the differentially expressed genes (DEGs). A threshold of |log2FC|> 1 and Padj < 0.05 was applied to enhance the reproducibility and biological relevance of transcriptomic findings [22, 23]. To further analyze the expression patterns of ADGR family members, their expression was compared between normal and cancerous gastric tissues using the Wilcoxon rank-sum test. The results were then normalized using the DEseq2 package for further bioinformatic analyses.

Survival analysis

Along with the gene expression profile, the clinical data of patients with STAD, including follow-up records and survival status, were also retrieved from TCGA database. Patients with follow-up durations shorter than 30 days or missing follow-up records and/or survival data were excluded. The survival data were then combined with the ADGR gene expression data. The survival package in R (version 3.7) was used to stratify the patients into high and low ADGR expression cohorts based on the median expression of each ADGR gene. The median was selected as the cutoff because it yielded the most significant log-rank test result among all tested thresholds, while it also ensures equal distribution of sample sizes between groups and provides an unbiased stratification to avoid type I error inflation [24]. Kaplan–Meier curves were subsequently generated for these groups using the log-rank test.

GEO dataset analysis

To independently validate the expression and prognostic significance of ADGRF4 and ADGRL4 in STAD, three publicly available GEO datasets were used. Gene expression data for GSE54129 and GSE79973 were downloaded and processed using the GEOquery (version 2.72.0) and Biobase (version 2.64.0) R packages. For genes represented by multiple probes, the probe with the highest average expression was selected as the representative. Differential expression between tumor and adjacent normal gastric tissue samples was assessed using the Wilcoxon rank-sum test. For survival validation, the GSE62254 dataset was used. Patients were stratified into high and low expression groups based on the median expression of each gene, and overall survival was compared using Kaplan–Meier analysis with the log-rank test. Univariate Cox proportional hazards regression was additionally performed to evaluate the association between gene expression and survival risk.

TME and immune cell infiltration analysis

The ESTIMATE R package (version 1.0.13) and single-sample gene set enrichment analysis algorithm were used to obtain stromal, immune, and ESTIMATE scores to evaluate the TME composition of STAD tissues. The TME composition was compared between the high and low ADGR expression groups using the Wilcoxon rank-sum test. The TIMER database (https://cistrome.shinyapps.io/timer/) was employed to analyze the association between ADGR expression and immune cell infiltration in STAD. The Cox proportional hazards model was used to evaluate the prognostic significance of ADGRs and six types of tumor-infiltrating immune cells. In addition, the proportions of 22 types of immune cells were analyzed using the CIBERSORT algorithm (v1.03; 1,000 permutations; samples with a CIBERSORT p-value ≥ 0.05 were excluded from subsequent analyses) and compared between the high and low ADGR expression groups using Wilcoxon’s rank-sum test.

Co-expression analysis

To characterize the immunomodulatory interactions of ADGRF4 and ADGRL4, Spearman correlation analysis was performed between their expression levels and the estimated proportions of 22 immune cell types derived from CIBERSORT and the expression levels of representative immune-related molecules, including chemokines (CXCL12, CCL2, CCL22), cytokines (IL-10, TGF-β1), macrophage-associated markers (CD163, MRC1), and immune checkpoint molecules (HAVCR2, CTLA-4, CD274). Differential expression of these immune-related molecules between high and low ADGRF4/ADGRL4 expression groups was assessed using the Wilcoxon rank-sum test.

Construction and validation of the prognostic model

The clinicopathological features (demographic features, tumor staging, and TNM classification) of patients with STAD (patients with missing values were excluded) from the TCGA database were merged with ADGR gene expression profiles. The association between ADGR expression and clinicopathological characteristics was assessed using Wilcoxon’s rank-sum test. Univariate Cox proportional hazards regression analysis was conducted using the survival package to identify significant predictive markers among age, gender, stage, TNM stage, and high/low ADGR gene expression. Subsequently, statistically significant variables were incorporated into multivariate Cox proportional hazards regression analysis to identify independent prognostic factors. The concordance index was calculated to evaluate model discrimination. The rms R package was used to construct nomograms for the prognostic model and predict the 1-, 3-, and 5-year survival rates of patients with STAD. The accuracy of the prognostic model was evaluated using a calibration curve. The time-dependent receiver operating characteristic (ROC) and the area under the curve (AUC) were calculated using the timeROC R package to assess model discrimination. The net clinical benefit of the models was evaluated by decision curve analysis (DCA) using the rmda and ggDCA packages in R.

Genetic alternation analysis

The genetic alterations, such as mutations, amplifications, and deletions, of ADGRs in patients with STAD were analyzed using data from the Genomic Data Commons portal (https://portal.gdc.cancer.gov/) [25].

Protein–Protein Interaction (PPI) analysis

The interactome of ADGRs was determined using the STRING database (https://cn.string-db.org/) with a confidence score cutoff of 0.7. The PPI network was visualized using Cytoscape software.

Functional enrichment analysis

To perform molecular function characterization, ADGR genes were divided into upregulated and downregulated groups based on their expression in cancerous gastric tissues compared with that in normal tissues. Subsequently, the Database for Annotation, Visualization and Integrated Discovery (DAVID) (https://davidbioinformatics.nih.gov/) was used to perform Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Fisher’s exact test was used to determine statistical significance.

Drug sensitivity analysis

The oncoPredict R package was used to analyze drug (AZD7762) sensitivity based on ADGR expression patterns from patients with STAD in TCGA database. The Wilcoxon rank-sum test was used to calculate the sensitivity of the interaction between ADGRs and targeted drugs.

Gene Set Enrichment Analysis (GSEA)

GSEA was performed using clusterProfiler (version 4.14.3) in R (version 4.4) to determine the pathways associated with ADGRL4 expression. Genes were ranked by their correlation with ADGRL4 expression. Hallmark and Reactome gene set collections from the Molecular Signatures Database were used as reference gene sets. A normalized enrichment score (NES) with an adjusted p value < 0.05 was considered statistically significant.

Immunohistochemistry (IHC)

Formalin-fixed, paraffin-embedded sections were processed for IHC as previously reported [26]. In brief, sections were rehydrated, and antigen retrieval was conducted by a heat-induced method using ethylenediaminetetraacetic acid (pH 8.0) in a water bath. Endogenous peroxidase was blocked by incubating the sections with 3% hydrogen peroxide (Sinopharm, Beijing, China) at room temperature for 30 min. The sections were blocked with 5% normal goat serum (Sigma-Aldrich, St. Louis, MO, USA) for 1 h at room temperature, followed by incubation with rabbit anti-ADGRF4 IgG (1:500, NBP2-58,896, Novus, Centennial, CO, USA) and mouse anti-ADGRL4 IgG primary antibodies (1:300, NBP2-59,041, Novus) overnight at 4 °C. Subsequently, the sections were washed and incubated with horseradish peroxidase-conjugated secondary antibody (goat-anti rabbit IgG (1:200, #164,117, Jackson ImmunoResearch, West Grove, PA, USA) or goat-anti mouse IgG (1:200, #164,315, Jackson ImmunoResearch) at room temperature for 30 min. Then, the sections were washed and incubated with 3,3′-diaminobenzidine (DAB, SK-4100, Vector Laboratories, Newark, CA, USA) for 10 min at room temperature. Sections were then rinsed with water, processed for dehydration, counterstained with hematoxylin, and sealed with mounting medium (H-5700–60, Vector Laboratories).

Staining for IHC was analyzed by a relative quantification method using FIJI software (version 2.1.0/153c) via the color deconvolution plugin. The “H DAB” vector was applied to separate hematoxylin and DAB channels. All sections were stained with primary and secondary antibodies under the same dilutions and developed with DAB for the same duration to ensure their comparability. For each section, multiple representative regions of interest (ROIs) were selected and analyzed. The DAB-positive area was quantified using a unified threshold and measured according to the optical density (OD). Specifically, hematoxylin and DAB channels were separated and a unified OD threshold (0.05–1) was applied for all sections to define DAB-positive regions. Then, the OD was calculated for each region of interest and the average OD of multiple ROIs of each section was used as the expression of the sample. Comparisons between groups were performed using Student’s t-test, and P < 0.05 was considered statistically significant.

Results

Using bioinformatic approaches, we explored the DEGs in STAD and normal gastric tissues. Before investigating the DEGs, the effectiveness of batch effect correction was validated by PCA. Before correction, TCGA and GTEx samples were separated along PC1, which accounted for 90.08% of the total variance, indicating source-related technical variation. After correction, this separation was reduced significantly (Supplementary Fig. S1). Among all 33 ADGR family members, 17 genes were significantly upregulated in tumor tissues, whereas four genes were significantly downregulated (Fig. 2). Then, we determined the prognostic value of these ADGRs along with six types of tumor-infiltrating immune cells. The results illustrated that plasmacytoid dendritic cells (hazard ratio [HR] = 0.010, P = 0.02), ADGRF4 (HR = 1.202, P = 0.03), and ADGRL4 (HR = 1.730, P = 0.05) were significantly correlated with the clinical outcomes of STAD (Fig. 3). Subsequently, we conducted Kaplan–Meier survival analysis using the data retrieved from the TCGA database and found that eight ADGRs (ADGRB3, ADGRD1, ADGRE1, ADGRF4, ADGRF5, ADGRG5, ADGRL3, and ADGRL4) were significantly associated with overall survival in patients with STAD (Fig. 4). The results demonstrated that the median survival in patients with high ADGRF4 expression was approximately 25 months, compared with 60 months for patients with low ADGRF4 expression. Similarly, patients with high ADGRL4 expression had a median survival of approximately 25 months, versus 70 months for those with low ADGRL4 expression (Fig. 4). Interestingly, ADGRF4 and ADGRL4 exhibited statistical significance in both clinical relevance analyses, highlighting the importance of these two genes in STAD prognosis. Furthermore, we validated the expression patterns of ADGRF4 and ADGRL4 in STAD tissues using IHC. The results indicated that both ADGRF4 and ADGRL4 were significantly upregulated in cancerous gastric tissues (Fig. 5).

Fig. 2.

Fig. 2

Open in a new tab

Differential expression analysis of ADGR family members in STAD. Box plots demonstrating the expression of 33 ADGRs in STAD compared with that in normal gastric tissues

Fig. 3.

Fig. 3

Open in a new tab

Prognostic significance of ADGRs and tumor-infiltrating immune cells. Forest plot presenting the prognostic significance of ADGRs and tumor-infiltrating immune cells. ADGRF4 (HR = 1.202, P = 0.03) and ADGRL4 (HR = 1.73, P = 0.05) were significantly correlated with the clinical outcomes of patients with STAD

Fig. 4.

Fig. 4

Open in a new tab

Kaplan–Meier survival analysis of ADGRs. Survival curves presenting the association between ADGR expression and overall survival in patients with STAD. Patients with high ADGRF4 and ADGRL4 expression exhibited reduced median survival times

Fig. 5.

Fig. 5

Open in a new tab

Experimental validation of ADGRF4 and ADGRL4 expression. Representative IHC images and analysis results in paired STAD tissue/normal tissue samples (n = 50). ADGRF4 and ADGRL4 expression was significantly elevated in tumor tissues (P < 0.05; scale bar = 50 µm)

To validate the expression patterns identified in TCGA, we analyzed ADGRF4 and ADGRL4 expression in two datasets (GSE54129 and GSE79973) retrieved from the GEO database. In both datasets, ADGRL4 was significantly upregulated in tumors compared with normal gastric tissues, which is consistent with our findings. However, the expression of ADGRF4 in tumor tissues and normal tissues in both cohorts is comparable. For survival validation, Kaplan–Meier analysis in the GSE62254 population demonstrated that patients with high ADGRL4 expression had significantly worse overall survival than those with low ADGRL4 expression. However, high ADGRF4 expression did not show a significant trend (Fig. 6).

Fig. 6.

Fig. 6

Open in a new tab

Validation of ADGRF4 and ADGRL4 expression and prognostic significance using GEO datasets. The expression of ADGRF4 and ADGRL4 in tumor and normal gastric tissues in GSE54129 (A and B) and GSE79973 (C and D). E and F Kaplan–Meier survival curves were plotted showing the association between ADGRF4 (E) and ADGRL4 (F) expression and overall survival using GSE62254 dataset

Despite the growing interest in ADGRs in cancer research, investigations on ADGRF4 and ADGRL4 in STAD remain limited. ADGRF4 has been studied in lung adenocarcinoma [27] and pancreatic ductal adenocarcinoma [28]. ADGRL4 was found to promote the progression of glioma [29] and colorectal cancer [30]. Based on their significant prognostic value, we interrogated their significance in STAD development and their potential as novel diagnostic markers and therapeutic targets. We constructed a prognostic prediction model incorporating ADGRF4/ADGRL4 expression, clinical stage, lymph node metastasis status (N stage), and age. This model effectively predicted 1-, 3-, and 5-year survival rates for patients with STAD (Fig. 7A and C). Subsequently, the efficacy of our prediction model was validated using calibration curves. The results demonstrated that the predictive results were consistent with the actual records at each point (Fig. 7B and D). These results suggested that our prediction model was reliable and accurate. Time-dependent ROC analysis revealed acceptable discriminative performance for both the ADGRF4- and ADGRL4-based prognostic models. In addition, DCA illustrated that both models exhibited net clinical benefit. Together, these findings indicate the potential value of ADGRF4 and ADGRL4 as biomarkers for prognostic prediction in patients with STAD.

Fig. 7.

Fig. 7

Open in a new tab

Construction and validation of clinical prognostic prediction models. Nomograms predicting 1-, 3-, and 5-year survival incorporating clinical stage, N stage, age, and ADGRF4 (A) or ADGRL4 (C) expression. B and D Calibration curves validating model accuracy

Thus far, we have demonstrated that ADGRF4 and ADGRL4 can influence the prognosis of patients with STAD. We further explored the underlying molecular mechanisms. For this purpose, we focused on the TME and immune cell infiltration, as emerging evidence suggests that ADGR is associated with the TME in non-small cell lung cancer [31]. We calculated TME scores and found significant differences in stromal, immune, and ESTIMATE scores between the high and low ADGR expression cohorts (Fig. 8A and C).

Fig. 8.

Fig. 8

Open in a new tab

TME and immune cell infiltration analysis. A and C Dot plot presenting TME scores in high and low ADGRF4 (A) and ADGRL4 expression (C) groups. B and D. CIBERSORT analysis revealing differential immune cell infiltration patterns between high and low ADGRF4 (B) and ADGRL4 expression (D) groups

The high expression groups exhibited higher scores, indicating the involvement of ADGRF4/ADGRL4 in modulation of the TME composition. Moreover, the CIBERSORT algorithm was employed to dissect the specific patterns of immune cell infiltration. Significant differences were observed in the infiltration levels of various immune cells (memory B cells, CD8+ T cells, CD4+ naïve T cells, resting NK cells, monocytes, M0 macrophages, and M1 macrophages) between the high and low ADGRF4 expression groups (Fig. 8B). Similarly, ADGRL4 expression was associated with significant differences in the infiltration of naïve B cells, plasma cells, CD8+ T cells, follicular T helper cells, regulatory T cells, resting NK cells, activated NK cells, monocytes, and M2 macrophages (Fig. 8D). These findings suggested that ADGRF4 and ADGRL4 expression are associated with altered immune cell infiltration and TME composition. These changes may ultimately contribute to their association with STAD prognosis. To investigate the immunomodulatory roles of ADGRF4 and ADGRL4, we examined their co-expression with representative immune-related molecules such as chemokines, cytokines, macrophage-associated markers, and immune checkpoint molecules (Fig. 9). ADGRL4 exhibited positive correlations with multiple immunosuppressive molecules while ADGRF4 showed a weaker and slightly different pattern (Fig. 9A and B). Moreover, the ADGRL4-high group showed significantly elevated expression of immunosuppressive molecules compared with ADGRL4-low group (Supplementary Fig. S2). ADGRF4 showed a relatively weaker and different pattern (Supplementary Fig. S3). Collectively, these observations indicate that ADGRL4 may play a more dominant role in forming an immunosuppressive microenvironment in gastric cancer. Meanwhile, we constructed a PPI network using these two proteins as seed proteins. The results revealed different interacting profiles for each seed protein (Fig. 10), suggesting that ADGRF4 and ADGRL4 harness distinct primary pathways to contribute to STAD progression.

Fig. 9.

Fig. 9

Open in a new tab

Co-expression patterns of ADGRF4 and ADGRL4 with immune-related molecules in STAD. A Spearman correlation heatmap showing the associations between ADGRF4/ADGRL4 expression and CIBERSORT-estimated immune cell fractions. B Spearman correlation heatmap showing the co-expression patterns of ADGRF4/ADGRL4 with representative chemokines, cytokines, macrophage-related markers, and immune checkpoint molecules

Fig. 10.

Fig. 10

Open in a new tab

PPI networks of ADGRF4 and ADGRL4. PPI networks generated using the STRING database present the interacting profiles of ADGRF4 (left) and ADGRL4 (right). Line thickness indicates the strength of data support

To further delineate the potential molecular pathways harnessed by ADGRF4 and ADGRL4 to influence STAD, GO and KEGG enrichment analyses were conducted using DAVID and the results from RNA-seq and PPI network analyses. Functional analysis of the DEGs demonstrated that ADGRF4 was significantly associated with epidermis development, skin development, and epidermal cell differentiation (Fig. 11A), whereas ADGRL4 was enriched in muscle system processes, muscle contraction, and collagen-containing extracellular matrix organization (Fig. 11C). These results indicate that ADGRF4 might be involved in epithelial tissue homeostasis, whereas ADGRL4 might be associated with functions that are critical for tumor cell migration. KEGG pathway analysis revealed both distinct and shared pathways. ADGRF4 displayed enrichment in cytoskeleton in muscle cells and estrogen signaling pathway (Fig. 11B), whereas ADGRL4 was associated with the calcium signaling pathway and cytoskeleton in muscle cells (Fig. 11D). Interestingly, both ADGRF4 and ADGRL4 exhibited significant enrichment in neuroactive ligand–receptor interaction pathways, indicating that these two proteins might share regulatory mechanisms in neural signaling during STAD progression.

Fig. 11.

Fig. 11

Open in a new tab

Function enrichment analysis based on RNA-seq results. GO and KEGG analysis of ADGRF4 (A and B) and ADGRL4 (C and D) based on RNA-seq results

To gain deeper mechanistic insights, GO and KEGG enrichment analyses were performed using all proteins identified in the ADGRF4 and ADGRL4 PPI networks. The results validated and extended our RNA-seq findings, confirming the enrichment of neuroactive ligand–receptor interaction and calcium signaling pathways (Fig. 12). In addition, another layer of importance was added to ADGRF4 and ADGRL4 in critical procedures in cancer development, such as angiogenesis and endothelial cell migration. These findings highlight the shared biological processes by which ADGRF4 and ADGRL4 regulate tumorigenesis despite their distinct PPI networks.

Fig. 12.

Fig. 12

Open in a new tab

Functional enrichment analysis based on PPI network proteins. GO and KEGG analysis of PPI network proteins confirming neuroactive ligand–receptor interactions and revealing additional cancer-related pathways including angiogenesis

Given the significant associations of ADGRF4 and ADGRL4 with STAD prognosis and their potential roles in TME modulation, we investigated their utility as therapeutic targets. Although being largely unexplored, as members of GPCRs, a large family of drug targets [32], ADGRs exhibit promising therapeutic potential. For instance, in vitro and in vivo studies revealed that ADGRD1 is a potential therapeutic target in glioblastoma [8]. To explore the potential of ADGRF4 and ADGRL4 as drug targets in STAD, we analyzed their relationships with drug sensitivity. Interestingly, we found that ADGRL4 expression was significantly correlated with sensitivity to the checkpoint kinase 1/2 (Chk1/Chk2) inhibitor AZD7762 (Fig. 13A). This finding suggests a potential association between ADGRL4 expression and sensitivity to AZD7762. To explore the biological basis of this association, GSEA was performed to determine pathways enriched in ADGRL4-high expression group. The results showed significant enrichment of pathways related to DNA damage response and checkpoint (Fig. 13B, C and Supplementary Table S1).

Fig. 13.

Fig. 13

Open in a new tab

Drug sensitivity analysis of AZD7762. A Correlation between ADGRL4 expression and sensitivity to AZD7762. The results indicate that higher ADGRL4 is associated with increased drug sensitivity. B, C Representative GSEA enrichment plots showing significant enrichment of signaling pathways related to DNA damage response and checkpoint in the ADGRL4 high expression group

Discussion

Using bioinformatic approaches, we analyzed the expression patterns of ADGRs in STAD using data retrieved from TCGA and GTEx. We found that ADGRF4 and ADGRL4 were upregulated compared with their expression in normal gastric tissues, and these findings were validated by IHC in clinical samples. We also revealed that ADGRF4 and ADGRL4 were associated with poor patient prognosis. Consistently, upregulated ADGRs have been documented in various cancerous tissues. For instance, overexpressed ADGRE5 has been reported in undifferentiated thyroid carcinoma [33], ADGRF1 has been reported to be upregulated in lung and prostate cancers [34], and ADGRG1 was revealed to be overexpressed in glioblastoma [29]. Together, these findings indicated that the differential expression of ADGRs might represent a common feature across different cancers.

After exploring the expression patterns, we observed that ADGRF4 and ADGRL4 upregulation resulted in significant changes in stromal component and immune cell populations in STAD tissues. These findings suggest that ADGRF4 and ADGRL4 expression are correlated with changes in TME composition toward an immunosuppressive state. This aligns with findings from previous studies. For instance, ADGRE1 was demonstrated to facilitate M2 macrophage polarization [21] and ADGRE5 was reported to suppress effector T-cell functions [35]. These investigations provide a mechanistic foundation for our observations regarding the association of upregulated ADGRF4 and ADGRL4 with TME modulation and immune cell infiltration in STAD.

Previous studies on other ADGR members, to some degree, supported our findings. For instance, pan-cancer analysis built an association between ADGR-related genes with immune cell infiltration and poor prognosis in STAD [14]. Meanwhile, GPCR-mediated signaling has been shown to be critical for establishing an immunosuppressive TME [16]. Our findings on ADGRF4 and ADGRL4 extended this evidence by showing their specific associations with TME composition and prognosis in STAD. Importantly, co-expression analysis revealed that ADGRL4 was correlated with immunosuppressive molecules (PD-L1, CTLA-4, and HAVCR2) and M2 macrophage-associated markers (CD163 and MRC1), which indicates that ADGRL4 may contribute to immune evasion. Meanwhile, ADGRF4 demonstrated a weaker and different correlation pattern. The differences reflect the differential immunoregulatory characteristics of these two molecules.

Furthermore, our GO and KEGG enrichment analyses revealed distinct yet complementary mechanisms by which ADGRF4 and ADGRL4 influence STAD progression. The results of RNA-seq indicated that ADGRF4 primarily regulates epidermal development, which was partially supported by a recent study illustrating the involvement of ADGRF4 in epidermal differentiation and the regulation of epithelial structure in human skin [36]. On the contrary, ADGRL4 was associated with extracellular matrix organization and calcium signaling pathways. Similarly, results from other groups highlighted the involvement of ADGRL4 in extracellular matrix remodeling and calcium-dependent cellular activities [12, 13]. Despite the distinct primary locations of ADGRF4 and ADGRL4 along the signaling transduction cascade in STAD progression, both proteins participated in neuroactive ligand–receptor interactions, highlighting their potential significance in neural signaling regulation.

Meanwhile, our PPI network-based functional enrichment analysis further validated the findings of RNA-seq analysis. The analysis confirmed the shared enrichment of both ADGRF4 and ADGRL4 in neuroactive ligand–receptor interactions, as well as cancer-related pathways that were not identified in RNA-seq analysis, such as angiogenesis. Together, these findings revealed shared regulatory roles of ADGRF4 and ADGRL4 in calcium signaling, angiogenesis, and neural communication pathways, providing mechanistic insights into their functions in STAD. Notably, although ADGRF4 and ADGRL4 share a regulatory role in neuroactive ligand–receptor interactions, no shared nodes were identified in their PPI networks. This indicates that these proteins regulate similar cellular functions through different upstream signaling pathways. Thus, ADGRF4 and ADGRL4 might provide complementary prognostic information, and their combined detection potentially boosts their prognostic value.

The independent PPI network of ADGRF4 and ADGRL4, to some degree, reflects the molecular heterogeneity of STAD. It is possible that different subgroups depend on different molecular pathways, laying a molecular basis for individualized treatment strategies. In addition, our drug sensitivity analysis indicated that high ADGRL4 expression was significantly associated with increased sensitivity to AZD7762. Despite the fact that this result was obtained from computational analysis, it raises the possibility that ADGRL4 expression may be associated with the DNA damage response pathway. Consistently, GSEA demonstrated significant enrichment of DNA damage response and checkpoint signalings in the ADGRL4 high expression group. These findings support the predicted association between ADGRL4 expression and AZD7762 sensitivity. Future in vitro drug sensitivity investigations would be needed to validate and provide mechanistic foundations for this observation.

To validate our findings, we utilized multiple datasets retrieved from the GEO database. The validation yielded consistent results for ADGRL4 as reflected by its significant upregulation in tumor tissues and association with poor overall survival. However, for ADGRF4, the validation results were not consistent. Differential expression between cancerous and normal tissues did not reach statistical significance in both datasets (GSE54129 and GSE79973), and its prognostic trend did not achieve significance. This inconsistency could be attributed to lower sensitivity of microarray platforms compared to the RNA-seq for genes with moderate expression differences. It is also possible that differences in molecular subtype composition between different populations contributed to this inconsistency [37]. Together, these findings confirmed that ADGRL4 can serve as a biomarker for STAD, while the role of ADGRF4 in STAD needs further investigation.

Nevertheless, this study had several limitations. First, our findings regarding the roles of ADGRF4 and ADGRL4 in tumor immune cell infiltration were based on correlation analyses, and the lack of mechanistic validation limits our ability to establish causality. Second, although IHC confirmed the upregulation of these genes in STAD tissues, we did not perform functional experiments, such as gene overexpression or knockdown, to investigate their direct impact on tumor progression or immune modulation. Third, we did not explore the potential crosstalk between ADGRF4/ADGRL4 and other signaling pathways known to be dysregulated in STAD. Future studies are needed to validate these findings and elucidate the underlying molecular mechanisms. Specifically, functional experiments and biological models should be utilized to validate the roles of ADGRF4 and ADGRL4 in STAD progression. Gene overexpression or silencing experiments in gastric cancer cell lines could help determine their direct effects on cell proliferation, migration, and invasion. Additionally, xenograft or genetically engineered mouse models could provide insights into the in vivo functions of these receptors. Moreover, the mechanisms by which ADGRF4 and ADGRL4 modulate the TME and immune cell infiltration should be investigated.

Conclusions

In conclusion, our bioinformatics and IHC study demonstrates that ADGRF4 and ADGRL4 are significantly overexpressed in STAD and independently associated with poor prognosis. Their expression is correlated with immune cell infiltration and changes in TME, which indicate their potential significance in shaping the immunological landscape in STAD. Enrichment analyses further reveal their involvement in tumor progression, as reflected by their association with pathways such as angiogenesis and calcium signaling. Together, these findings suggest that ADGRF4 and ADGRL4 could serve as biomarkers for prognosis, and they may also possess therapeutic implications in patients with STAD.

Supplementary Information

Supplementary Material 1. (544.6KB, docx)

Acknowledgements

We thank Medjaden Inc. for scientific editing of this manuscript.

Abbreviations

STAD

Stomach adenocarcinoma

GPCRs

G protein-coupled receptors

TME

Tumor microenvironment

TCGA

The Cancer Genome Atlas

GTEx

Genotype-Tissue Expression

ROC

Receiver operating characteristic

AUC

Area under the curve

DCA

Decision curve analysis

PPI

Protein–protein interaction

DAVID

Database for Annotation, Visualization and Integrated Discovery

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

IHC

Immunohistochemistry

OD

Optical density

Authors’ contributions

JHW and PH designed and coordinated the study; JHW, MJL and DDZ performed the experiments, acquired and analyzed data; JHW, XWW and PH interpreted the data; JHW and PH wrote the manuscript; all authors revised the manuscript, read and approved the final version of the article.

Funding

The study was supported by 2022 Panjin Natural Science Foundation Program Guiding Program Project (2022YL008).

Data availability

The RNA-seq data for STAD samples (n = 446) were obtained from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/projects/TCGA-STAD). Normal gastric tissue gene expression data (n = 503) were retrieved from the Genotype-Tissue Expression (GTEx) database (https://gtexportal.org/home/tissue/Stomach, dbGaP Accession phs000424.v8.p2).

Declarations

Ethics approval and consent to participate

This research was approved by the Ethics Committee of Panjin Central Hospital (Approval No.: 20250120LWLL-KS-02) and was performed in compliance with the Declaration of Helsinki. Written informed consent was obtained from all participants for participation in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209–49. [DOI] [PubMed] [Google Scholar]
  • 2.Xue W, Dong B, Wang Y, Xie Y, Li P, Gong Z, et al. A novel prognostic index of stomach adenocarcinoma based on immunogenomic landscape analysis and immunotherapy options. Exp Mol Pathol. 2022;128:104832. [DOI] [PubMed] [Google Scholar]
  • 3.Xu X, Lambert NA, Wu G. Sequence-directed concentration of G protein-coupled receptors in COPII vesicles. iScience. 2023;26:107969. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hauser AS, Attwood MM, Rask-Andersen M, Schiöth HB, Gloriam DE. Trends in GPCR drug discovery: new agents, targets and indications. Nat Rev Drug Discov. 2017;16:829–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kolakowski LF Jr. GCRDb: a G-protein-coupled receptor database. Recept Channels. 1994;2:1–7. [PubMed] [Google Scholar]
  • 6.Bjarnadóttir TK, Fredriksson R, Schiöth HB. The adhesion GPCRs: a unique family of G protein-coupled receptors with important roles in both central and peripheral tissues. Cell Mol Life Sci. 2007;64:2104–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hamann J, Aust G, Araç D, Engel FB, Formstone C, Fredriksson R, et al. International Union of Basic and Clinical Pharmacology. XCIV. Adhesion G protein-coupled receptors. Pharmacol Rev. 2015;67:338–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Byrne KF, Pal A, Curtin JF, Stephens JC, Kinsella GK. G-protein-coupled receptors as therapeutic targets for glioblastoma. Drug Discov Today. 2021;26:2858–70. [DOI] [PubMed] [Google Scholar]
  • 9.Ng KF, Chen TC, Stacey M, Lin HH. Role of ADGRG1/GPR56 in tumor progression. Cells. 2021. 10.3390/cells10123352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Lei P, Wang H, Yu L, Xu C, Sun H, Lyu Y, et al. A correlation study of adhesion G protein-coupled receptors as potential therapeutic targets in uterine corpus endometrial cancer. Int Immunopharmacol. 2022;108:108743. [DOI] [PubMed] [Google Scholar]
  • 11.Ganesh RA, Venkataraman K, Sirdeshmukh R. GPR56: an adhesion GPCR involved in brain development, neurological disorders and cancer. Brain Res. 2020;1747:147055. [DOI] [PubMed] [Google Scholar]
  • 12.Masiero M, Simões FC, Han HD, Snell C, Peterkin T, Bridges E, et al. A core human primary tumor angiogenesis signature identifies the endothelial orphan receptor ELTD1 as a key regulator of angiogenesis. Cancer Cell. 2013;24:229–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Guihurt Santiago J, Burgos-Tirado N, Lafontaine DD, Mendoza Sierra JC, Camacho RH, Vecchini Rodríguez CM, et al. Adhesion G protein-coupled receptor, ELTD1, is a potential therapeutic target for retinoblastoma migration and invasion. BMC Cancer. 2021;21:53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Lyu G, Li D. ZNF165: A Pan-Cancer Biomarker with Prognostic and Therapeutic Potential. Protein Pept Lett. 2025;32:206–23. [DOI] [PubMed] [Google Scholar]
  • 15.Li L, Li Y, Zhou X. Advancements in the Research of GEF-H1: Biological Functions and Tumor Associations. Curr Mol Pharmacol. 2024;17:e18761429274883. [DOI] [PubMed] [Google Scholar]
  • 16.Ye Q, Li D, Zou Y, Yuan Y. The Role and Treatment Strategies of Ammonia-Related Metabolism in Tumor Microenvironment. Curr Gene Ther. 2025;25:199–209. [DOI] [PubMed] [Google Scholar]
  • 17.Zhang Y, Zhao Z, Huang W, Kim BS, Lin L, Li X, et al. Pan-Cancer Single-Cell Analysis Revealing the Heterogeneity of Cancer- Associated Fibroblasts in Skin Tumors. Curr Gene Ther. 2025;25:793–821. [DOI] [PubMed] [Google Scholar]
  • 18.Peng Z, Fang C, Tong Z, Rao Q, Ren Z, Hu K. Crosstalk Between Cancer-associated Fibroblasts and Myeloid Cells Shapes the Heterogeneous Microenvironment of Gastric Cancer. Curr Genomics. 2024;25:390–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Wang Q, Shao X, Zhang Y, Zhu M, Wang FXC, Mu J, et al. Role of tumor microenvironment in cancer progression and therapeutic strategy. Cancer Med. 2023;12:11149–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zhang W, Lee A, Tiwari AK, Yang MQ. Characterizing the tumor microenvironment and its prognostic impact in breast cancer. Cells. 2024. 10.3390/cells13181518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Akter R, Kim K, Kwon HY, Kim Y, Eom YW, Cho HM, et al. EMR1/ADGRE1 expression in cancer cells upregulated by tumor-associated macrophages is related to poor prognosis in colorectal cancer. Biomedicines. 2022. 10.3390/biomedicines10123121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Balikci Cicek I, Kucukakcali Z. Transcriptomic profiling of gastric cancer reveals key biomarkers and pathways via bioinformatic analysis. Genes. 2025. 10.3390/genes16070829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Siavoshi A, Piran M, Sharifi-Zarchi A, Ataellahi F. Integration of Gastric Cancer RNA-Seq Datasets Along With PPI Network Suggests That Nonhub Nodes Have the Potential to Become Biomarkers. Cancer Rep (Hoboken). 2025;8:e70126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Cancer Genome Atlas Research Network. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 2014;513:202–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Heath AP, Ferretti V, Agrawal S, An M, Angelakos JC, Arya R, et al. The NCI Genomic Data Commons. Nat Genet. 2021;53:257–62. [DOI] [PubMed] [Google Scholar]
  • 26.Hou Y, Sun X, Gheinani PT, Guan X, Sharma S, Zhou Y, et al. Epithelial SMYD5 Exaggerates IBD by Down-regulating Mitochondrial Functions via Post-Translational Control of PGC-1α Stability. Cell Mol Gastroenterol Hepatol. 2022;14:375–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Wang Y, Shi M, Yang N, Zhou X, Xu L. GPR115 Contributes to Lung Adenocarcinoma Metastasis Associated With LAMC2 and Predicts a Poor Prognosis. Front Oncol. 2020;10:577530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Nishiwada S, Shimura T, Yamamura K, Nakagawa K, Nagai M, Nakamura K, et al. Clinical significance and functional role of adhesion G-protein-coupled receptors in human pancreatic ductal adenocarcinoma. Br J Cancer. 2023;128:321–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Huang H, Georganaki M, Conze LL, Laviña B, van Hooren L, Vemuri K, et al. ELTD1 deletion reduces vascular abnormality and improves T-cell recruitment after PD-1 blockade in glioma. Neuro Oncol. 2022;24:398–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sun J, Zhang Z, Chen J, Xue M, Pan X. ELTD1 promotes invasion and metastasis by activating MMP2 in colorectal cancer. Int J Biol Sci. 2021;17:3048–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lv M, Li X, Tian W, Yang H, Zhou B. ADGRD1 as a Potential Prognostic and Immunological Biomarker in Non-Small-Cell Lung Cancer. Biomed Res Int. 2022;2022:5699892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lorente JS, Sokolov AV, Ferguson G, Schiöth HB, Hauser AS, Gloriam DE. GPCR drug discovery: new agents, targets and indications. Nat Rev Drug Discov. 2025;24:458–79. [DOI] [PubMed] [Google Scholar]
  • 33.Hoang-Vu C, Bull K, Schwarz I, Krause G, Schmutzler C, Aust G, et al. Regulation of CD97 protein in thyroid carcinoma. J Clin Endocrinol Metab. 1999;84:1104–9. [DOI] [PubMed] [Google Scholar]
  • 34.Lum AM, Wang BB, Beck-Engeser GB, Li L, Channa N, Wabl M. Orphan receptor GPR110, an oncogene overexpressed in lung and prostate cancer. BMC Cancer. 2010;10:40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Sutavani RV, Bradley RG, Ramage JM, Jackson AM, Durrant LG, Spendlove I. CD55 costimulation induces differentiation of a discrete T regulatory type 1 cell population with a stable phenotype. J Immunol. 2013;191:5895–903. [DOI] [PubMed] [Google Scholar]
  • 36.Winkler R, Quaas M, Glasmacher S, Wolfrum U, Thalheim T, Galle J, et al. The adhesion G-protein-coupled receptor GPR115/ADGRF4 regulates epidermal differentiation and associates with cytoskeletal KRT1. Cells. 2022. 10.3390/cells11193151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Li B, Zhang F, Niu Q, Liu J, Yu Y, Wang P, et al. A molecular classification of gastric cancer associated with distinct clinical outcomes and validated by an XGBoost-based prediction model. Mol Ther Nucleic Acids. 2023;31:224–40. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (544.6KB, docx)

Data Availability Statement

The RNA-seq data for STAD samples (n = 446) were obtained from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/projects/TCGA-STAD). Normal gastric tissue gene expression data (n = 503) were retrieved from the Genotype-Tissue Expression (GTEx) database (https://gtexportal.org/home/tissue/Stomach, dbGaP Accession phs000424.v8.p2).

Articles from BMC Gastroenterology are provided here courtesy of BMC

RESOURCES