RNA-sequencing (RNA-seq) data from 448 gastric cancer (STAD) samples, including 410 tumor and 38 normal samples, were collected, along with clinical data from 383 patients with STAD, sourced from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/) (37). The Limma package in R 4.5.1 (https://bioconductor.org/packages/release/bioc/html/limma.html) facilitated the standardization of RNA-seq information (38). The transcriptomic details for all samples were preserved to investigate the nuanced differences between adenocarcinomatous and adjacent tissues. However, 17 samples with incomplete clinical information were excluded from subsequent survival analysis.

The RNA-seq data and associated clinical/prognostic information for 300 patients diagnosed with gastric cancer were methodically extracted from the GSE66254 dataset in the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE66254) (39). The original chip data from GSE66254 underwent a rigorous annotation and standardization process in R, utilizing the Limma package.

From the biobank of the Department of Gastrointestinal and Glandular Surgery at The First Affiliated Hospital of Guangxi Medical University (Nanning, China), tissue specimens (including gastric cancer tumor and adjacent non-tumorous tissues) were collected from 48 patients with gastric cancer who underwent surgery. Postoperative pathological reports indicated lymph node metastasis in 24 of these patients. These tissue samples were subsequently utilized to analyze the correlation between lymph node metastasis and the activation of the PI3K/AKT/mTOR and EMT signaling pathways, as well as the expression of GPR176.

The cell lines HGC-27 and NCI-N87 were procured from the American Type Culture Collection (The Global Bioresource Center; http://www.atcc.org/) cell repository. Culturing these cells involved a comprehensive medium consisting of DMEM (cat. no. 10566016; Gibco, Thermo Fisher Scientific, Inc.), supplemented with 10% FBS (cat. no. 10099141C; Gibco; Thermo Fisher Scientific, Inc.), 1% streptomycin and 1% penicillin (cat. no. P1400; Beijing Solarbio Science & Technology Co., Ltd.). The controlled environment for these cell cultures was maintained in incubators with 5% CO₂ and 37°C.

RNA extraction from each sample (gastric cancer tumor and adjacent non-tumorous tissues) was carried out using the TRI Reagent^™, by following the manufacturer's protocol (cat. no. AM9738; Invitrogen; Thermo Fisher Scientific, Inc.). Subsequently, reverse transcription (RT) into cDNA was carried out according to the manufacturer's instructions for the PrimeScript RT Reagent Kit with gDNA Eraser (cat. no. RR047B; Takara Bio, Inc.). The primers were designed using Primer3 (https://primer3.ut.ee/) and are listed in Table SI. qPCR was performed using the FastStart Universal SYBR^® Green Master Mix (cat. no. 06402712001; Roche Diagnostics GmbH) on an Applied Biosystems 7500 (Thermo Fisher Scientific, Inc.) with the following thermocycling conditions: Initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. Gene expression was analyzed using the 2^−∆∆Cq method (40).

Design and packaging of overexpression (OE)/RNA interfering (RNAi) lentiviruses into cells was carried out by Shanghai GeneChem Co., Ltd. Lentiviruses for GPR176 overexpression, GPR176 knockdown (sh-GPR176), PIP5K1A overexpression and PIP5K1A knockdown (sh-PIP5K1A) were constructed using the wild-type sequences of GPR176 and PIP5K1A, respectively. Lentiviruses containing empty vectors were used as the control group for infection. The infection concentration of all lentiviruses in the present study was set to 5 multiplicity of infection (MOI), with an infection duration of 12 h in 37°C cell incubator. Puromycin was added at a concentration of 10 ng/ml to select cells with off-target effects and the selection period was 5 days in in 37°C cell incubator. The GPR176 overexpression lentivirus contained the wild-type sequence of the GPR176 gene, while the PIP5K1A overexpression lentivirus contained the wild-type sequence of the PIP5K1A gene. The functional sequences of sh-GPR176 and sh-PIP5K1A are provided in Table SII. The efficacy of these lentiviruses was verified through a using RT-qPCR assay and western blotting (WB).

Extraction of proteins from cells (HGC-27 and NCI-N87) and tissues (gastric cancer and adjacent normal gastric) entailed a combination of RIPA reagent (cat. no. R0010; Beijing Solarbio Science & Technology Co., Ltd.) and 1% PMSF (cat. no. P0100; Beijing Solarbio Science & Technology Co., Ltd.). Quantification of protein concentration was carried out using the BCA protein assay kit (cat. no. P0009; Beyotime Institute of Biotechnology). The proteins (2 µg) underwent separation using 10% SDS-PAGE electrophoresis and were then transferred to PVDF membranes. Following blocking with 5% skim milk at room temperature for 30 min, the PVDF membranes were incubated overnight at 4°C with the diluent for the primary antibody. After two washes with PBS-Tween (1%), the membranes were incubated with secondary antibody diluent at 23°C for 1 h. The visualization of protein bands was carried out using the Bio-Rad ChemiDoc MP Imaging System (Bio-Rad Laboratories, Inc.). The information and usage concentrations of the corresponding primary antibodies are listed in Table SIII. HRP-conjugated Goat Anti-Mouse IgG (H+L) (1:100; cat. no. SA00001-1, Proteintech Group Inc.) and HRP-conjugated Goat Anti-Rabbit IgG (H+L) (1:100; cat. no. SA00001-2, Proteintech Group Inc.) were used as the secondary antibody.

IF staining was carried out following standard protocols (https://www.ptgcn.com/support/protocols/#if). Briefly, 1×10⁵ HGC-27 and NCI-N87 cells were fixed with 4% paraformaldehyde at room temperature for 15 min, permeabilized with 0.2% Triton X-100 (cat. no. 9002-93-1; Beijing Solarbio Science & Technology Co., Ltd.) and blocked with 5% normal goat serum (cat. no. SL038, Beijing Solarbio Science & Technology Co., Ltd.) at room temperature for 1 h. Samples were then incubated with primary antibodies (Vimentin, 1:200; E-cadherin, 1:200) at 4°C overnight, followed by Multi-rAb^® CoraLite^® Plus 488-Goat Anti-Rabbit Recombinant Secondary Antibody (H+L) (1:500; cat. no. RGAR002; Proteintech Group Inc.) and Multi-rAb^® CoraLite^® Plus 594-Goat Anti-Rabbit Recombinant Secondary Antibody (H+L) (1:500; cat. no. RGAR004; Proteintech Group Inc.) at room temperature for 1 h and DAPI (cat. no. 28718-90-3, Beijing Solarbio Science & Technology Co., Ltd.) nuclear counterstaining at room temperature for 15 min. Fluorescence images were captured with an Olympus inverted fluorescence microscope (Olympus IX73; Olympus Corporation) equipped with appropriate filter sets and a digital camera and analyzed using Olympus CellSens software 2.3.

The cell suspension was obtained by digesting the cells with 1 ml trypsin (cat. no. T1321; Beijing Solarbio Science & Technology Co., Ltd.) for 2 min, adding 5 ml of complete medium, and then repeatedly pipetting the mixture. Before adding the cell suspension in 6.5 mm Transwell^® with 8.0 µm Pore Polycarbonate Membrane Inserts (cat. no. 3422; Corning, Inc.), the Matrigel (cat. no. 356234; Beijing Solarbio Science & Technology Co., Ltd.) was prepared according to the manufacturer's instructions. Cells were plated in the upper chambers, and the culture medium was added to the lower chambers. Specifically, 200 µl of a serum-free cell suspension containing 100,000 cells was added to the upper Transwell inserts, while 600 µl of DMEM supplemented with 10% FBS was added to the lower chambers of a 24-well plate. The plate was then incubated for 48 h at 37°C and 5% CO₂. After incubation, the insert was fixed with 4% paraformaldehyde at room temperature for 30 min and stained with a 1% crystal violet solution at room temperature for 20 min. Subsequent actions included the erasure of cells on the upper layer, which were then removed, washed with PBS, and the invaded cells were observed under an Olympus IX53.

Cells (HGC-27 cells and NCI-N87) were grouped according to the regulation of GPR176 and PIP5K1A, with the details stated in Construction of lentivirus and stable cell lines. To observe cell proliferation (HGC-27 and NCI-N87), a precise wound was induced by scratching the plate with a 200 µl suction tip when the cells covered 90–100% of the culture plate area. After washing twice with PBS, serum-free medium was added to the plate, and cells were cultured in incubators with 5% CO₂ and 37°C. Subsequently, the width of the wound was observed and imaged consecutively under the Olympus IX53 at 0 and 48 h.

Male nude mice (n=40; 6–8 weeks old; 18–22 g) were housed under specific pathogen-free conditions with a controlled temperature of 22±2°C, humidity of 50±10% and a 12-h light/dark cycle. All animals had ad libitum access to a standard laboratory diet and autoclaved water. The nude mice were randomly assigned into four groups (10 per group) and received subcutaneous injections of HGC-27 cells: sh-NC, sh-GPR-176, Lenti-null and Lenti-GPR-176, respectively. HGC-27 cells in a stable growth state were resuspended in pre-cooled PBS at a concentration of 1×10⁷ cells/200 µl. This suspension was aseptically injected subcutaneously into the right flank of each mouse.

After transplantation, the general health status and body weight of the mice were monitored daily. Tumor volume was measured every two days using a digital caliper and calculated using the formula V=(length × width² × π)/6. To minimize observer bias, all measurements were conducted in a blinded fashion. Humane endpoints were strictly defined as follows: Tumor burden >1.5 cm in diameter, significant weight loss (>20% of initial body weight) or signs of severe distress. None of the animals reached these endpoints prior to the scheduled experimental conclusion. At the end of the experiment (day 28 post-inoculation), all mice were euthanized by CO₂ asphyxiation with a chamber displacement rate of 30–50% per min, followed by cervical dislocation to ensure mortality.

Each experiment was carried out to at least three independent replicates. Data analysis was conducted using IBM SPSS Statistics software (version 26.0; IBM Corp.). Results are expressed as the mean ± standard error. Differences between two groups were assessed using an unpaired Student's t-test, while multiple comparisons were carried out by one-way or two-way ANOVA. Linear regression and correlation analysis were employed to assess the correlation between GPR176 and PIP5K1A expression. P<0.05 was considered to indicate a statistically significant difference.

Expression matrices and corresponding clinical data were respectively obtained from TCGA-STAD dataset and GSE66254 dataset (37,39). The expression of GPR176, its prognostic significance and its association with clinicopathologic factors were evaluated. Analysis of the pan-cancer data from TCGA revealed significant differences in GPR176 expression between cancer and adjacent tissues in cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), stomach adenocarcinoma (STAD) and uterine corpus endometrial carcinoma (UCEC) (Fig. 1A). In the TCGA-STAD dataset, the expression of GPR176 was significantly higher in gastric cancer tissues when compared with adjacent tissues (Fig. 1B). Additionally, GPR176 expression levels were lower in patients with TNM stage I gastric adenocarcinoma when compared with patients with stage II/III/IV gastric adenocarcinoma (Fig. 1C), and GPR176 expression levels were significantly lower in patients with stage T1 when compared with patients with pathologic stage T2/T3/T4 (Fig. 1D).

Statistical analysis confirmed the association of GPR176 with tissue type and tumor stage, suggesting its importance in the occurrence and progression of gastric adenocarcinoma. Subsequently, in the TCGA-STAD and GSE66254 datasets, significant association was observed between GPR176 expression levels and overall survival/recurrence-free survival in gastric adenocarcinoma, with high GPR176 expression levels associated with worse prognosis (Fig. 1E-G). Using the median expression of GPR176 as a cut-off point, patients were categorized into GPR176-High and GPR176-Low groups. An increased proportion of patients at T3/T4 in the GPR176-High group was observed (Fig. 1H). The area under the receiver operating curve reached 0.8315, indicating GPR176 as a good biomarker to distinguish gastric adenocarcinoma from normal gastric tissue (Fig. 1I). Correlation analysis of GPR176 with PIP5K1A expression levels in gastric adenocarcinoma tissues using the TCGA-STAD dataset verified a linear positive association (Fig. 1J).

Nomogram were created based on GPR176 expression and clinicopathologic parameters. Univariate Cox regression analysis and multivariate Cox regression analysis were carried out using clinical case characteristics such as age, sex, TNM stage, T stage, N stage, M stage, histologic grade and GPR176 expression (Table SIV). The results of univariate Cox regression analysis results indicated that age, sex, TNM stage, T stage, N stage, M stage and histological grade were associated with the overall survival of patients with gastric adenocarcinoma. In Multivariate Cox regression analysis, only age, histologic grade and GPR176 expression were associated with overall survival of patients with gastric adenocarcinoma. The nomogram graph based on age, sex, TNM stage, T stage, N stage, M stage, histological grade and GPR176 expression were constructed to assess the risk of mortality for specific patients (Fig. 2A). The predictive power of the histogram was evaluated by comparing the grade between the training group and the validation group. The nomogram showed a high degree of overlap between the self-validation cohort and the training group in predicting the 1-, 3- or 5-year prognosis (Fig. 2B-D).

Based on the expression levels of GPR176, patients in the TCGA-STAD dataset were categorized into high and low expression groups. Differential expression analysis was carried out using the Limma package for RNA-seq and a volcano plot was generated (Fig. 3A). The corresponding pie charts showed that 326 genes were upregulated, and 581 genes were downregulated. Functional enrichment analysis of differentially expressed genes associated with GPR176 revealed enrichment in signaling pathways such as ‘PI3K-Akt signaling pathway’; Fig. 3B), and cellular functions such as ‘Cell adhesion’ and ‘G-protein-coupled receptor signaling’ (Fig. 3C). Gene Set Enrichment Analysis indicated associations between GPR176 and tumor-related signaling pathways such as ‘cell adhesion molecules CAMs’, ‘Hedgehog signaling pathway’, ‘Jak-stat signaling pathway’ and ‘MAPK signaling pathway’ (Fig. 3D). CDK2, FOXO1, FOXO3 and VEGFA are downstream target genes of the PI3K/AKT/mTOR signaling pathway, reflecting the activation/inhibition status of the pathway (41). Linear regression analysis using RNA-seq data from gastric adenocarcinoma tissues in TCGA-STAD revealed an association between GPR176 and CDK2, FOXO1, FOXO3 and VEGFA (Fig. 3E).

To confirm the effect of GPR176 on the biological behavior of gastric cancer cells and explore the corresponding mechanisms, lentiviral vectors were used to overexpress and knockdown GPR176 expression levels. RT-qPCR and WB confirmed the satisfactory efficiency of OE lentivirus and RNAi lentivirus in HGC-27 and NCI-N87 cells (Fig. S1A). Subsequently, scratch-wound healing and Transwell assays were carried out to evaluate the effects of modulating GPR176 expression on cell migration and invasion. After upregulation of GPR176, the migration and invasion ability of cells increased significantly, while downregulation of GPR176 resulted in a significant decrease in this ability (Fig. 4A and B). Subsequent RT-qPCR analysis revealed a significant increase in the mRNA levels of EMT-activating genes, such as CDH2, VIM and SNAI1, following upregulation of GPR176 (Fig. 4C). By contrast, the expression of the gene CDH1 which encodes E-cadherin, associated with the inhibition of the EMT pathway (42), significantly decreased (Fig. 4C). Conversely, downregulation of GPR176 led to a significant decrease in the mRNA levels of CDH2, VIM and SNAI1, accompanied by a significant increase in the mRNA levels of CDH1, associated with the inhibition of the EMT pathway (Fig. 4C). WB results corroborated the RT-qPCR results and revealed an increase in the protein concentration of N-cadherin, vimentin and Snai1, and a decrease in the protein concentration of E-cadherin after upregulation of GPR176 (Fig. 4D), with the corresponding bar chart shown in Fig. S2A and B. IF showed that GPR176 upregulation increased vimentin and reduced E-cadherin, whereas GPR176 knockdown had the opposite effect (Fig. S1E). The results from RT-qPCR, WB, and IF experiments showed that GPR176 expression was associated with the activation of the EMT signaling pathway and increased migration and invasion of gastric adenocarcinoma cells.

Previous studies have confirmed that PIP5K1A is an upstream molecule in the PI3K/AKT/mTOR signaling pathway (30–33). Analysis of RNA-seq data from the TCGA-STAD dataset in the present study revealed a significant positive association between GPR176 and PIP5K1A expression (Fig. 1J). Therefore, we hypothesized that GPR176 may activate the PI3K/AKT/mTOR signaling pathway by inducing upregulation of PIP5K1A expression. Interference and overexpression experiments of PIP5K1A were performed based on the overexpression of GPR176 and the repression of GPR176. Analysis revealed that interference with PIP5K1A expression significantly prevented the migration and invasion induced by GPR176 overexpression (Fig. 5A and B). Conversely, downregulation of PIP5K1A also significantly reversed the slowed migration and invasion of gastric cancer cells caused by GPR176 knockdown (Fig. 5A and B). Following this, WB was carried out to evaluate the phosphorylation levels of molecules within the PI3K/AKT/mTOR pathway and the EMT signaling pathway, along with assessing the expression levels of PIP5K1A. The activation of the PI3K/AKT/mTOR pathway and EMT signaling pathway after GPR176 upregulation was attenuated by the downregulation of PIP5K1A. Conversely, the inhibition of the PI3K/AKT/mTOR pathway and EMT signaling pathway after GPR176 downregulation was counteracted by the OE of PIP5K1A (Fig. 6A and B), with the corresponding bar chart shown in Fig. S3A and B. The results of the RT-qPCR assay were in concordance with the aforementioned experiments, indicating a preventing in changes to the mRNA expression levels of molecules in the EMT signaling pathway after GPR176 upregulation, which was suppressed by the downregulation of PIP5K1A in HGC-27 and NCI-N87 cells; similarly, in HGC-27 and NCI-N87 cells, the downregulation of GPR176 resulted in prevention of the mRNA expression levels of molecules in the EMT signaling pathway, under the upregulating influence of PIP5K1A (Fig. 7A-L).

Firstly, patients with gastric cancer were categorized into two groups based on the presence or absence of lymph node metastasis: Those with lymph node metastasis and those without. Further classification was made based on histological type, distinguishing between the para-cancer group and the cancer group. Initially, WB was employed to assess the protein levels of E-cadherin, N-cadherin, PIP5K1A, GPR176, AKT, p-AKT, mTOR and p-mTOR in each group. The findings revealed that irrespective of lymph node metastasis, the expression of E-cadherin in the cancer group was significantly reduced compared with the para-cancer group. Conversely, levels of N-cad, PIP5K1A, GPR176, p-AKT, p-mTOR, p-AKT/AKT and p-mTOR/mTOR were increased in the cancer group compared with the para-cancer group (Fig. 8A-F). The representative blots of WB experiments are referenced in Fig. 8G.

Subsequently, validation of the expression levels of key genes in the EMT pathway and PI3K/AKT/mTOR signaling pathway, as well as PIP5K1A and GPR186, was conducted using RT-qPCR. Among patients without lymph node metastasis, no significant differences were observed in GPR186 and CDH2 (N-cadherin) expression levels between the para-cancer and cancer groups (Fig. 8I and K). However, CDH1 (E-cadherin) expression was significantly reduced in the cancer group compared with the para-cancer group (Fig. 8H). Conversely, PIP5K1A expression was markedly higher in the cancer group compared with the para-cancer group (Fig. 8J). In patients with lymph node metastasis, there were no significant differences in CDH1 expression between the para-cancer and cancer groups (Fig. 8L). Nevertheless, GPR186, CDH2 and PIP5K1A expression levels were significantly increased in the cancer group compared with the para-cancer group (Fig. 8M-O). In summary, the expression levels of PIP5K1A and GPR176 were increased in the cancer group compared with the para-cancer group. Additionally, activation of the EMT pathway and PI3K/AKT/mTOR pathway was evident in the cancer group.

In vitro experiments provided evidence that GPR176 could activate the PI3K/AKT/mTOR pathway, facilitating EMT and promoting cell proliferation by inducing the overexpression of PIP5K1A. To further validate the role of GPR176 in gastric cancer, a nude mouse subcutaneous tumor experiment was designed in the present study to assess the influence of GPR176 expression on the PI3K/AKT/mTOR pathway and EMT. In comparison with the Lenti-NC group, the GPR176 upregulation group exhibited significantly larger tumor volume and weight. Conversely, relative to the sh-NC group, the sh-GPR176 group displayed a substantial reduction in tumor volume and weight (Fig. 9A). Subsequent examination of the mRNA expression levels of EMT pathway molecules, consistent with the in vitro results, revealed activation of the EMT pathway following GPR176 upregulation, while the EMT signaling pathway was inhibited after GPR176 downregulation (Fig. 9B). Western blot analysis suggested the activation of the EMT and PI3K/AKT/mTOR pathways following GPR176 upregulation, as indicated by increased levels of key pathway markers. Conversely, GPR176 downregulation resulted in changes indicative of pathway inhibition. (Fig. 9C), with the corresponding bar chart shown in Fig. S4A-G.