Fam198b as a novel biomarker for gastric cancer and a potential therapeutic target to prevent tumor cell proliferation dysregulation

Highlights

• •Increased Fam198b expression in GC is associated with poor prognosis.
• •The function of Fam198b in gastric cancer was testified by bioinformatics and in vitro experiments.
• •Fam198b promoted the proliferation, migration and invasion of gastric cancer cell lines.
• •Fam198b regulates Bcl-2 expression through the classical PI3K/AKT pathway to promote GC progression.
• •Fam198b is a novel biomarker for gastric cancer and a potential therapeutic target to prevent dysregulated tumor cell proliferation.

Abstract

Background

It has been reported that the human family with sequence similarity 198, member B (Fam198b) play an important role in the occurrence and development of various cancers. Nevertheless, its function in gastric cancer is not completely clear. Hereby, we investigated the function and prognostic value of Fam198b in gastric cancer and further validated the results in gastric cancer through a series of in vitro experiments.

Methods

We used R software and online bioinformatics analysis tools-GEPIA2, TIMER2, Kaplan-Meier plotter, cBioPortal, TISIDB COSMIC, and STRING to study the characteristics and functions of Fam198b in GC, such as aberrant expression, prognostic value, genomic alterations, immune microenvironment, anticancer drug sensitivity, and related signaling pathways. In addition, in vitro experiments such as immunohistochemistry (IHC), cell function experiments, and signaling pathway experiments were performed to validate the key conclusions.

Result

Fam198b is obviously highly expressed in gastric cancer, and its expression is intensively correlated with tumor prognosis. The etiology of abnormal Fam198b expression was superficially investigated and validated by associating genomic alterations and the immune microenvironment. Furthermore, Fam198b is intensively correlated with the sensitivity of multiple antitumor drugs. It was demonstrated by functional enrichment analysis that Fam198b was linked to myogenesis, angiogenesis, epithelial mesenchymal transition and cytokine binding. It was observed in vitro experiments that knockdown Fam198b could significantly inhibit tumor cell proliferation and migration. These results were reversed when Fam198b was overexpressed. It was validated by signaling pathway experiments that Fam198b promoted gastric cancer progression by up-regulating the PI3K/AKT/BCL-2 signaling pathway.

Conclusion

As a novel biomarker to predict GC prognosis and tumor progression, Fam198b is a promising therapeutic target to reverse tumor progression.

Introduction

Gastric cancer (GC) is one of the most invasive and the fourth most common cause of cancer-related death in the world, despite it is the fifth most common type of cancer [1]. Cancer statistics revealed that there were over 1 million gastric cancer cases and over 768,000 gastric cancer related deaths worldwide in 2020 [2]. It is reported that more than 95 % of gastric cancers pathological types are adenocarcinoma. Although the 5 year survival rate of early gastric cancer was excessive to 90 %[3]. In the fact, most of patients are diagnosed as comparatively advanced cancer rather than early stage due to occult onset and rapid progression of GC. Despite advances have taken place in the surgical treatment and adjuvant chemotherapy and/or chemoradiotherapy of GC over the past few decades, the 5 year survival rate of patients with stage II and advanced stages still worsen rapidly [4]. Therefore, there is an urgency to identify a novel biomarkers for early diagnosis or targeted therapy of GC.

The human family with sequence similarity 198, member B (Fam198b); also known as GASK1B (Golgi Associated Kinase 1B); a gene whose function is uncomplete explored, is predicted to be a membrane-bound glycoprotein localized on Golgi apparatus[5,6], regulated by the upstream FGF receptor signaling pathway[7], probably related to resistance to PAC[8], CIS as well as TOP[9]. Based on the previous limited studies, Fam198b is regarded as an important role in the regulation of cancer progression. In lung adenocarcinoma, high expressed Fam198b is associated with an indication to good prognosis[10]. Meanwhile it has been reported that the knockdown Fam198b promotes proliferation and metastasis in ovarian cancer [11]. However, Fam198b acts as a precancer gene in colorectal cancer progression, increased Fam198b expression in macrophages was associated with an indication to poor prognosis[12]. Currently, the role of Fam198b in GC has not been elucidated, it is necessary to explore the exact biochemical mechanism of Fam198b in gastric cancer.

We screened out gastric cancer intensively related biomarkers after preliminary bioinformatics analysis. It was found that Fam198b was overexpressed in GC compared to adjacent normal tissues. Instructed by this finding, we performed further analysis and necessary experiments to explore its characteristics and prognostic value in GC. In conclusion, this study aimed to confirm the potential capability of GC as a novel biomarker and therapeutic target, while providing new insights into the underlying molecular mechanisms of GC.

Methods and materials

Bioinformatics

Data from public databases and identification of the common DEG

Transcriptomic data of gastric cancer (GC) and normal samples were obtained from GSE118916, GSE112369 and GSE79973 datasets at gene expression omnibus(GEO)(https://www.ncbi.nlm.nih.gov/geo/)[13].The R package "limma" was used to identify the significant DEG. In this study, the genes meeting the criteria outlined in (Supplementary Table 1) were considered as DEGs. We identified the intersection of the three datasets as the best DEG by using the “venn” script.

Gene expression arrays and clinical information of GC patients were obtained from the TCGA database (http://xena.ucsc.edu/)[14].

WGCNA analysis

In order to identify GC related DEGs, we utilized the GSE79973 data set to develop a weighted gene co-expression network analysis (WGCNA) model with the "WGCNA" script. The correlation between each gene module and the clinical features was calculated. The p value < 0.05 was considered statistically significant. We identified modules with | correlation coefficient |> 0.6 as GC correlation modules. Finally, the 'venn' script was employed to overlap DEGs and GC associated modules to capture their potential common genes for further investigation.

Machine learning and hub-gene screening

To identify hub-gene of GC, we analyzed DEGs using three different machine learning algorithms-SVM-REF, randomForest and LASSO. Ultimately, common genes identified by all three machine learning algorithms were considered hub genes for GC.

Verify the selected hub-gene on an external database

The identified hub gene was differentially analyzed using the GEPIA(//gepia.cancer-pku.c) online tool[15], and survival analysis was conducted using Kaplan Meier plotter(https://kmplot.com/analysis/) to examine the prognostic value of hub genes in GC patients[16]. Finally, hub genes were selected based on their integrated highest degree of LogFC, prognostic value of survival, and novelty. In addition, we performed a whole-cancer analysis of Fam198b expression using the tumor immune estimation resource 2.0 (TIMER) database (http://timer.cistrome.org/)[17].

Diagnosis and prognosis value of Fam198b

The "pROC" and "timeROC" packages were utilized to generate receiver operating characteristic (ROC) curves to assess the diagnostic efficacy and prognostic value of Fam198b in gastric cancer. Cox regression models and nomogram plots were constructed using the "survival" and "rms" packages to analyze independent risk indicators for gastric cancer and predict 1-, 3-, and 5-year overall survival (OS) in GC patients.

Analysis of genetic variation

The genetic mutation of Fam198b and its correlation with the survival of gastric cancer patients were studied using the cBioPortal database(https://www.cbioportal.org/)[18]. In addition, the distribution of genetic changes of Fam198b was explored through The Catalogue of Somatic Mutations online database (http://www.sanger.ac.uk/cosmic/)[19].

Gene-set enrichment analysis and drug treatment

We retrieved Hallmark gene sets from the MSigDB gene database and categorized samples into high and low groups based on the median expression levels of Fam198b. Upregulated and downregulated genes were identified by comparing the differences in gene level between the two groups. The gene set with | NES |> 1 and FDR <0.05 were considered significant.

Drug sensitivity data was obtained from the CellMiner online database (https://discover.nci.nih.gov/cellminer/home.do), and then the expression data of Fam198b and drug sensitivity data were integrated for the Pearson correlation test to determine the correlation.

Functional analysis of related genes and protein-protein interaction (PPI) analysis

The top 100 genes most associated to Fam198b were obtained using the GEPIA, then the "clusterProfiler" and " org.Hs.eg. The db " packages were utilized to perform the functional analysis in terms of gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG). GO categorizes genes based on molecular function (MF), biological process (BP), and cellular components (CC) to describe the function of gene products. KEGG is a collection of databases for systematically analyzing gene function and associating related gene sets with specific pathways. Furthermore, we conducted protein-protein interaction of Fam198b analysis using the String database (https://string-db.org/) [20].

In vitro experiments

Immunohistochemistry (IHC)

For this study, 80 pairs of paraffin tissues, collected from Northern Jiangsu People's Hospital(Clinical Medical College of Yangzhou University, Yangzhou, China) between July 2015 and May 2017, were used as clinical specimens for the experiments. This study was approved by the Ethics Committee of Northern Jiangsu People's Hospital(Ethics number 2019 KY-022).

Tissue microarrays (TMA) were generated as previous mentioned methods, including tissues from 80 histologically proven gastric cancer patients and 80 matched normal tissue samples [21]. Tissue sections were dried at 60 °C for 2 h. Paraffin was removed with xylene and samples were hydrated in a graded alcohol series and citrate buffer, then blocked with 3 % hydrogen peroxide. Subsequently, the sections were incubated overnight with primary antibody to Fam198b (Proteintech; diluted in a ratio of 1:50) at 4 °C, followed by HRP-labeled goat anti-rabbit IgG (1:100; catalog number ab288151; Abcam) for 30 min at room temperature, and the slides were counterstained by DAB staining with hematoxylin. Five high-power fields (400 x magnification) were randomly selected and each slide was photographed. The cytoplasmic expression score was calculated by the staining intensity plus the percentage of positive cells. The staining intensity score is as follows: colorless (0), yellow (1), brown (2), dark brown (3). Based on the percentage of positive cells, scores were assigned as follows: 0 points for 0–5 %, 1 point for 5–25 %, 2 points for 25–50 %, 3 points for 50–75 %, and 4 points for >75 %. Fam198b expression levels were classified as low or high based on a total score of 5.

Cell culture

The normal gastric epithelial cell lines GES-1 and the GC cell lines HGC-27, AGS, and NCI-N87 were purchased from the Cell Bank of the Chinese Academy of Sciences. Cell lines were maintained in RPMI-1640 supplemented with 10 % fetal calf serum (Invitrogen, Carlsbad, CA, USA), 100 U/mL penicillin and 100 μg/ml streptomycin (SV30010, Hyclone, Cytiva) at 37 °C in a 5 % CO2 atmosphere.

RT-qPCR analysis

Total RNA were extracted from cell lysates using TRIzol reagent (Invitrogen Corporation, Beijing, China). Following the manufacturer's instructions, the cDNA was synthesized from 1.0 μg of total RNA in 20μl reaction mixture using a kit, then RT-qPCR was performed with kit in the StepOnePlus Real-time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc). Primer designs for Fam198b and GAPDH are provided in Supplementary Table 2. The results of the real-time qPCR experiments were calculated using the 2−ΔΔCt method. Experiments were conducted in triplicate for each reaction.

Western blot analysis

Proteins were extracted from normal gastric epithelial cells and GC cells using RIPA lysis buffer. Total protein was quantified using a BCA kit, then separated on a 10 % SDS-PAGE gel and electrotransferred onto a polyvinylidene difluoride (PVDF) membrane (Thermo Fisher Scientific, Waltham, MA, USA). After blocking with 5 % BSA for 1.5 h, they were incubated overnight with various primary antibodies in a 4 °Cshaker, including anti-Fam198b (Proteintech), anti-PI3K, anti-p-PI3K (Tyr467/199), anti-Akt, anti-p-Akt (Thr308), anti-Cyclind-D1, anti-Bcl-2, anti-Bad and anti-Bax (all purchased from Abmart). After the membrane was incubated with the HRP-conjugated secondary antibody for 1.5 h at room temperature, the bands were detected using the enhanced chemiluminescence detection reagent (Applygen Technologies Inc, Beijing, China). Image J software was used to measure the intensity of the bands on the protein blot. Experiments were performed in triplicate for each reaction.

Cell transfection

To manipulate Fam198b expression, we employed small interfering RNA (siRNA) to knock down Fam198b in NCI-N87 and AGS cells, while Fam198b was overexpressed using a plasmid in the HGC-27 cell line. Gastric cancer cells were seeded into 6-well plates at a concentration of 15 × 10^5 cells/well and allowed to reach 75 % confluence before transfection. The siRNA and pcDNA3.1-Fam198b plasmid were transfected with Lipofectamine® 2000 (Thermo Fisher, Waltham, MA, USA) according to the manufacturer's instructions. Experiments were conducted 48 h after transfection, and transfection efficiency was validated through RT-qPCR and western blotting. Sequences used are shown in Supplementary Table 3.

Cell counting kit-8 (CCK-8) and colony formation assays

To assess cell viability, 1000 cells per well were seeded into 96-well plates and cultured overnight at 37 °C in a 5 % CO2 environment. At 24, 48, 72, and 96 h, 10 μL of Cell Counting Kit-8 (CCK-8) solution (Beyotime Institute of Biotechnology, Shanghai, China) was added to each well and incubated for 2 h. Absorbance was measured at 450 nm using a microplate reader. Furthermore, a colony formation assay was performed to assess cell proliferative capacity. A total of 500 GC cells per well were seeded into 6-well plates and cultured for 10 days at 37 °C in a 5 % CO2 atmosphere. Colonies were fixed with 4 % paraformaldehyde for 10 min and then stained with 0.1 % crystal violet for 7 min. The number of colonies (> 50 cells) was manually counted under the microscope. Each experiment was performed independently in triplicate.

Wound healing assay

NCI-N87, AGS, and HGC-27 cells were seeded into six-well plates and cultured in medium without FBS. When the monolayer cells reached 80–90 % confluence, a 200- µl pipette tip was used to create a vertical wound at the bottom of the plate. The floating cells were then washed away with PBS. Next, the medium containing 3 % FBS was replaced to culture the remaining cells. Photographs were taken at 0 h and 24 h post-scratching. Wound closure condition was evaluated using the ImageJ software.

Cell migration and invasion assays

Migration and invasion assays were performed using 24-well plates with 8 µm pore transwell chambers (Corning, Inc). For migration assays, transfected cells were starved for 12 h, then 2 × 10 [5] cells were suspended in RPMI-1640 medium without FBS and appended to the upper compartment. Simultaneously, RPMI-1640 medium containing 10 % FBS was placed in the lower compartment. In the invasion assay, the Matrigel matrix was diluted at a 1:9 ratio and coated onto the upper side of the bottom membrane of the transwell chamber before proceeding as in the migration assay. After incubating the 24 wells in a 37 °C incubator for 48 h, the 24-well plates were taken out. Cells that migrated to the lower surface of the filter membrane were fixed with 4 % paraformaldehyde and stained with 1 % crystal violet. Remaining cells on the upper surface of the filter membrane were gently removed with a cotton swab. Images were captured under a microscope, and cell numbers were quantified using ImageJ software.

Statistical analysis

The statistical analysis results were visualized using R 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism 8 (GraphPad 8.0.1 Software, La Jolla, CA, USA), using T-test, chi-square test, Kaplan-Meier method and log-rank test. All data were expressed as at least three independent experiments; * p <0.05, * * p <0.01, * * * p <0.001 were considered as significant.

Results

Hub-genes screening

For datasets GSE118916, GSE112369 and GSE79973,52 common DEG (Fig. 1A) were screened out by "limma" and "venn" packages of R . In order to explore the key DEG in GC development, dataset GSE79973 constructed a gene co-expression network and calculated the correlation between clinical features (Fig. 1B) and modules. We selected the module gene set with | correlation coefficient |> 0.5 to overlap with the common DEG to obtain 34 common genes (Fig. 1C). To further identify the hub-gene of GC, we screened these 34 genes out using three machine learning algorithms to select their common genes based on the top 20 genes of SVM-REF, the top 20 genes of randomForest and the results of LASSO (Fig. 1D-G).

Fig. 1.

Screening for hub genes in gastric cancer and its differential analysis and prognosis. Identifying common DEGs using the GEO dataset(A). WGCNA analysis identify GC-related modules(B). Common genes of DEGs and GC-related modules(C). Machine learning and hub-gene screening, SVM-REF analysis(D), randomForest analysis(E), LASSO analysis(F). hub-genes(G). Differential analysis and prognosis, Fam198b(H), CLIC6(I), LOXL1(J). Differential analysis of Fam198b in pan-cancers(K).

External database validation for hub-genes expression and prognosis

To validate the expression of the selected genes and the relationship with overall survival in STAD patients, we used GEPIA and Kaplan Meier plotter containing TCGA and GTEx GC samples for analysis. The correlation between the expression level of hub-genes and the overall survival of STAD patients is shown in Fig. 1H-J. The results showed significant discrepancy in expression levels of Fam198b, CLIC 6 and LOXL1 in STAD samples compared to normal tissues. Moreover, the upregulation of Fam198b and LOXL1 was associated with worse OS in STAD patients. Some studies have shown that LOXL1 is correlated with the development of GC[22]. While the role of Fam198b in GC remains unclear, we chose Fam198b for further investigation. In TIMER2 pan-cancer data we found that Fam198b was upregulated in CHOL, HNSC, KICH, KIRC, LIHC, PRAD and STAD and downregulated in BLCA, BRCA, CESC, KIRP, LUSC, PCPG, READ, SKCM and UCEC (all P <0.05)(Fig. 1K),and the specific information of these cancer types were described in (Supplementary Table 4).

Genetic alterations and mutation prognosis of Fam198b

We investigated Fam198b gene mutations and alterations using cBioPortal. The highest alteration frequency was observed in STAD patients (5 %), with "Mutation" and "amplification" being the most dominant alterations (Fig. 2A-B). Additionally, we explored specific mutation types and sites (Fig. 2C) of Fam198b in cancer. The results indicated that Fam198b alterations included nonsense substitutions, missense substitutions, synonymous substitutions, frameshift insertions, inframe deletions, and frameshift deletions, with missense substitutions being the most common (33.79 %). Nucleotide changes were observed for multiple mutations, with G > A and C > T being the most dominant (26.94 %; 24.92 %) (Fig. 2D). We also investigated the effect of Fam198b gene mutations on survival using the cBioportal database. Survival curves indicated that genetic mutations in Fam198b did not significantly affect overall survival (p = 0.168) or disease-free survival (p = 0.111) in STAD patients (Fig. 2E-F).

Fig. 2.

Genetic and epigenetic alterations of Fam198b.Genetic mutations in Fam198b(A).Fam198b mutation in pan-cancers(B). Site of Fam198b mutation in pan-cancers (C). Types of Fam198b mutation in pan-cancers (D). Correlation of Fam198b gene mutation with OS (E) and DFS (F).

Fam198b associates with tumor immune microenvironment

To gain insight into the mechanism of tumor immune infiltration, we used the TISIDB database to explore the correlation between lymphocytic infiltration, immunosuppressants, immunostimsants and Fam198b expression in the GC, and presented four main results for each section. It indicates a positive correlation between Fam198b and the infiltration of major immune cells, including Macrophage (Spearman:rho=0.5,p<2.2E-16),Treg(Spearman:rho=0.464,p<2.2E-16), NK(Spearman:rho=0.417, p<2.2E-16), Tfh(Spearman:rho=0.41, p<2.2E-16) (Fig. 3A-B).Besides, Fam198b expression was positively correlated with multiple immunosuppressant agents. The four immunostimulatory agents with a strong correlation were CSF1R (Spearman:rho=0.429, p<2.2E-16), KDR (Spearman:rho=0.405,p<2.2E-16), PDCD1LG2 (Spearman:rho=0.383, p = 3.49E-16) and TGFBR1 (Spearman:rho=0.383, p = 4.06E-16) (Fig. 3C-D). The four immunostimulants were ENTPD1 (Spearman: rho=0.647, p<2.2E-16), CXCL12(Spearman:rho=0.54,p<2.2E-16), CXCR4(Spearman: rho=0.366, And p <1.55E 14) and CD28 (Spearman: rho=0.355, p = 1.18E − 13) (Fig. 3E-F). The above results indicated that Fam198b could modulate multiple immune components through various pathways, affecting tumor immune infiltration in GC.

Fig. 3.

Correlation analysis of Fam198b levels and immunity. Relationship with tumor-infiltrating lymphocytes in cancer(A), and the top 4 correlations in GC(B). Relationships with immune inhibitors in cancer(C), and the top 4 correlations(D). Relationship with stimulants in cancer(E), and the top 4 correlations in GC(F).

Gene set enrichment analysis and drug efficacy analysis

To further explore the significantly altered molecular pathways of Fam198b^High and Fam198b^Low in STAD, we performed a GSEA analysis based on the Hallmarks gene set. The results showed that Fam198b mainly positively regulated signaling pathways, including UV_RESPONSE_DN, PANCREAS_BETA_CELLS, MYOGENESIS, ANGIOGENESIS, and EPITHELIAL_MESENCHYMAL_TRANSITION (Fig. 4A), which have been reported to be involved in tumor proliferation and progression[23], [24], [25].

Fig. 4.

Gene set enrichment analysis and drug efficacy analysis of Fam198b. Gene set enrichment analysis of Fam198b(A). Drug efficacy analysis of Fam198b(B).

In the CellMiner database, we examined the correlation between Fam198b gene expression and drug sensitivity. Results showed a positive correlation of Fam198b with Okadaic acid (Cor = 0.397, p = 0.002), AZD-5363 (Cor = 0.391, p = 0.002), Kahalide F (Cor = 0.333, p = 0.009) and a negative correlation with Allopurinol (Cor = −0.383, p = 0.002) (Fig. 4B). This information could aid in optimizing the individualized treatment of STAD patients with high Fam198b expression.

Experiments verification

Fam198b increased in Gc and associated with poor prognosis in Gc patients

To examine the expression of Fam198b in gastric cancer (GC), we employed immunohistochemistry (IHC) to assess Fam198b levels in 80 pairs of GC and adjacent tissue samples. The results demonstrated significantly higher Fam198b expression in GC tissues compared to adjacent tissues (Fig. 5A-B). Furthermore, both Western blot and RT-qPCR analyses revealed significantly elevated Fam198b expression in GC cell lines in comparison to GES-1 cells(Fig. 5C-D).

Fig. 5.

Fam198b is highly expressed in gastric cancer tissues and gastric cancer cell lines and has a poor prognosis. Immunohistochemical analysis of Fam198b expression in tissue arrays of 80 GC patients. Representative images were shown(A). Paired t-test showed immunohistochemical results(B). Western blot (C) and Reverse transcription-quantitative PCR (D) showed the expression of Fam198b in human gastric mucosal epithelial cells (GES-1) and gastric cancer cell lines. Survival analysis of gastric cancer patients with different expression levels of Fam198b(E).

Moreover, we investigated the association between Fam198b expression and clinicopathological features. The analysis indicated that increased Fam198b expression was associated with parameters such as tumor size, Lauren type, T stage, pathological stage, degree of differentiation, vascular invasion, and nerve invasion (p < 0.05) (Table 1). To further investigate the connection between Fam198b expression and GC prognosis, we evaluated overall survival (OS) among 80 GC patients. Kaplan-Meier curve analysis revealed a positive correlation between elevated Fam198b protein expression and poorer survival outcomes in GC patients (P = 0.0017; Fig. 5E).

Table 1.

Relationship between Fam198b expression and clinicopathological features in 80 patients with gastric cancer.

		Expression level of Fam198b
Histopathological parameters	Total number (80)	High	Low	P-value
Age, years				0.975
<65	25	16	9
>=65	55	35	20
Sex				0.204
Male	66	40	26
Female	14	11	3
Tumor size				0.002
<5	29	20	22
>=5	51	31	7
Lauren type				0.021
Intestinal type	62	36	26
Diffuse type	17	15	2
T stage				0.036
T1, T2	12	10	10
T3, T4	68	41	19
N stage				0.624
N0, N1	31	20	13
N2, N3	49	31	16
M stage				0.447
M0	74	48	26
M1	6	3	3
Pathologic stage				0.038
I-II	20	13	14
III-IV	60	38	15
Histological grade				0.048
G1-G2	25	12	13
G3	55	39	16
Venous invasion				0.001
No	41	18	22
Yes	39	33	7
Nerve invasion				0.024
No	38	20	19
Yes	42	31	10

Furthermore, we conducted univariate and multivariate analyses to identify risk factors for GC. Univariate regression analysis identified factors affecting OS, including Fam198b expression, tumor size, T stage, pathological stage, vascular invasion, and nerve invasion. Multivariate analysis demonstrated that Fam198b expression, tumor size, T stage, pathological stage, and nerve invasion were independent risk factors for GC progression (Table 2).In order to further validate the diagnostic and prognostic value of Fam198b in gastric cancer, we constructed the receiver operating curve (ROC) and Nomogram using the STAD data of TCGA database, and the AUC was 0.793(0.754–0.832), indicating that Fam198b has good value in the diagnosis of STAD (Fig. 6A). Univariate cox regression analysis showed that Fam198b level, age, T stage, and pathological stage were all significantly associated with OS (all p <0.05). In the multivariate cox regression analysis, Fam198b expression (p = 0.025) was also significantly associated with age (p = 0.024), indicating that Fam198b and age were independent prognostic factors for GC patients (Fig. 6B-C).

Table 2.

Prognostic factors for overall survival of 80 gastric cancer patients were analyzed by univariate and multivariate Cox proportional hazards models.

	Univariate analysis		Multivariate analysis
	HR (95 % CI)	P-value	HR (95 % CI)	P-value
Age	0.651(0.285–1.488)	0.309
Sex	0.591(0.237–1.476)	0.260
Tumor size	3.658(1.583–8.449)	0.002	3.088(1.210–7.878)	0.018
Lauren type	1.258(0.543–2.918)	0.593
T stage	5.349(1.570–18.23)	0.007	5.372(1.459–19.78)	0.011
N stage	1.518(0.663–3.474)	0.323
M stage	0.502(0.067–3.746)	0.502
Pathologic stage	4.462(1.605–12.41)	0.004	3.774(1.037–13.74)	0.044
Histological grade	1.520(0.624–3.700)	0.356
Venous invasion	2.451(1.089–5.518)	0.030	0.593(0.215–1.633)	0.312
Nerve invasion	3.359(1.426–7.910)	0.006	2.848(1.039–7.807)	0.042
Fam198b	4.745(1.619–13.90)	0.005	3.623(1.118–11.75)	0.032

Fig. 6.

Diagnostic and prognostic efficacy of Fam198b in GC. Diagnostic ROC analysis with the AUC of Fam198b in GC(A). Univariate regression analysis and multivariate survival methods (overall survival) for prognostic covariates of gastric cancer patients based on TCGA data(B-C). A nomogram for predicting the probability of 1-, 3-, and 5-year OS for GC patients(D). Calibration plots of the nomogram for predicting the probability of OS at 1, 3, and 5 years(E). Prognostic efficacy of Fam198b in GC(F).

Building on these multivariate results, we integrated T, N, M stage, and pathological stage into a Nomogram to better elucidate the clinical relevance of Fam198b. Simultaneously, we generated a calibration curve to assess the Nomogram's performance. The calibration curve shows that the Nomogram functions well in predicting OS and actual OS at 1,3, and 5 years (Fig. 6D-E).Time-dependent ROC analysis indicated that Fam198b had good predictive capacity for OS in GC patients, with AUC of 0.601,0.599,5 and 0.725 (Fig. 6F), respectively.

Fam198b promotes the proliferation of NCI-N87, AGS, and HGC-27 gastric cancer cells

Based on the comprehensive findings discussed earlier, Fam198b emerges as a key driver of tumorigenesis and progression in stomach adenocarcinoma (STAD). To substantiate this, we selected NCI-N87 and AGS cell lines for investigating the effects of reduced Fam198b expression, and HGC-27 cell lines for examining the consequences of Fam198b overexpression. Specific siRNA and overexpression vectors were utilized to reduce Fam198b levels in NCI-N87 and AGS cells, and elevate Fam198b expression in HGC-27 cells. The efficacy of Fam198b knockdown and overexpression was confirmed through Western blot and RT-qPCR analyses (Fig. 7A-C).

Fig. 7.

Fam198b knockdown strongly inhibited the proliferation of GC cells, and the results were reversed upon overexpression. The expression of Fam198b was significantly reduced in cells transfected with knockdown Fam198b compared to cells transfected with si-Ctrol(A-B). Compared with cells transfected with the empty vector, the expression of Fam198b was significantly increased in cells transfected with overexpression of Fam198b(C). Cell Counting Kit-8 was used to detect the proliferation of NCI-N87 and AGS cells after Fam198b knockdown(D-E) and HGC-27 cells after Fam198b overexpression(F). Colony formation assays were used to detect the proliferation of NCI-N87 and AGS cells after Fam198b knockdown(G-H) and HGC-27 cells after Fam198b overexpression(I).

To confirm whether Fam198b could effect on the proliferation of NCI-N87, AGS, and HGC-27 cells, CCK 8 and colony formation assays were performed. In CCK 8 experiments, the proliferation rate of NCI-N87 and AGS cells treated with si-Fam198b was significantly lower than that in the siCtrl group (Fig. 7D-E). After Fam198b overexpression, HGC-27 cells showed more proliferative capacity than the Vector group (Fig. 7F). Likewise, the number of colonies formed by si-Fam198b-treated cells was significantly reduced compared to the siCtrl group (Fig. 7G-H). In contrast, HGC-27 cells overexpressing Fam198b demonstrated increased colony formation (Fig. 7I).

Fam198b upregulation promotes the migration and invasion of GC cells

To investigate the influence of Fam198b on the migration and invasion of GC cells, transwell assay and wound healing assay were performed. Results from the transwell migration assay and wound healing assay indicated significantly reduced migration in the si-Fam198b group in NCI-N87 and AGS cells compared to the si-Ctrl group. Moreover, the transwell assay with Matrigel showed decreased invasive ability in the si-Fam198b group (Fig. 8A-B). Moreover, we evaluated the migration and invasion of HGC-27 cells after the transfection compared with the vector group. Migration and invasion of HGC-27 cells of overexpressed Fam198b increased(Fig. 8C).

Fig. 8.

Fam198b knockdown strongly inhibited the migration and invasion of GC cells, and the results were reversed upon overexpression. Transwell assays were used to detect the migration and invasion of NCI-N87 and AGS cells after Fam198b knockdown(A-B) and HGC-27 cells after Fam198b overexpression(C). The wound healing assay was used to detect the migration of NCI-N87 and AGS cells after Fam198b knockdown(D-E) and HGC-27 cells after Fam198b overexpression(F).

Fam198b regulation of GC progression by activating the PI3K / AKT / Bcl-2 axis

To further elucidate the specific mechanism of Fam198b regulating GC progression, we obtained the top 100 genes associated with Fam198b from the GEPIA2 tool for GO and KEGG analyses. It was indicated in GO analysis that Fam198b-related genes may be related to processes like extracellular matrix organization, endoplasmic reticulum lumen, growth factor binding, and cytokine binding, which were all involved in cell proliferation. The enriched pathways in the KEGG analysis results attracted our attention. Because PI3K-Akt signaling pathway has a well-recognized role in promoting the proliferation and metastasis of gastric cancer KEGG[26]. A protein-protein interaction (PPI) network was constructed using the STRING tool, highlighting the top 10 proteins interacting with Fam198b. Moreover, we conducted co-expression analysis by examining the top 10 genes from GEPIA2 that co-expressed with Fam198b in various cancers, as revealed through heat map visualization (Fig. 9D).

Fig. 9.

Enrichment analysis of Fam198b-related genes and Fam198b regulates GC progression by activating PI3K/Akt/Bcl-2 signaling pathway. GO/KEGG analysis of the top 100 genes associated with Fam198b(A-B). PPI analysis of Fam198b protein by STING tool(A). The expression heatmap of the top 10 out of 100 genes in cancer using TIMER2 tool(D). The protein levels of p-PI3K, PI3K, p-Akt, Akt, Cyclin-D1, Bcl-2, Bad, and Bax in NCI-N87 were analyzed by western blot after dealing with si-Ctrl and si-Fam198b together(E, H) while after dealing with Vector and OE-Fam198b together(F, I). Fam198 regulation of PI3K/Akt signaling pathway was shown(G).

To validate the regulatory relationship between Fam198b and the PI3K/Akt pathway, along with apoptosis-related Bcl-2 family proteins, we performed Western blotting. Fam198b knockdown with siRNA in NCI-N87 cells could down-regulate the protein levels of phospho-PI3K (p-PI3K), phospho-Akt (p-Akt), Cyclin-d1 and Bcl-2 proteins, and upregulated the protein levels of Bad and Bax (Fig. 9E, H). Conversely, Fam198b overexpression in HGC-27 cells led to the opposite effects, upregulating p-PI3K, p-Akt, Cyclin D1, and Bcl-2, while reducing Bad and Bax (Fig. 9F, I). The results indicated that Fam198b may contribute to gastric cancer development to some extent partly through the activation of the PI3K / Akt / Bcl-2 signaling pathway (Fig. 9G).

Discussion

Gastric cancer (GC) stands out as one of the most aggressive malignancies. Despite advancements in cancer research and treatment, the five-year survival rate for advanced GC remains distressingly low, imposing a substantial socioeconomic burden on society [27]. Consequently, there is an urgent need for further investigation into effective early diagnostic biomarkers and therapeutic targets in GC.

In this study, the research team utilized weighted gene co-expression network analysis (WGCNA) and machine learning techniques to pinpoint Fam198b as a pivotal biomarker for GC [28,29]. Fam198b, a gene associated with cancer, exhibits a dual role in various cancer types. Previous studies have indicated that Fam198b functions as a tumor suppressor gene, as it is downregulated in conditions such as lung adenocarcinoma and ovarian cancer. Fam198b has been shown to inhibit lung adenocarcinoma progression by blocking the ERK-mediated MMP-1 expression pathway[10]. Additionally, Fam198b is involved in the CELF2/Fam198b axis, suppressing ovarian cancer proliferation and metastasis[11]. In contrast, Fam198b is a precancer gene in colorectal cancer. It promotes colorectal cancer progression by targeting SMAD2 to regulate M2 polarization in macrophages. Despite these seemingly contradictory effects of Fam198b in cancer, these studies collectively underscore the significant role of Fam198b in cancer progression. Thus, the research team embarked on a study to further elucidate its role in gastric cancer.

Utilizing data from the cancer genome atlas (TCGA) database, the researchers identified Fam198b as being overexpressed in various tumors, including GC, with statistical significance (p < 0.05). Subsequently, immunohistochemistry (IHC) experiments confirmed that Fam198b was markedly upregulated in gastric cancer tissues when compared to adjacent tissues. Furthermore, the expression level of Fam198b was found to be associated with several adverse prognostic risk factors, including tumor size, Lauren type, T stage, pathological stage, degree of differentiation, vascular invasion, and nerve invasion.

To assess the prognostic value of Fam198b, survival analysis was performed. It was demonstrated in by Kaplan Meier plotter that GC patients with high Fam198b expression tended to have a shorter OS (HR = 1.55,95 %CI 1.24–1.94, P <0.001). This conclusion was further confirmed by the survival analysis results of the tissue microarray. Furthermore, Cox regression analysis based on TCGA data indicated that Fam198b was a certain risk factor for GC. Similarly, Cox regression analysis of tissue microarray also confirmed that Fam198b were GC-independent prognostic risk factors like tumor size, T stage, pathological stage, and nerve invasion. In addition, Fam198b demonstrated good efficacy in the diagnostic ROC and time-dependent ROC. These results collectively suggest that Fam198b is not just an overexpressed gene but also a critical risk factor influencing the prognosis of GC patients, underscoring its pivotal role in the development of gastric cancer.

Given the findings above, we conducted a series of analyses to investigate the downstream mechanisms of its carcinogenic and risk effects.

It was reported that mutations in protein-encoding genes induced expression changes in cancer[30], We therefore explored the association of Fam198b in genetic alterations. Despite the genetic alterations were observed in Fam198b and showing the highest level in STAD patients, it does not affect patient outcomes.

The tumor immune microenvironment consists of infiltrating immune cells, chemokines, and cytokines, and is an important regulator of tumorigenesis, invasion, and metastasis[31,32]. We performed an immune infiltration analysis of the common types of immune cells. IL-17, the signature cytokine of Thl7 cells, is a proinflammatory factor, and Treg reduces tissue damage in vivo by regulating the immune response and releasing an anti-inflammatory cytokine (IL-10)[33].The imbalance between proinflammatory Th17 cells and inhibitory Treg cells is a key factor in inflammation, autoimmune diseases, and cancer development. [34]. This association suggests that increased Fam198b expression may promote the development and evolution of GC cells by affecting immune cell composition in the tumor microenvironment.

The study explored various treatment options for cancer, including surgery, hormone therapy, gene therapy, immunotherapy, radiotherapy, laser therapy, combination therapies, and targeted therapies. Chemotherapy is the most common and promising method of cancer treatment[35,36]. However, drug resistance remains a significant challenge in cancer treatment, limiting the potential for a cure in many cancer patients [37].. The researchers investigated the potential correlation between drug sensitivity and Fam198b expression using the Cellminer database. The results show that the expression of Fam198b is positively correlated with the sensitivity of many classical antitumor drugs, for example, Okadaic acid, AZD-5363, Kahalide F. This suggests that Fam198b could potentially serve as a predictor of tumor response to chemotherapy.

To further understand the role of Fam198b in gastric cancer (GC), we performed a series of cellular experiments. In vitro experiments demonstrated that knocking down Fam198b inhibited malignant behaviors in GC cell lines, including proliferation, migration, and invasion. Conversely, these behaviors were enhanced when Fam198b was overexpressed. This supports the idea that Fam198b could be a valuable prognostic indicator in GC. Mechanistically, the study suggested a potential oncogenic pathway for Fam198b in GC involving the PI3K/Akt signaling pathway on the KEGG database. The PI3K / Akt signaling pathway, as a classical pathway of tumor regulation, promotes the growth, proliferation, and survival of cancer cells by enhancing the activity of nutrient transporters and metabolic enzymes[38,39]. The results indicated that Fam198b may regulate the expression of Bcl-2, a protein involved in apoptosis, through the PI3K/Akt pathway to promote GC progression.

Our study also has some shortcomings for some ineluctable reasons. Firstly, it only recorded overall survival (OS) in patients and did not track disease-free survival. Secondly, it did not investigate the effect of Fam198b on the cell cycle. Finally, more meticulous in vivo experiments also need to be conducted to confirm the current achieved findings and help to reveal novel characteristics for Fam198b role in GC, and this part is the key point we are going to make more exploration in the future.

Conclusion

Based on multiple authoritative databases, clinical samples and experimental evidence, this study revealed that Fam198b plays an important role in gastric cancer. Fam198b is a novel biomarker that can be utilized to predict the occurrence, prognosis, immune response and drug sensitivity of patients with gastric cancer, and has a promising therapeutic target to reverse tumor development.

Funding

This study was sponsored by Yangzhou City Science and Technology Project (Funding No.YZ2020159)

Ethics approval

It is approved by the Ethics Committee of Northern Jiangsu People's Hospital, Yangzhou, China (approval no. 2019KY-022).

Availability of data and materials

The bioinformatics datasets provided in this study are available in online repositories and the data used during the experiments are available from the corresponding authors upon reasonable request.

CRediT authorship contribution statement

Bangquan Chen: Conceptualization, Methodology, Software, Data curation, Writing – original draft, Validation. Maladho Tanta Diallo: Writing – review & editing, Visualization. Yue Ma: Visualization, Investigation. Wenhao Yu: Software, Investigation. Qing Yao: Data curation, Investigation. Shuyang Gao: Visualization, Investigation. Yantao Yu: Formal analysis, Investigation. Qiannan Sun: Data curation, Investigation. Yong Wang: Software, Investigation. Jun Ren: Conceptualization, Visualization, Investigation, Supervision. Daorong Wang: Supervision, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare no competing interests.