Abstract
Background/Aim: Clear cell renal cell carcinoma (ccRCC), the most prevalent form of kidney cancer, often presents or recurs as an advanced, aggressive, and lethal disease. Thus, biomarkers are needed to identify patients at risk of developing advanced-stage or treatment-resistant ccRCC. SYNJ2BP, a cytoplasmic scaffolding protein, regulates ACVR2 activity, a key mediator of signaling pathways involved in tumor progression and metastasis. This study aimed to ascertain if SYNJ2BP, a gene highly expressed in normal kidney tissue, may serve as a predictive biomarker for ccRCC.
Materials and Methods: Bioinformatic analysis and immunohistochemistry were applied to investigate the relationship between SYNJ2BP expression, the immune landscape, and survival outcomes in ccRCC. We utilized data from publicly available databases, including The Cancer Genome Atlas, Gene Set Cancer Analysis (TCGA), and various other databases.
Results: In-silico analyses revealed that SYNJ2BP expression was significantly down-regulated in ccRCC (Log2FC=0.40, p=2.65E-36; FDR=9.73E-34), compared to normal tissue. Moreover, SYNJ2BP expression was significantly reduced in advanced stages and grades (III and IV; p<0.001) compared to lower stages and grades (I and II). Decreased expression was associated with nodal invasion and metastasis (p<0.0001), unresponsive to treatment (p=0.0052), post-treatment recurrence (p=0.002), lower median overall survival (HR=0.39, 95% CI=0.28-0.54, p<0.0001), disease-specific survival (HR=0.16, 95% CI=0.09-0.27, p<0.0001) and shorter progression-free survival (HR=0.24, 95% CI=0.17-0.35, p<0.0001). Survival trends remained consistent in multivariate Cox regression, where expression remained independently associated with outcome. Consistent with transcript-level findings, immunohistochemistry demonstrated reduced protein expression of SYNJ2BP in ccRCC patients (p<0.05).
Conclusion: SYNJ2BP is a novel prognostic biomarker for ccRCC and the down-regulation of SYNJ2BP expression is associated with poor survival outcomes and reduced treatment response.
- SYNJ2BP
- ccRCC
- renal carcinoma
- phosphatidylinositol signaling
- ACVR2
- biomarkers
- treatment response
- mitochondria
Introduction
It is estimated that 80,980 new cases of kidney cancer will be diagnosed in 2025, and approximately 14,510 patients will die from their disease (1). Clear cell renal cell carcinoma (ccRCC) is the predominant form of kidney cancer and is considered highly curable when diagnosed as a localized disease. Notwithstanding, ccRCC can initially present or recur as a more aggressive form associated with progressed disease. While the treatment landscape for ccRCC has drastically changed over the past several years with the advancements in immuno-oncology and tyrosine kinase inhibitors [TKIs; (2, 3)], identification of biomarkers predictive of aggressive disease phenotypes is essential to improve diagnostic accuracy and for the discovery of adjuvant and novel therapies. Biomarker discovery in kidney cancer remains an active area in research, with multiple potential biomarker candidates that are focused on utilizing gene mutations, chromosomal copy number variants, immunotherapy markers, multigene signatures, and multi-omics subclassifications that are based on angiogenesis, immune, cell-cycle, metabolism, and stromal differences (4). Further, next-generation sequencing advancements enable comprehensive molecular profiling of tumor tissue to identify genetic changes that may drive cancer growth and spread. In a recent proteomics-based study, Park et al. identified the secretory factor serpin family A member 3 (SERPINA3) as a potential biomarker for ccRCC with increased expression, which correlated with both disease progression and poor survival outcomes (5). Moreover, a transcriptional repressor ZNF844 was found to be significantly down-regulated across all tumor stages and associated with worse overall survival outcomes (6).
Synaptojanin-2 binding protein (SYNJ2BP), also known as ARIP2 and OMP25, is a PDZ (PSD-95/Dlg/ZO-1) domain cytoplasmic scaffolding protein responsible for organizing proteins into functional complexes. SYNJ2BP has been found to localize to the mitochondria and the plasma membrane (7). Localization of SYNJ2BP to the mitochondria provides a binding partner for RRBP1, located on the rough endoplasmic reticulum (ER), and is important for mitochondrial DNA (mtDNA) replication. Mitochondrial DNA is organized into nucleoids localized near the mitochondrial-ER contact sites, enabling a close spatial organization within the ER that allows mtDNA replication to occur (8-10). It has been shown that knockdown of the SYNJ2BP gene does not allow for proper nucleoid formation in the cell and rather, results in abnormal mtDNA aggregates called mitobulbs, resulting in a decrease and enlargement of nucleoids (11). This suggests that SYNJ2BP plays a key role in mitochondrial nucleoid maintenance (11). The formation of mitobulbs still allows for active mtDNA replication. Notwithstanding, this results in higher levels of reactive oxygen species (ROS), which may lead to oxidative damage and mitochondria dysfunction (12). Therefore, aberrations in SYNJ2BP gene and protein expression may result in mitochondrial dysfunction, which has been linked to cancer progression through chemotherapy resistance, apoptosis resistance, and immune suppression (13, 14). In addition, SYNJ2BP is involved in the regulation of activin type II receptors (ActRII or ACVR2), primary ligand binding receptors that are constitutively active serine/threonine kinases, leading to the transcription of selected genes via intracellular SMAD signaling (15, 16). SYNJ2BP enhances endocytosis of ACVR2, thereby suppressing activin-induced transcription (15). Activin signaling through ACVR2 is involved in tumor proliferation, apoptosis, migration, invasion, and metastases and high expression of ACVR2 is associated with aggressive disease in multiple malignancies, including renal cell carcinomas [RCC (17, 18)]. Despite evidence from studies that have investigated ACVR (19) as a biomarker in RCC and other malignancies, SYNJ2BP expression has not been associated with histopathologic features or survival outcomes in patients with RCC. Here, we report that SYNJ2BP is differentially expressed in ccRCC, and that reduced expression is associated with poor survival and reduced treatment response.
Materials and Methods
This study involved multi-step bioinformatic and immunohistochemical analyses to investigate the relationship between SYNJ2BP expression, the immune landscape, and survival outcomes in ccRCC (Figure 1). We utilized data from publicly available databases, including The Cancer Genome Atlas, Gene Set Cancer Analysis (TCGA), University of California Santa Cruz (UCSC), LinkedOmics and various other databases.
Schematic workflow for SYNJ2BP analysis in renal carcinoma. A filtering scheme was applied that included analysis of SYNJ2BP gene expression, associations with patient outcomes, clinicopathological features, and response to treatment, as well as immunohistochemical validation, functional annotation, and network mapping. Cox: Cox regression; DEG: differential gene expression; KEGG: Kyoto Encyclopedia of Genes and Genomes; KICH: kidney chromophobe; KIRC: kidney renal clear cell carcinoma; KIRP: kidney renal papillary cell carcinoma; KM: Kaplan-Meier; ROC: receiver operating characteristic.
Gene expression and cancer data analysis. Gene Set Cancer Analysis (GSCA; http://bioinfo.life.hust.edu.cn/GSCA accessed May 9, 2022) is a comprehensive genomic and immunogenomic database with over 10,000 multi-dimensional genomic datasets across 33 cancer types from The Cancer Genome Atlas (TCGA) and more than 750 small molecule drugs from Genomics of Drug Sensitivity in Cancer and cancer therapeutics response portal (20). GSCA was utilized to assess the differential expression of SYNJ2BP in kidney renal clear cell (KIRC), kidney renal papillary cell (KIRP), and kidney chromophobe (KICH) carcinoma datasets derived from the TCGA database. Differential expression analysis was performed on tumor samples and paired normal tissue (n=72 paired KIRC, n=32 paired KIRP, and n=25 paired KICH samples). The following screening conditions were applied: selection of “Input gene set: SYNJ2BP”, “Cancer types: KIRC, KIRP and KICH”, “Single gene level: Differential expression, Expression & Survival and Expression & Stage”, and “Differential Gene Set Variation Analysis (GSVA) and GSVA & Pathway activity”.
Gene expression validation. To validate gene expression results obtained from GSCA TCGA-KIRC clinical information and the RNA sequencing dataset (N=533) for SYNJ2BP were downloaded from the University of California Santa Cruz (UCSC) Xena portal (https://xena.ucsc.edu/public/ accessed May 9, 2022) to assess associations with patient outcomes (21). The following thresholds were the selection of TCGA-KIRC study, genomic data for SYNJ2BP, phenotypic data for the sample type, sex, age at initial pathologic diagnosis, pathologic N, pathologic stage, neoplastic histological grade, and genomic information for ID TCGA-HiSeqV2 data. Duplicate samples were removed by selecting primary tumors as a filter, leaving 533 samples. To ascertain the association between patient survival and SYNJ2BP expression, Kaplan-Meier (KM) analysis was selected. Raw data was used to generate KM plots to evaluate the impact of SYNJ2BP on patient survival, perform receiver operating characteristics/area under the curve (ROC/AUC) and univariate and multivariate analyses to determine whether SYNJ2BP expression was a predictive biomarker for ccRCC. ROC analysis was conducted with an area under the curve (AUC) of 0.5, indicating no discrimination, 0.6-0.7 moderate predictive value, 0.7-0.8 good performance, 0.8-0.9 strong predictive performance, and >0.9 indicating excellent performance. Hazard ratios (HR), 95% confidence intervals (CIs), and p-values were calculated by univariate and multivariate Cox proportional hazard models.
Immunohistochemical (IHC) analysis. Tissue microarrays (TMA) were prepared from archival formalin-fixed paraffin-embedded tissue samples, with duplicate cores procured from each de-identified ccRCC patient and matched normal tissue under Mayo Clinic Institutional Review Board approval (IRB# 14-004094). Immunostaining was performed according to the manufacturer’s protocol and as previously described (22-25). Briefly, TMA paraffin blocks were cut into 5 μm thick sections, mounted on slides, deparaffinized, blocked with Diluent (Dakocytomation, Glostrup, Denmark) for 30 minutes and probed for SYNJ2BP (1:750 dilutions; Sigma-Aldrich, HPA000866). The slides were scanned at 20X magnification with ScanScope XT (Aperio Technologies Inc, Vista, CA, U.S.A.) and were analyzed with ImageScope software (Aperio Technologies Inc). Scoring analysis based on signal intensity (0-300) was conducted by using algorithm-based macros in ImageScope by a trained researcher (H scores). H scores were calculated by the sum of the percentage of stained cells multiplied by an ordinal value that corresponded to signal intensity (0 = none, 1 = weak, 2 = moderate, and 3 = strong). Samples with insufficient tumor tissue for H score analysis were excluded from the study.
Gene set enrichment and functional annotation. LinkedOmics is a multi-omics and clinical database for 32 cancer types comprising 11,158 patients from the TCGA project (26). The genes co-expressed with SYNJ2BP were identified using the gene ontology (GO) enrichment analysis in LinkedOmics (http://www.linkedomics.org accessed May 9, 2022). The “LinkFinder” module in LinkedOmics was used to identify genes correlated with SYNJ2BP expression in the TCGA-KIRC dataset. The “LinkInterpreter” module was used to perform a gene set enrichment analysis (GSEA) from significantly correlated genes identified from “LinkFinder” to identify co-expressed genes positively and negatively correlated with SYNJ2BP within the gene ontology (GO) Biological Processes, KEGG, and Panther databases. Enrichr is a comprehensive database containing a large variety of curated libraries (Pathways, GO, transcription factors, etc.) used for functional annotation of differentially expressed genes [DEG; (27, 28)]. Genes interacting with SYNJ2BP derived from the Search Tool for the Retrieval of Interacting Genes (STRING) were used to perform a functional enrichment analysis in Enrichr (https://maayanlab.cloud/Enrichr accessed May 12, 2022). Differential enrichment analysis was conducted across GO: molecular functions and KEGG pathways.
Functional interaction and network mapping. STRING, a database that integrates known and predicted functional protein associations, was used to predict and construct protein-protein interaction (PPI) networks [v11.5; https://string-db.org/ accessed May 12, 2022; (29)]. Networks were created using the following criteria: a minimum interaction confidence score of 0.4, integrating sources from experimental data, curated pathway databases, co-expression, and text mining. GSVA was used to score genes identified from the SYNJ2BP PPI network.
Statistical analysis. SYNJ2BP gene expression levels were compared across tumor stages, grades, metastasis and nodal involvement using a one-way analysis of variance (ANOVA) followed by a Tukey-Kramer post hoc analysis (GraphPad Prism, La Jolla, CA, USA). Adjusted p-values were calculated to correct for multiple comparisons using the False Discovery Rate (FDR) procedure (30) and was calculated using the Multiple Correction Testing WebTool (31). The gene set enrichment results were analyzed for statistical significance using Pearson’s correlation coefficient test with the p-value and FDR set at 0.05. The ROC and KM survival outcome analyses, Cox risk hazard ratio (HR), 95% CI, and log-rank p-values were calculated by the MedCalc statistical software (v20; Ostend, Belgium) and the TCGA-KIRC dataset downloaded from UCSA Xena. SYNJ2BP expression was sorted into high (coded as 1) and low (coded as 0) groups and used consistently in KM and Cox regression analyses. ROC curves were generated using continuous gene expression values. The Mann-Whitney U-test was used to assess differences in SYNJ2BP expression between high and low tumor stages/grades and non-responder to responder groups. The p-value significance was set at p<0.05. H-score statistical analysis was performed using a 2-tailed paired Student t-test (GraphPad Prism), with p>0.05 considered as statistically significant.
Institutional Review Board statement. Mayo Clinic IRB reviewed and approved the use of de-identified patient samples used for immunohistochemistry in this study.
Data availability statement. Clear Cell Renal Cell Carcinoma data was downloaded from the UCSC XENA portal (https://xena.ucsc.edu/public/) to create Kaplan-Meier curves as well as to perform ROC analysis and multivariate analysis of clinicopathological features relative toSYNJ2BP expression. Data downloaded was filtered to only include ccRCC primary tumors (n=542) from the KIRC TCGA database or Solid Normal Tissue (n=72). Demographic information such as sex and age of initial diagnosis as well as gene expression, survival times, follow-up treatment success, stage, and grade were also downloaded. Fifteen (15) samples were excluded due to inefficient information on histological grade, and five (5) were excluded because of insufficient information on pathological stages.
Results
SYNJ2BP expression in renal cancers. To investigate the potential role of SYNJ2BP in renal carcinomas, the GSCA database was leveraged to analyze gene expression across three main subtypes: KIRC, KICH, and KIRP carcinomas. SYNJ2BP transcripts were significantly decreased in (p<0.001) in KIRC [Log2fold change (FC)=0.40, p=2.65E-36; FDR=9.73E-34] and KIRP (log2FC=0.60, p=3.84E-13; FDR=4.42E-11) between matched tumor and normal tissue. In KICH subtype, SYNJ2BP expression was found to be significantly increased (Log2FC=1.73, p=3.47E-7; FDR=3.40E-6; Figure 2A). Notably, the most significant reduction was observed in KIRC (Figure 2A). Across all pathological stages, SYNJ2BP exhibited significant differential expression between tumor and normal kidney tissues in KIRC (p=2.16E-11, FDR=6.34E-9) and KIRP (p=4.03E-4, FDR=6.97E-3) compared to KICH (p=0.72, FDR=0.86; Figure 2B).
SYNJ2BP mRNA expression in renal carcinoma histologic subtypes (KICH, KIRC, and KIRP) from the GSCA-TCGA RNA-seq datasets. (A) SYNJ2BP expression in ccRCC subtypes. Expression was significantly reduced in KIRC and KIRP tumor tissues compared to normal samples and significantly increased in the KICH subtype. (B) SYNJ2BP expression across pathological stages of ccRCC subtypes. Expression differed significantly across all pathological stages between tumor and normal kidney samples in KIRC and KIRP. The size of each bubble correlates with FDR significance. Statistical significance was defined as p<0.001. FC: Fold-change; FDR: false discovery rate; KICH: kidney chromophobe; KIRC: kidney renal clear cell carcinoma; KIRP: kidney renal papillary cell carcinoma.
Prognostic relevance of SYNJ2BP expression - overall survival. Recognizing the need for ccRCC prognostic biomarkers, differences in SYNJ2BP expression levels and the potential link to survival outcomes were assessed. Kaplan-Meier (KM) analysis was used to assess the correlation between SYNJ2BP expression and overall survival (OS). The survival analysis revealed that individuals with high expression had significantly longer overall survival than those with low expression (HR=0.39, 95% CI=0.28-0.54, p<0.0001; Figure 3A). Median OS was 5.2 years (95% CI=4.3-6.5 years) in the low-expression group, whereas the high-expression group did not reach a median overall survival. To establish whether SYNJ2BP expression could discriminate between patients with poor or favorable survival outcomes, a receiver operating characteristic (ROC) analysis was conducted. The ROC AUC for OS was 0.696 (95% CI=0.655-0.735, p<0.0001), with 60.0% sensitivity and 72.1% specificity at the optimal cutoff. These results indicate moderate discriminative ability, suggesting that SYNJ2BP expression may have prognostic potential for predicting OS outcomes (Figure 3B).
Prognostic value of SYNJ2BP expression in disease-specific, progression-free, and overall survival in ccRCC. (A) Kaplan-Meier (KM) curves showing overall survival (OS) of SYNJ2BP expression in patients with clear cell renal cell carcinoma (ccRCC). Median OS for the low expression group was 5.2 years (95% CI=4.3-6.5) and was not reached for the high expression group. (B) Receiver operating characteristic (ROC) area under the curve (AUC) of SYNJ2BP expression predicting OS (AUC=0.696). (C) KM curves demonstrating progression-free survival (PFS) of SYNJ2BP expression in ccRCC patients. Median PFS for the low expression group was 4.5 years (95% CI=3.3-5.6) and was not reached for the high expression group. (D) ROC curve of SYNJ2BP expression predicting PFS (AUC=0.724). (E) KM curves showing disease-specific survival (DSS) of SYNJ2BP expression in ccRCC patients. The median DSS for low expression was 7.1 years (95% CI=5.2-9.7) and was not reached for the high expression group. (F) ROC curve of SYNJ2BP expression predicting DSS (AUC=0.749). The hazard ratios shown were derived from univariate Cox regression analysis. In the KM analyses, high expression was used as a reference (coded as 1).
Progression-free survival. To evaluate whether SYNJ2BP expression is associated with disease progression, a KM analysis was conducted. The low-expression group was associated with significantly shorter progression-free survival (PFS) compared to the high-expression group (HR=0.24, 95% CI=0.17-0.35, p<0.0001; Figure 3C). Median PFS was 4.5 years (95% CI=3.3-5.6 years) in the low-expression group, whereas the median PFS in the high-expression group was not reached (Figure 3C). The ROC AUC for PFS was 0.724 (95% CI=0.683-0.761, p<0.0001), with 80.0% sensitivity and 61.6% specificity, indicating that SYNJ2BP expression is a good discriminator of disease progression (Figure 3D).
Disease-specific survival. Disease-specific survival (DSS) was assessed to determine if SYNJ2BP expression levels were associated with ccRCC-related mortality. KM analysis revealed that low expression was significantly associated with reduced DSS compared to the high-expression group (HR=0.16, 95% CI=0.09-0.27, p<0.0001; Figure 3E). The median DSS for the SYNJ2BP low expression group was 7.1 years (95% CI=5.2-9.7 years), as opposed to the high expression group where the DSS median was not reached, suggesting that low expression may be associated with worse DSS (Figure 3E). The ROC AUC for DSS was 0.749 (95% CI=0.709-0.786, p<0.0001) with 85.3% sensitivity and 59.3% specificity at the optimal cutoff, demonstrating good discriminatory performance of SYNJ2BP expression for DSS (Figure 3F).
To further evaluate whether the associations between SYNJ2BP expression and survival outcomes observed in the KM and ROC were independently associated with OS, DSS, and PFS, univariate and multivariate Cox regression analyses were conducted. Clinicopathological variables of age, sex, grade, stage, and SYNJ2BP expression were assessed. Univariate analysis revealed that low SYNJ2BP expression was significantly associated with reduced survival outcomes, including OS (HR=0.39, 95% CI=0.28-0.54, p<0.0001), PFS (HR=0.24, 95% CI=0.17-0.35, p<0.0001), and DSS (HR=0.16, 95% CI=0.09-0.27, p<0.0001), relative to the high-expression group (Table I). Similarly, advanced tumor stage and high tumor grade were a significant predictor of poor outcomes for DSS, PFS, and OS, while age was only a significant predictor for OS and PFS; no significance was observed for sex (Table I).
Univariate and multivariate Cox regression analysis was performed between SYNJ2BP expression and clinicopathological features within the TCGA-KIRC database.
In a multivariate Cox regression analysis, adjusting for age, sex, tumor stage and grade, only low SYNJ2BP expression remained as an independent predictor across all three survival parameters for DSS (HR=0.17, 95% CI=0.10-0.29, p<0.0001), PFS (HR=0.35, 95% CI=0.24-0.53, p<0.0001) and OS (HR=0.64, 95% CI=0.45-0.91, p=0.0123) compared to the high expression group. Other clinicopathological variables, including sex, age, stage, and grade, were not consistently significant across all three parameters. These results suggest that SYNJ2BP expression is an independent risk factor and prognostic biomarker for ccRCC.
SYNJ2BP gene expression and clinicopathological features. The relationship between SYNJ2BP expression levels and clinicopathological features in ccRCC was investigated using TCGA-KIRC database to determine the association between gene expression, disease characteristics, and clinical outcomes. Significant associations (p<0.05) were observed between SYNJ2BP expression and tumor grade (p<0.001), stage (p<0.001), age (p=0.0449), and sex (p=0.0454; Table II). Subsequent analyses assessed SYNJ2BP gene expression differences across tumor grades and stages. SYNJ2BP transcript levels were observed to be significantly lower in all histological Grades 1-4 (F[4, 592]=131.9, p<0.0001, FDR=0.0311) when compared to normal tissues (Figure 4A). Tukey post-hoc analysis demonstrated significant differences between Grades 1 and 3 (p=0.0311), Grades 1 and 4 (p<0.0001), Grades 2 and 3 (p=0.0011), Grades 2 and 4 (p<0.0001) and Grades 3 and 4 (p=0.0001; Figure 4A). No significant differences were observed between Grades 1 and 2. Comparing low-grade (Grades 1-2) with high-grade (Grades 3-4) tumors, SYNJ2BP expression was significantly decreased in patients with high-grade tumors (p<0.001; Figure 4B). Across pathological stages, gene expression levels were significantly lower relative to normal tissue (F[4, 598]=118.630, p<0.0001, FDR=0.0275; Figure 4C). Significant differences in SYNJ2BP expression were observed between Stages I and III (p<0.0001), Stages I and IV (p<0.0001), Stages II and III (p=0.0275), and Stages II and IV (p<0.0016). No significant differences were observed between Stages I and II or between Stages III and IV. Comparing early-stage (Stages I-II) with late-stage disease (Stages III-IV), SYNJ2BP expression was found to be significantly reduced in patients with advanced-stage tumors (p<0.001; Figure 4D). A similar pattern was observed in metastatic (M1; defined as spread to distant tissue) and non-metastatic cancer (M0) with significantly lower gene expression compared to normal tissue (F[2,567]=203.4, p<0.0001, FDR<0.001; Figure 4E). When comparing M0 with M1, SYNJ2BP expression was significantly lower (p<0.0001, FDR<0.001) in metastatic tissue relative to normal or non-metastatic tissue (Figure 4E). Comparing gene expression levels with nodal involvement, SYNJ2BP transcript levels differed significantly between primary tumors (N0) and lymph nodes (N1), compared to normal tissue (F[2, 323]=184.3, p<0.0001, FDR<0.001; Figure 4F). Similarly, SYNJ2BP expression was significantly lower in N1 tumors relative to N0 tumors (p=0.0004).
Association between SYNJ2BP gene expression and Clinicopathological features in ccRCC.
SYNJ2BP expression across clinicopathological features in ccRCC. (A) The association of SYNJ2BP expression with tumor grades. Transcript levels were significantly decreased across all tumor grades compared to normal tissue. One-way ANOVA with Tukey post-hoc analysis revealed significant differences in gene expression between individual tumor grades and normal tissue. (B) SYNJ2BP expression comparing low-grade (Grades 1-2) with high-grade (Grades 3-4) tumors. Patients with high-grade tumors had significantly reduced expression levels. (C) SYNJ2BP expression across tumor stages. Transcript levels differed significantly between tumor stages and normal tissue. Tukey-adjusted comparisons revealed significant differences in SYNJ2BP expression between individual tumor stages and normal tissue. (D) SYNJ2BP expression between early-stage (Stages I-II) and advanced-stage (Stages III-IV) tumors. Transcript levels were significantly lower in patients with high-grade tumors. (E) The association of SYNJ2BP expression with metastasis. Compared to normal tissue, transcript levels were significantly reduced in both M1 and M0 cancers. Relative to M0 cancers, M1 cases showed significantly decreased expression. (F) SYNJ2BP expression in relation to nodal involvement. Transcript levels were significantly lower in primary tumors (N0) and lymph node (N1) involvement compared to normal tissue. Compared to N0 tumors, N1 tumors showed significantly reduced expression. Statistical significance was assessed by one-way ANOVA or two-tailed t-test. Adjusted p-values are denoted as #p<0.05 and ***p<0.001. Non-significant associations (p>0.05) are labeled as ns. ccRCC: Clear cell renal cell carcinoma; FDR: false discovery rate; N0: node negative; N1: node positive; M1: metastatic; M0: non-metastatic.
SYNJ2BP expression and treatment response. Lower expression was observed to be associated with more advanced ccRCC stages and shorter PFS, indicating that SYNJ2BP expression is a good discriminator of disease progression. Therefore, we investigated whether the SYNJ2BP was differentially expressed in patients who responded and those who did not respond to anti-tumor therapies (Figure 4A-D). A Mann-Whitney U-test used to compare gene expression in responders (complete or partial response) to non-responders (stable or progressive disease) revealed that SYNJ2BP levels were significantly lower in non-responders (p=0.0052; Figure 5A). To establish whether expression could discriminate between treatment outcomes, a ROC analysis was conducted. The ROC AUC for treatment success was 0.679 (95% CI=0.566-0.793, p=0.002), with 49.5% sensitivity and 80.0% specificity at the optimal cutoff, indicating moderate discriminative ability. These results suggest that SYNJ2BP expression may have prognostic potential for predicting successful treatment outcomes (Figure 5B). Following initial treatment, the emergence of new tumor development was assessed. Results revealed that SYNJ2BP expression was significantly lower in patients with tumor recurrence (p=0.0027; Figure 5C). The ROC AUC for recurrence after treatment was 0.703 (95% CI=0.596-0.809, p<0.001), with 48.4% sensitivity and 82.6% specificity at the optimal cutoff (Figure 5D). Collectively, this suggests that SYNJ2BP expression is a good predictor of treatment success.
SYNJ2BP expression and treatment success in ccRCC. (A) SYNJ2BP expression in responders and non-responders. Mann-Whitney U test revealed significantly lower expression levels in non-responders compared to responders. (B) The receiver operating characteristic (ROC) area under the curve (AUC) of SYNJ2BP expression predicts treatment success (AUC=0.679). (C) SYNJ2BP expression and recurrence after treatment. The analysis demonstrated significantly lower expression levels in patients with tumor recurrence. (D) ROC curve of SYNJ2BP expression predicting recurrence after treatment (AUC=0.703). The symbol ** denotes statistical significance with a p<0.01. Treatment defined as local therapy (surgery or radiation therapy), or systemic therapy which could include oral (sunitinib, sorafenib, everolimus, pazopanib), or intravenous therapies (temsirolimus, bevacizumab, interferon-alpha, chemotherapy).
SYNJ2BP protein expression. Alterations in gene expression do not always correspond to changes in protein abundance, attributed to post-transcriptional regulation, protein synthesis efficiency, and breakdown. Accordingly, to establish whether SYNJ2BP protein expression is significantly associated with both clinical stages and histological grades, patient tissue microarrays (TMA) were prepared from matched normal and ccRCC tissue samples from Stages I, II, III, and IV (primary and metastatic). At the protein level, immunohistochemistry (IHC) staining of SYNJ2BP confirmed reduced levels in ccRCC patient samples at all stages compared to normal-matched tissue (Figure 6A). Further, H score analysis revealed a statistically significant difference in SYNJ2BP protein levels between ccRCC patient and matched normal samples (Figure 6B).
SYNJ2BP expression profile in normal renal tissue and clear cell renal cell carcinoma (ccRCC). Tissue microarray (TMA) of ccRCC patient samples versus matched normal tissue stained for SYNJ2BP expression in Stages I, II, III, IV (primary and metastatic; normal n=45, 27, 37, 13 and tumor n=52, 31, 41, 36 respectively). (A) Immunohistochemical staining (IHC) of SYNJ2BP. H-score ± standard deviation from the mean is shown. SYNJ2BP expression was quantified by an H-Score (staining intensity, with an index between 0-300 calculated as a percent of the total area). The symbol *** denotes statistical significance with a p<0.001, ****p<0.0001 from an unpaired t-test. (B) Stage I, (C) Stage II, (D) Stage III, (E) Stage IV (primary and metastatic).
Gene enrichment pathway analysis. Co-expressed genes often share molecular pathways and are involved in common cellular biological processes. For this reason, Gene Set Enrichment Analysis (GSEA) was conducted on genes co-expressed with SYNJ2BP. Gene Ontology-Molecular Function (GO-MF), KEGG, and Panther databases were used to investigate enriched distinct molecular functions and biological pathways of genes co-expressed with SYNJ2BP in ccRCC. All three databases indicated that genes positively co-expressed with SYNJ2BP were associated with phosphatidylinositol signaling pathways (Figure 7). The related normalized enrichment scores (NES) were NES=1.70 (p<0.0001, FDR=0.041; GO-MF), NES=1.76 (p<0.0001, FDR=0.037; KEGG), and NES=1.74 (p<0.0001, FDR=0.016; Panther; Figure 7). The encoded proteins of the co-expressed genes were mapped to the search tool for retrieval of interacting genes (STRING) to access protein-protein interaction (PPI) networks for SYNJ2BP. Five co-expressed genes, had high-confidence associations with SYNJ2BP (Figure 8A). The GSCA database was leveraged to analyze gene expression of the co-expressed genes in KIRC carcinoma to determine differential expression between normal and ccRCC patients (n=72 pairs; Figure 8B). Within the co-expressed genes, ACVR2A (Log2FC=0.73, p=1.34E-12, FDR=8.69E-12), PEX26 (Log2FC=0.76, p=2.53E-9, FDR=1.02E-8), RALBP1 (Log2FC=0.42, p=2.29E-14, FDR=1.94E-13), SACM1L (Log2FC=0.50, p=3.83E-19, FDR=6.96E-18), SYNJ2 (Log2FC=0.73, p=1.08E-3, FDR=2.05E-3), TMIGD1 (Log2FC=0.29, p=6.73E-5, FDR=1.52E-4) were significantly down-regulated (Figure 8B). GSVA analysis was performed using a gene set of SYNJ2BP and its co-expressed genes to estimate pathway activity scores and compare differences between normal and tumor tissue. In KIRC, the pathway showed significantly lower GSVA scores in tumor tissue (mean=1.32) relative to normal tissue (mean=1.62), with a log2 fold change of −0.30 (p=1.99E-72; Figure 8C). Comparing pathological stages in KIRC, GSVA scores decreased progressively with the advancing stage, with mean scores of 1.4 (Stage I), 1.3 (Stage II), 1.3 (Stage III), and 1.2 (Stage IV) with statistically significant differences across stages (p=2.74E-10). Statistical significance was not reached in the test for a decreasing trend in GVSA scores (p=0.09; trend score=−1.70; Figure 8D).
Pathway enrichment of SYNJ2BP co-expressed genes. Gene set enrichment analysis (GSEA) of genes positively correlated with SYNJ2BP using the Gene Ontology-Molecular Function (GO-MF), KEGG, and Panther databases. All three enrichment databases indicated significant and consistent pathway-level involvement in the phosphatidylinositol signaling pathway. Significance was set at p<0.05 and FDR<0.05.
Pathway-level and network analysis of SYNJ2BP expression in ccRCC. (A) Protein-protein interaction (PPI) network of genes co-expressed with SYNJ2BP. Five of the co-expressed genes (RRBP1, SYNJ2, ACVR2A, PEX26 and SYNJ2BP-COX16) had high-confidence associations (combined score>0.700; STRING). (B) Pathway enrichment of DEGs in KIRC. Several genes (ACVR2A, PEX26, RALBP1, SACM1L, SYNJ2 and TMIGD1) showed significant down-regulation (Log2FC<0.80, FDR<0.05; GSEA). (C) Differential pathway activity between tumor and normal tissues. Tumor samples showed significantly lower scores (log2FC of −0.30, p=1.99E-72; GSVA). (D) Stage-specific variation in pathway activity. Across pathological stages, the GSVA scores decreased progressively with the advancing stage, with statistically significant differences across all stages (p=2.74E-10). ACVR2A: Activin receptor type-2A; DEG: differentially expressed genes; FC: fold change; GSEA: gene set enrichment analysis; GSVA: gene set variation analysis; KIRC: kidney renal clear cell; PEX26: peroxisome assembly protein 26; RRBP1: ribosome binding protein 1; SYNJ2: synaptojanin-2.
Discussion
This study identified SYNJ2BP as a potential biomarker to predict survival outcomes and response to therapy in patients with ccRCC. Our findings offer novel insights into the potential role of SYNJ2BP-associated pathways in shaping clinical outcomes. These results warrant further investigation into the interaction between current and emerging immune therapies and the biological mechanisms mediated by SYNJ2BP. SYNJ2BP is a multifunctional protein. It has been shown to participate in mitochondrial DNA replication and cell surface localization to regulate the ACVR2 receptor. Additionally, SYNJ2BP has been shown to promote GRK-dependent phosphorylation, MAP-kinase (ERK1/2) signaling (32), and Delta-like canonical Notch ligand 4 (DLL4) binding to promote Notch receptor signaling (33). It has been shown to regulate angiogenesis and vascular remodeling by stabilizing DLL4 ligands, further intensifying Notch-dependent signaling pathways (34, 35). Moreover, in a recent study, SYNJ2BP was thought to play a protective role for the mitochondria, whereby it safeguards mRNA during times of stress and facilitates post-stress mitochondrial recovery and resumption of oxidative phosphorylation (36).
Emerging evidence suggests that SYNJ2BP may be involved in tumorigenesis. Highly expressed in breast cancer, SYNJ2BP is thought to promote cell migration, invasion, and epithelial-mesenchymal transition (EMT) via the activation of the PI3K/AKT/GSK3β/SNAI1 signaling pathway (37). In esophageal squamous cell carcinoma, SYNJ2BP is overexpressed (38). Conversely, SYNJ2BP expression was noted to be down-regulated in hepatocellular carcinoma (HCC), renal cell carcinoma, and related cell lines (34, 39). Low expression of SYNJ2BP was associated with larger tumor size, increased tumor nodules, vascular invasion, and higher TNM stages. Moreover, SYNJ2BP expression was shown to be an independent risk factor for HCC survival outcomes. Collectively, this data suggests that dysregulation of SYNJ2BP may contribute to tumor progression in a tissue-specific manner.
In this study, we characterized SYNJ2BP gene expression in ccRCC. Similar to HCC (34), SYNJ2BP expression was significantly reduced in ccRCC. In addition, lower expression of SYNJ2BP was associated with significantly worse overall survival, disease-specific survival, and progression-free survival. ROC analysis revealed that SYNJ2BP was a good predictor of overall survival and an even stronger progression-free and disease-specific survival predictor. Notably, both univariate and multivariate models demonstrated that low expression was an independent risk factor associated with overall, progression-free, and disease-specific survival. Hence, SYNJ2BP represents a novel prognostic biomarker for ccRCC. Analysis of SYNJ2BP expression within clinicopathological features revealed that SYNJ2BP is differentially expressed across tumor stages and grades, with significantly lower (p<0.05) expression in higher grades and more advanced-stage tumors. SYNJ2BP transcripts were reduced in metastatic tissue relative to normal tissue and non-metastatic cancer. Given that low transcript levels are associated with cancer progression, we examined differential gene expression in cancer treatment patients. Our analysis indicated that patients who were responsive to treatment (defined as a complete or partial response) had increased SYNJ2BP transcript levels relative to patients who were unresponsive (defined as a stable or progressive disease). This suggests that SYNJ2BP may behave similarly to a tumor suppressor gene and that its down-regulation may portend treatment unresponsiveness. The tumor suppressive effect of the SYNJ2BP protein may be the result of its regulation and endocytosis of ACVR2, whereby activin signaling through this receptor is associated with more aggressive cancer, tumor proliferation, apoptosis, migration, invasion, and metastases (17-19).
Given that SYNJ2BP expression is strongly associated with clinicopathological features, we conducted enrichment pathway analyses of genes co-expressed with SYNJ2BP in ccRCC to identify potential pathways in which the encoded protein may be involved in. All three GSEA databases converged on phosphatidylinositol signaling as a pathway over-represented in the genes associated with the SYNJ2BP gene. This finding corresponds with an earlier study by Liu et al. 2016 (34), which suggested a role for SYNJ2BP in modulating phosphatidylinositol 3-kinase (PI3K)/AKT activity in HCC. Of note is that PI3K/AKT and related pathways are dysregulated in ccRCC. Further, many of the PI3K/AKT pathway members are constitutively active and associated with aggressive cancers, including ccRCC (40-43). Moreover, activation of members and their downstream targets are associated with EMT, cancer cell survival, chemoresistance, and tumor-induced angiogenesis (44-48).
PI3K pathways represent a vast network of numerous proteins. Therefore, we conducted a protein interaction network analysis to define further the biological role SYNJ2BP plays in ccRCC. Interestingly, many of the proteins in the SYNJ2BP network are down-regulated in ccRCC, suggesting that SYNJ2BP and its interacting partners may represent a unique gene signature for ccRCC. Collectively, the proteins are involved in phosphatidylinositol signaling, as revealed by pathway enrichment analysis, which aligns with previous GSEA of genes co-expressed with the SYNJ2BP gene. In addition, SYNJ2BP partners, most notably SYNJ2, DLL1, and ACVR2A, are implicated in tumorigenesis or cancer progression (49-54). SYNJ2 is believed to regulate phosphatidylinositol signaling pathways that promote tumor invasiveness (55, 56), whereas DLL1 (a paralog of DLL4) may contribute to tumorigenesis by affecting cellular proliferation and invasiveness (57, 58). In addition, DLL1 may alter the tumor microenvironment by regulating angiogenesis and lymphoid differentiation via activation of the Notch signaling pathway (59-61). ACVR2A, via its kinase activity, can activate Smad proteins (Smad 2,3 and 4) to induce TGF-regulated gene transcription (62). TMIGD1 has been shown to inhibit tumor growth and cell migration in renal tumors by altering the phosphorylation status of cell cycle regulatory proteins (63). Hence, SYNJ2BP expression can potentially alter numerous critical pathways associated with tumorigenesis.
The limitations of the study include those related to a small sample size, retrospective analyses, and sources of data from database registries. Not knowing the specific systemic therapies utilized provides questions on how this biomarker would perform with current standard therapies with immune checkpoint inhibitors. This provides the opportunity to explore SYNJ2BP expression in patients treated with dual immune checkpoint inhibitors or immune checkpoint inhibitors with TKI therapies. Furthermore, investigating how SYNJ2BP can impact the tumor microenvironment, particularly the immune cells, could provide insight into immune-resistant or responsive tumors. In summary, our results suggest that SYNJ2BP may regulate tumor progression and treatment response via alterations of major phosphatidylinositol-related signaling pathways and is a prognostic biomarker for ccRCC with properties consistent with a tumor suppressor gene.
Footnotes
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Authors’ Contributions
The Authors contributed to the preparation in the following ways: Conceptualization, SOH, MDS, RRR; methodology, SOH, ET, JAC, TNT; resources, SOH, MDS, ET, JAC; writing – original draft preparation, SOH, MDS; writing – review and editing, RRR, ET, TNT, JAC, JPR, AMK; funding acquisition, SOH, ET, JAC. All Authors have read and agreed to the published version of the manuscript.
Funding
This research was supported by the National Cancer Institute of the National Institutes of Health under award numbers U54MD007582-40 (RRR) and P20CA264075 (SOH, MDS). The John Q. Ranney Research Fund for Renal Cell Carcinoma (JAC).
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received June 28, 2025.
- Revision received July 23, 2025.
- Accepted August 11, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.