Bioinformatics and Quantitative Real-Time Polymerase Chain Reaction Analysis of SUCNR1 and GPR37L1 in Schizophrenia

Podsumowanie

This study evaluates SUCNR1 and GPR37L1 as candidate schizophrenia-associated molecular markers through integrated bioinformatics analysis of the GSE54913 dataset, quantitative real-time polymerase chain reaction (qRT-PCR) validation, and correlation analysis with verbal memory in an independent cohort.

Streszczenie

Schizophrenia is a severe, complex, and multifactorial mental disorder involving numerous genetic susceptibility elements, leading to substantial disability, morbidity, and mortality. Despite significant progress in understanding its pathophysiology and etiology, specific diagnostic biomarkers for schizophrenia remain elusive. This study aimed to identify candidate molecular markers associated with schizophrenia. An integrated bioinformatics analysis was performed on the public microarray dataset GSE54913. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses revealed that the most significantly enriched GO terms were related to channel activity, including passive transmembrane transporter activity, ion channel activity, gated channel activity, and substrate-specific channel activity. The top five enriched KEGG pathways were insulin secretion, cAMP signaling pathway, nucleotide excision repair, TNF signaling pathway, and glutathione metabolism. Validation was conducted using quantitative real-time polymerase chain reaction (qRT-PCR) on an independent sample set from Wuhan Rongjun Youfu Hospital. The qRT-PCR results were largely consistent with the microarray analysis (Pearson r = 0.89, 95% CI: 0.66–0.97). Protein-protein interaction (PPI) network analysis identified two hub genes, SUCNR1 and GPR37L1, which were significantly associated with the GO term ‘ion channel activity’ and enriched in the KEGG pathway ‘insulin secretion’. Furthermore, SUCNR1 expression showed a negative correlation with verbal memory scores (r = -0.54, P = 0.015), whereas GPR37L1 expression showed a positive correlation (r = 0.59, P = 0.0034). These findings suggest that altered SUCNR1 and GPR37L1 expression may be associated with schizophrenia and may represent candidate molecular markers for further investigation.

Wprowadzenie

Schizophrenia is a chronic and complex mental disorder of unidentified etiology, characterized by severe brain dysfunction, cognitive impairment, and psychosocial deficits¹. It poses a major global health burden, affecting over 21 million people worldwide². Although diagnostic and therapeutic approaches have evolved considerably over the past fifty years, the core pathogenesis remains unclear, and long-term outcomes associated with substantial disability, morbidity, and mortality have not markedly improved³. Therefore, identifying potential pivotal genes and regulatory targets is imperative.

The pathophysiology of schizophrenia, while not fully elucidated, is widely attributed to genetic polymorphisms and expression variations⁴. For instance, genetic variation in the estrogen alpha gene may influence susceptibility to schizophrenia through alternative gene regulation and transcript processing⁵. Similarly, a functional promoter variant of NRG1 has been linked to schizophrenia and correlated with reduced expression of the type III NRG1 isoform⁶. Postmortem studies have shown significantly reduced expression of PDE4B isoforms in schizophrenic brains, suggesting its predictive potential⁷. Other candidates include the dopamine transporter (DAT), vesicular monoamine transporter (VMAT2), and monoamine oxidase (MAO), which regulate synaptic dopamine levels and may serve as biomarkers⁸. Additionally, TCP1 may contribute to cytoskeletal deficits via improper actin folding in schizophrenia⁹. Thus, elucidating gene expression profiles in the pathogenesis of schizophrenia could provide insights into risk prediction, mechanistic understanding, and therapeutic evaluation.

Recent computational approaches have advanced the identification of disease-associated genes through integrated network analysis and machine learning methods^10,11,12. Despite numerous genetic studies, no reliable blood-based diagnostic biomarkers have been translated into clinical practice for schizophrenia. To address this gap, the present study employed an integrated bioinformatics approach to reanalyze the GSE54913 dataset, which contains blood-derived transcriptomic data from well-characterized patients with schizophrenia and healthy controls. The dataset was selected because (1) peripheral blood samples are minimally invasive and clinically practical, (2) it includes a relatively large sample size among publicly available schizophrenia microarray datasets, and (3) raw data were available for reanalysis. The objectives of this study were to: (1) identify differentially expressed genes in schizophrenia patients compared to healthy controls using the GSE54913 dataset; (2) perform functional enrichment and PPI network analyses to identify hub genes; (3) validate the expression of candidate genes by qRT-PCR in an independent cohort; and (4) explore the correlation of hub gene expression with verbal memory performance. To the authors’ knowledge, this is the first study to identify SUCNR1 and GPR37L1 as candidate blood-based molecular markers for schizophrenia.

Protokół

This study was approved by the Ethics Committee of Wuhan Rongjun Youfu Hospital (project identification code, YF-IRB202310215) and conducted in accordance with the principles of the Declaration of Helsinki. All participants were Chinese community residents and provided written informed consent.

Study subjects and blood sampling
Ten adults diagnosed with schizophrenia and ten healthy control volunteers with no family history of mental illness within three generations were enrolled. Inclusion criteria for schizophrenia patients: (1) diagnosis of schizophrenia according to DSM-5 criteria confirmed by two independent senior psychiatrists; (2) age 18–65 years; (3) no change in antipsychotic medication for at least 4 weeks prior to blood sampling; (4) willingness to provide written informed consent. Exclusion criteria: (1) comorbid major medical illnesses (e.g., diabetes, cardiovascular disease, cancer); (2) substance abuse or dependence within the past 6 months; (3) intellectual disability; (4) pregnancy or lactation. Inclusion criteria for healthy controls: (1) no personal or family history (within three generations) of any mental illness; (2) no current or past antipsychotic medication use; (3) age- and sex-matched to the schizophrenia group; (4) no major medical illnesses. Mean disease duration for schizophrenia patients was 12.5 ± 6.8 years; all patients were on stable antipsychotic medication (6 on risperidone, 4 on olanzapine); mean PANSS total score was 76.4 ± 12.3. Controls were matched for age (±5 years) and sex (5 males, 5 females per group). Verbal memory was assessed using the total recall score on the Hopkins Verbal Learning Test–Revised (HVLT-R). All subjects were recruited from Wuhan Rongjun Youfu Hospital between January 1 and July 31, 2025. Baseline demographics are summarized in Table 1.

Sample size was determined using sample size calculation software for a two-tailed independent t-test with an effect size of 1.2, α = 0.05, and power (1-β) = 0.80, yielding a minimum of nine subjects per group.

Venous blood was drawn into EDTA-anticoagulated tubes and processed without delay. Peripheral blood mononuclear cells (PBMCs) were then separated from the samples by density gradient centrifugation using density gradient medium. Briefly, blood was diluted 1:1 with phosphate-buffered saline (PBS, pH 7.4), layered onto the density gradient medium, and centrifuged at 400 × g for 30 min at 20 °C with the brake off. After centrifugation, the PBMC-containing interface was carefully transferred to a new tube. The cells were rinsed twice with PBS, with each wash followed by centrifugation at 300 × g for 10 min at 4 °C, and the final pellet was suspended in PBS. Viability was evaluated using the trypan blue dye-exclusion method, and only preparations with ≥95% viable cells were used. Isolated PBMCs were aliquoted and stored at −80 °C for a maximum of 3 months before RNA extraction.

Microarray data
The workflow of the study is shown in Figure 1. The microarray dataset GSE54913 was downloaded from the GEO database (http://www.ncbi.nlm.nih.gov/geo/). This dataset was selected because (1) it contains transcriptomic data from peripheral blood samples, which are minimally invasive and clinically practical for biomarker discovery; (2) it includes a relatively large sample size among publicly available schizophrenia microarray datasets (18 patients, 12 controls); (3) the raw data were available for reanalysis. According to the GEO record, the samples were derived from peripheral blood mononuclear cells (PBMCs), not plasma. The original description ‘plasma samples’ in the previous version was an error and has been corrected.

Data preprocessing and differentially expressed gene (DEG) screening
Raw data (.CEL files) were preprocessed using microarray preprocessing software. Background correction was performed using the Robust Multichip Average (RMA) method, followed by quantile normalization and log2 transformation. Probesets without any gene annotation were filtered out. Probes with >20% missing values across samples were excluded; missing values for remaining probes were imputed using the k-nearest neighbors algorithm (k = 10) implemented in missing-value imputation software. Batch effects were not present as all samples were processed in a single batch according to the GEO record. After preprocessing, expression values for 17,200 genes were obtained for downstream analysis. DEGs between schizophrenia patients and healthy controls were identified using differential expression analysis software. Genes with a false discovery rate (FDR) < 0.05, absolute fold change (FC) > 1.2, and P < 0.05 were considered differentially expressed. A relatively low FC threshold (absolute FC > 1.2) was chosen because schizophrenia is a complex psychiatric disorder in which individual gene expression differences are often subtle rather than dramatic.

GO and pathway enrichment analyses
Functional enrichment analyses for Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were performed using functional enrichment analysis software. The background gene set consisted of all genes that passed preprocessing (17,200 genes). Enrichment terms and pathways with raw P-values < 0.05 were reported. Because raw P-values were used for GO and KEGG enrichment outputs, these results should be interpreted as exploratory.

PPI network analysis and hub gene identification
A protein-protein interaction database was used to construct a PPI network, with a combined interaction score > 0.9 set as the threshold. The network was visualized using network visualization software. Hub genes were identified using a hub gene identification plugin with the Degree algorithm. The top 10 nodes with the highest degree scores were selected as hub genes. The subnetwork of key hub genes was extracted using a subnetwork extraction plugin with default parameters (degree cutoff = 2, node score cutoff = 0.2, K-core = 2, max depth = 100).

Total RNA isolation and qRT-PCR
β-Actin was used as the internal reference (housekeeping) gene. The stability of β-actin expression across samples was confirmed by no significant difference in Ct values between the schizophrenia and control groups (P > 0.05). Relative quantification was performed using the 2^−ΔΔCt method. All reactions were performed in triplicate, and the mean Ct value was used for calculation (see primer sequences in Table 2). The top ten DEGs (five most significantly upregulated and five most significantly downregulated by fold change) were selected for initial qRT-PCR validation to confirm the overall reliability of the microarray data. Subsequently, SUCNR1 and GPR37L1 were selected for focused validation based on three criteria: (1) they were identified as hub genes in the PPI network analysis (degree ≥ 7); (2) they were significantly associated with the top enriched GO term ‘ion channel activity’ and the KEGG pathway ‘insulin secretion’; (3) both encode GPCRs, which are known drug targets in schizophrenia.

Statistical analysis
Normality was evaluated with the Shapiro–Wilk test. Variables showing an approximately normal distribution (P > 0.05) were analyzed using parametric methods, including Student’s t-test for two-group comparisons and one-way ANOVA for comparisons involving more than two groups. When the normality assumption was not met, Mann–Whitney U or Kruskal–Wallis tests were used, as appropriate. All statistical tests were two-tailed. For multiple comparisons (e.g., in ANOVA post-hoc tests), the Benjamini-Hochberg false discovery rate (FDR) method was applied, with an FDR threshold of 0.05. For correlation analyses, no multiple-comparison correction was applied, as only two correlations were performed; the raw P-values are reported with caution.

Wyniki

Identyfikacja DEG i klasteryzacja hierarchiczna
Analiza zbioru danych GSE54913 pozwoliła zidentyfikować 473 geny różnicowo wyrażone (DEG), w tym 357 genów o zwiększonej i 116 genów o zmniejszonej ekspresji, pomiędzy pacjentami ze schizofrenią a grupą kontrolną (Rycina 2A,B). Klasteryzacja hierarchiczna tych DEG pozwoliła odróżnić próbki pochodzące od osób ze schizofrenią od próbek kontrolnych (Rycina 2C).

Dyskusja

Schizophrenia is a severe, multifactorial disorder involving numerous genetic susceptibility factors^13,14. While progress has been made in understanding its pathophysiology, reliable diagnostic biomarkers are still lacking. In this study, an integrated bioinformatics approach identified 473 DEGs in blood samples from patients with schizophrenia. GO and KEGG analyses highlighted enrichment in channel activity-related terms and pathways, including insulin secretion...

Materiały

Lista materiałów użytych w tym artykule

Nazwa	Firma	Numer katalogowy	Komentarze
affy	Bioconductor	https://bioconductor.org/packages/affy/	microarray preprocessing software. Description: R package for microarray preprocessing
clusterProfiler	Bioconductor	https://bioconductor.org/packages/clusterProfiler/	functional enrichment analysis software. Description: R package for functional enrichment analysis
CytoHubba	National Institute of Bioinformatics	https://apps.cytoscape.org/apps/cytohubba	hub gene identification plugin. Description: Cytoscape plugin for hub gene identification
Cytoscape	Cytoscape Consortium	https://cytoscape.org/	network visualization software. Description: Software for network visualization and analysis
G*Power	Heinrich Heine University Düsseldorf	https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower	sample size calculation software. Description: Sample size calculation software
GraphPad Prism	GraphPad Software	https://www.graphpad.com/	statistical analysis and graphing software. Description: Statistical analysis and graphing software
Histopaque-1077	Sigma-Aldrich	10771	density gradient medium. Description: Density gradient medium for PBMC isolation
impute	Bioconductor	https://bioconductor.org/packages/impute/	missing-value imputation software. Description: R package for missing-value imputation
Limma	Bioconductor	https://bioconductor.org/packages/limma/	differential expression analysis software. Description: R package for differential expression analysis
MCODE	Cytoscape app	https://apps.cytoscape.org/apps/mcode	subnetwork extraction plugin. Description: Cytoscape plugin for subnetwork extraction
R/Bioconductor	R Foundation	https://www.r-project.org/	statistical computing environment. Description: Statistical computing environment
STRING	EMBL	https://string-db.org/	protein-protein interaction database. Description: Protein-protein interaction database
SYBR Green	Takara	RR820A	fluorescent dye for qRT-PCR. Description: Fluorescent dye for qRT-PCR
TRIzol	Takara	9109	RNA extraction reagent. Description: Reagent for RNA extraction