Integrative transcriptomic and proteomic analysis reveals that SERPING1 inhibits neuronal proliferation via the CaMKII-CREB-BDNF pathway in schizophrenia

Abstract

BACKGROUND

Schizophrenia (SZ), a chronic and widespread brain disorder, presents with complex etiology and pathogenesis that remain inadequately understood. Despite the absence of a universally recognized endophenotype, peripheral blood mononuclear cells (PBMCs) serve as a robust model for investigating intracellular alterations linked to SZ.

AIM

To preliminarily investigate potential pathogenic mechanisms and identify novel biomarkers for SZ.

METHODS

PBMCs from SZ patients were subjected to integrative transcriptomic and proteomic analyses to uncover differentially expressed genes (DEGs) and differentially expressed proteins while mapping putative disease-associated signaling pathways. Key findings were validated using western blot (WB) and real-time fluorescence quantitative PCR (RT-qPCR). RNAi-lentivirus was employed to transfect rat hippocampal CA1 neurons in vitro, with subsequent verification of target gene expression via RT-qPCR. The levels of neuronal conduction proteins, including calmodulin-dependent protein kinase II (caMKII), CREB, and BDNF, were assessed through WB. Apoptosis was quantified by flow cytometry, while cell proliferation and viability were evaluated using the Cell Counting Kit-8 assay.

RESULTS

The integration of transcriptomic and proteomic analyses identified 6079 co-expressed genes, among which 25 DEGs were significantly altered between the SZ group and healthy controls. Notably, haptoglobin (HP), lactotransferrin (LTF), and SERPING1 exhibited marked upregulation. KEGG pathway enrichment analysis implicated neuroactive ligand-receptor interaction pathways in disease pathogenesis. Clinical sample validation demonstrated elevated protein and mRNA levels of HP, LTF, and SERPING1 in the SZ group compared to controls. WB analysis of all clinical samples further corroborated the significant upregulation of SERPING1. In hippocampal CA1 neurons transfected with lentivirus, reduced SERPING1 expression was accompanied by increased levels of CaMKII, CREB, and BDNF, enhanced cell viability, and reduced apoptosis.

CONCLUSION

SERPING1 may suppress neural cell proliferation in SZ patients via modulation of the CaMKII-CREB-BDNF signaling pathway.

Key Words: Transcriptomics; Proteomics; Schizophrenia; SERPING1; Pathogenesis

Core Tip: The strength of this study lies in the integration of transcriptomics and proteomics data from the same sample source (peripheral blood mononuclear cells) to better observe mRNA-protein inconsistencies and mRNA-protein correlations, thus explaining the biology of schizophrenia in a holistic manner. Proteomics and transcriptomics correlation analyses provide a panoramic view of the expression profile of schizophrenia, enabling complementarity and integration at the transcriptional and protein levels. Again, this study is one of the few to investigate the role of the SERPING1 gene in neural cell proliferation and apoptosis.

INTRODUCTION

Schizophrenia (SZ) is a complex mental illness characterized by positive symptoms such as hallucinations, delusions, and speech disturbances, alongside negative symptoms including hypobulia, anhedonia, and social withdrawal, as well as cognitive deficits. Its progression typically spans three stages: prodromal (pre-psychotic), psychotic, and chronic phases[1,2]. Diagnostic approaches rely exclusively on clinical manifestations, as no auxiliary tests or reliable biomarkers are available[3], creating significant challenges in early detection, disease stratification, therapy selection, and prognosis. This diagnostic limitation often leads to misdiagnosis, delayed diagnosis, ineffective treatment, or relapse. The poorly understood pathogenesis of SZ is linked to neurodevelopmental abnormalities, genetic predispositions, and the cumulative influence of environmental factors[4,5]. Evidence further implicates dysregulated inflammation, immune system alterations[6,7], and disrupted energy metabolism in both the brain and the body[8]. Epidemiological findings by Benros et al[9] highlight an increased SZ risk associated with severe infections and autoimmune disorders. Moreover, sustained remission of psychotic symptoms in patients undergoing prolonged warfarin therapy for deep vein thrombosis underscores a potential connection between coagulation pathway dysfunction and SZ progression[10].

Genome-wide association studies have identified over 100 SZ-associated risk genes[11], encompassing key functional categories including neuronal, immune, and histone-related pathways[12]. However, the mechanisms through which these genes contribute to SZ pathogenesis remain poorly understood. Current SZ research primarily relies on post-mortem tissues, blood samples, and patient-derived induced pluripotent stem cells. While post-mortem samples provide valuable insights, their acquisition and preservation present significant logistical hurdles compared to the relative ease of obtaining blood samples. iPSC models are instrumental for investigating fundamental neural cell processes but are less effective for case-control studies due to statistical constraints and challenges in discerning individual-level cellular phenotypes[13]. Blood samples, by contrast, reflect immune system dynamics and provide indirect evidence linking central nervous system activity in SZ to alterations in peripheral blood gene expression[14]. Among blood-derived components, peripheral blood mononuclear cells (PBMCs) are widely utilized in SZ research due to their proven sensitivity and specificity as biomarkers across a range of diseases[15,16]. This study integrated transcriptomics and proteomics analyses of PBMCs from SZ patients with in vitro experiments to offer preliminary insights into SZ pathogenesis.

MATERIALS AND METHODS

Research subjects

SZ group: This study involved 150 SZ inpatients, comprising 6 participants for omics analysis and 150 for western blot (WB) detection, recruited from Hainan Anning Hospital between March and July 2022. Data collected included demographic and clinical parameters such as age, age of onset, gender, SZ subtype, family history, and head imaging results, with detailed clinical profiles of those included in the omics analysis provided in Table 1. Eligibility was determined by an SZ diagnosis based on International Classification of Diseases-10 criteria, independently confirmed by two trained specialists. Exclusion criteria encompassed intellectual disabilities, other psychiatric disorders, neurological conditions, significant systemic illnesses (e.g., cancer, leukemia), severe infections (e.g., acquired immune deficiency syndrome), and pregnancy or lactation.

Table 1 Clinical information of schizophrenia used in the Omics analysis.

SZ	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Gender	Male	Male	Male	Female	Female	Female
Age	36	25	19	16	29	36
Age of onset	24	19	17	15	20	28
Disease classification	Undifferentiated type	Undifferentiated type	Paranoid type	Undifferentiated type	Paranoid type	Undifferentiated type
Family history	No	No	No	No	No	No
Imaging examination	Abnormal	Abnormal	Normal	Abnormal	Normal	Normal

Healthy controls, matched for age and gender with the Schizophrenia group, was included during the same period. SZ: Schizophrenia.

Healthy control group: A cohort of 150 healthy controls (HC) from Haikou People's Hospital, matched by age and gender to the SZ group, was selected during the same timeframe. Participants were required to meet the following inclusion criteria: absence of personal or familial history of mental disorders, no intellectual disability, no prior or current alcohol, drug, or substance abuse, no chronic illnesses, no history of head trauma or neurological surgeries, and not being pregnant or lactating.

Neuronal cells in the CA1 region of the hippocampus of male healthy SD rats: The study was reviewed and approved by the Biomedical Ethics Committee of Haikou Hospital affiliated with Xiangya School of Medicine, Central South University (Haikou People's Hospital), under ethics approval number 2021- (Ethics Review)-263 (Ethical Statement 1). Additional approval was granted by the Biomedical Ethics Committee of Hainan Anning Hospital, referenced as approval number 2022- (Ethics Review)-2 (Ethical Statement 2). Written informed consent was obtained from participants or their guardians, with documentation validated by the respective Medical Ethics Committees before the commencement of the study. The experiment does not involve animal experiments and is based on rat neuronal cells sourced from the Animal Experiment Center of the School of Pharmacy, Hainan University, China.

Experimental methods and steps

Collection and PBMC extraction of clinical samples: Fresh fasting venous whole blood was collected from participants using 9 mL sodium citrate anticoagulant vacuum tubes (sodium citrate: blood ratio = 1:9). PBMCs were isolated via lymphocyte separation solution, followed by washing with PBS. After discarding the supernatant, the PBMCs were harvested and preserved at -80 °C.

RNA extraction and sequencing: PBMCs were ground in liquid nitrogen, transferred to 1.5 mL centrifuge tubes, and processed for RNA extraction using the Trizol method. RNA concentration and purity were assessed with an enzyme-labeled instrument, and high-quality RNA was preserved at -80 °C. mRNA was enriched via magnetic beads conjugated with Oligo (dT) targeting the poly-A structure at the 3' end, followed by reverse transcription into cDNA for library construction. Fragment size (25-1000 bp) and concentration (0.1-50 ng/µL) were verified using the DNA 1000 assay kit (Agilent Technology, China). Sequencing of qualified libraries was conducted on the Illumina HiSeq Xten platform. Raw data underwent quality control and alignment with the reference genome, with unqualified sequences filtered to obtain high-quality reads. HISAT2 (https://daehwankimlab.github.io/hisat2/) was utilized for genome alignment. Gene expression levels were quantified using featureCounts (http://bioinf.wehi.edu.au/featureCounts/) to compute FPKM values, enabling comparative analysis of read counts between samples. Differential expression analysis was performed using DESeq2 (https://bioconductor.org/packages/release/bioc/html/DESeq2.html) and edgeR (https://bioconductor.org/packages/release/bioc/html/edgeR.html). Hypothesis testing results were adjusted for multiple comparisons to yield P-adj values, with DEGs defined by P-adj < 0.05 and |log2foldchange| > 1.

Quantitative analysis of data-independent acquisition proteomics: PBMCs were lysed in SDT buffer, subjected to ultrasonication, and heated at 100 °C for 15 minutes. After centrifugation at 14000 g for 40 minutes, protein concentration in the supernatant was determined using a BCA protein assay kit (Bio-Rad). Equivalent aliquots from each sample were combined to form a pooled sample, which was fractionated by HPRP and analyzed via LC-MS/MS in QE-HFX_DDA mode. The DDA library database was generated using Spectronaut Pulsar X software (https://biognosys.com/resources/spectronaut-manual/). Individual samples were then processed in LC-MS/MS data-independent acquisition (DIA) mode, with the DDA database employed for qualitative and quantitative assessments.

Integrative transcriptomics and proteomics analysis: The integrative analysis was performed using R (https://www.r-project.org/) and associated tools, including Blast2 (https://www.blast2go.com/), to derive insights from transcriptomic and proteomic datasets.

Validation of protein expression by WB in clinical samples: Total protein extracted from PBMCs was resolved by SDS-PAGE and transferred onto PVDF membranes (Merck Millipore, Bedford, MA, United States). Membranes were blocked with 5% non-fat milk in PBST and incubated overnight at 4 °C with primary antibodies (anti-HP, anti-LTF, and anti-SERPING1, Abcam, United Kingdom). Secondary antibody incubation at room temperature for 40 minutes preceded visualization using an ECL kit (Beyotime, Shanghai, China). GAPDH was utilized as the internal control, and ImageJ software was used for grayscale analysis to quantify relative protein expression levels.

Real-time fluorescence quantitative PCR detection of expression in clinical samples: Total RNA extracted from PBMCs served as the template for cDNA synthesis following the Thermo real-time fluorescence quantitative PCR (RT-qPCR) Kit protocol. Quantification of mRNA expression was conducted via real-time fluorescence qPCR using TransGen Biotech^® SybrGreen qPCR SuperMix. Primer sequences specific to the target genes were detailed in Table 2. Relative mRNA expression levels were calculated using the 2-ΔΔCT method (Livak method).

Table 2 Real-time fluorescence quantitative PCR primer information.

ID	Forward primer (5'- > 3')	Reverse primer (5'- > 3')
GAPDH	TGTTCGTCATGGGTGTGAAC	ATGGCATGGACTGTGGTCAT
HP	AATGTGAAGCAGATGACGGC	GATCCCAGTCGCATACCAGG
LTF	TTGTCTTCCTCGTCCTGCTG	GGGAGCCAAAGAGCTGGAAT
SEEPING1	CTCCTACCCAGCCCACTACT	CCAAGTTGGCGTCACTGTTG

HP: Haptoglobin; LTF: Lactotransferrin.

Rat hippocampal tissue isolation and cell culture: Healthy male SD rats (250 ± 20 g) were housed in a controlled environment (21 ± 2 °C, 40%-70% humidity) with regular rodent chow and clean bedding. Anesthesia was administered using 0.3% sodium pentobarbital (1 mL/100 g), followed by perfusion and fixation with physiological saline and 4% paraformaldehyde. After dissection, the hippocampus was isolated and stored at -80 °C. CA1 region tissues were washed in D-Hanks solution, minced, and digested with 0.25% trypsin at 37 °C to obtain cell clusters or single cells. Following centrifugation at 500 rpm for 5 minutes, the supernatant was removed, and the cells were resuspended in complete neuron culture medium. The cell density, adjusted to 2 × 10⁶ cells/mL, was determined using a hemocytometer. Cells were seeded into culture flasks and maintained at 37 °C in a 5% CO₂ incubator.

Construction of RNAi-silencing lentivirus: The SERPING1 gene sequence (GENE ID: 2182) was retrieved from NCBI, and interference targets were designed as outlined in Table 3. The Plko.1-EGFP-PURO vector served as the cloning backbone, utilizing Age I (ACCGGT) and EcoR I (GAATTC) as restriction sites. Recombinant plasmids were amplified and purified using the OMEGA Endotoxin-Free Plasmid Mini Kit. The A260/280 ratio and plasmid concentration were quantified, and samples were stored at -20 °C. Serum-free, antibiotic-free DMEM (1 mL) was prepared for transfection, into which lentiviral packaging and expression plasmids were combined with the transfection reagent. The solution was vortexed and incubated at room temperature for 30 minutes to facilitate polymer formation. The polymer mixture was then gradually added to 293T cells cultured in fresh medium, followed by incubation. Viral supernatants were subsequently harvested, centrifuged, filtered, concentrated, purified, aliquoted, and stored at -80 °C. Lentivirus-infected 293T cells underwent flow cytometry and fluorescence microscopy analysis to assess lentiviral packaging titers. The corresponding results are presented in Supplementary material.

Table 3 The information of RNAi target.

RNAi target	Target sequence
SEEPING1-sh1	ACACTGTTTTGCAGCTTTTAT
SEEPING1-sh2	AACCAGTGGATCCACCAAAAT
SEEPING1-sh3	AGTCTTCATGGGCCGTGTATA

Lentiviral infection of target cells (neuronal cells in the CA1 region of the rat hippocampus): Lentiviral transfection in target cells (neuronal cells within the CA1 region of the rat hippocampus) was initiated by diluting 50 μL of si-SERPING1 Lentivirus in Opti-Medium and mixing gently via pipetting. Separately, 1.2 μL of Lipofectamine 2000 was diluted in 50 μL of Opti-Medium, mixed gently by pipetting, and allowed to stand at room temperature for 5 minutes. The Lentivirus solution and diluted transfection reagent were combined, mixed gently, and left to stand at room temperature for 20 minutes to form transfection complexes. A volume of 100 μL of these complexes was added to each well of a 24-well cell culture plate, followed by gentle rocking to ensure even distribution. The plate was incubated at 37 °C in a 5% CO₂ atmosphere for 24-48 hours, with the active transfection phase lasting 4-6 hours.

RT-qPCR detection of SEEPING1 expression in neuronal cells in the CA1 region of rat hippocampus: Neuronal cells were transfected with Lentivirus in four groups: negative control (NC), si-SERPING1#1, si-SERPING1#2, and si-SERPING1#3, corresponding to three distinct silencing targets. After 4-6 hours of transfection, RNA extraction was performed, followed by cDNA synthesis using the Thermo qPCR RT Kit. Real-time fluorescence-based quantitative PCR was carried out using TransGen Biotech^® SybrGreen qPCR SuperMix, with the following primer sequences designed for the rat SERPING1 gene:

Forward primer (5'- > 3'): CGCCTCTCTGAGCCTGTATG

Reverse primer (5'- > 3'): TCAGTTCCAACACCGTCTCG

The relative expression of the target gene mRNA was calculated using the 2-ΔΔCT method (Livak method).

WB detection of neuronal expression of neural-related conductance proteins calcium/calmodulin-dependent protein kinase II, cAMP response element-binding protein, brain-derived neurotrophic factor in neuronal cells of CA1 region of rat hippocampus: Lentivirus transfection was conducted across three groups: Control, NC, and si-SERPING1#1, which demonstrated the highest silencing efficiency among the tested candidates. WB detection was performed 4-6 hours post-transfection, following the procedural details specified in the previous WB detection.

Apoptosis of target cells detected by flow cytometry: Rat hippocampal neurons in the logarithmic growth phase were seeded into 6-well culture plates at a density of 7 × 10⁵ cells per well, with 2 mL of medium per well. Lentivirus transfection was conducted across the Control, NC, and si-SERPING1-1 groups. Following 4-6 hours of transfection, cells were harvested by centrifugation, rinsed twice with pre-chilled PBS (4 °C), and resuspended in 500 μL of binding buffer to achieve a final concentration of 10⁶ cells/mL. Subsequently, 100 μL of the suspension was transferred to a 5 mL flow cytometry tube, supplemented with 5 μL Annexin V-APC and mixed thoroughly, followed by 5 μL Propidium Iodide. The mixture was incubated at room temperature in the dark for 15 minutes. Apoptotic analysis was performed using a flow cytometer as per the Biolegend Annexin V-FITC/PI double-staining apoptosis detection kit protocol.

Proliferation of target cells detected by Cell Counting Kit-8: The viability of rat hippocampal neurons was evaluated using the Cell Counting Kit-8 (CCK-8, Solarbio, Beijing, China). Lentivirus transfection was applied to three groups: Control, NC, and si-SERPING1-1. Following transfection, cells from each group were trypsinized and seeded into 96-well plates at a density of 8000 cells per well. After stabilization, CCK-8 assays were performed on days 1, 2, 3, 4, and 5. At each time point, 10 μL of CCK-8 solution was added per well, and cells were incubated at 37 °C in a 5% CO₂ atmosphere for 1-4 hours. The optical density (OD) at 450 nm was then determined using a microplate reader (ELx800, BioTech, VT, United States). OD measurements were recorded, and a growth curve was constructed with time as the x-axis and OD values as the y-axis.

Statistical analysis

Data analysis and visualization were conducted using GraphPad Prism 9 software. Results were presented as mean ± SD. Two-group comparisons were performed using multiple t-tests, while one-way analysis of variance (ANOVA) was used for comparisons among multiple groups. Statistical evaluations included χ² tests where applicable. A P value < 0.05 was considered indicative of statistical significance.

RESULTS

RNA sequencing results

A total of 3317 differentially expressed genes (DEGs) were identified, including 1635 upregulated and 1682 downregulated genes. The 10 most statistically significant DEGs are illustrated in Figure 1A, with detailed information provided in Table 4. Clustering analysis revealed that the DEGs distinctly separated SZ patients from HCs, indicating substantial changes in PBMC gene expression among SZ patients (Figure 1B).

Figure 1 Analysis of transcriptomics sequencing results. A: Volcano plot illustrating the distribution of differentially expressed genes between the schizophrenia (SZ) and Healthy controls (HC) groups; B: Cluster heatmap of differentially expressed genes in the SZ and HC groups.

Table 4 Differential gene expression statistics.

Gene_id	Symbol	Disease_read count	Normal_read count	Log2FC	P-adj
ENSG00000107643	MAPK8	2141.92	980.48	1.1278	1.63E-36
ENSG00000108691	CCL2	965.58	14.1	6.0888	1.04E-31
ENSG00000169439	SDC2	726.54	12.05	5.9209	8.61E-31
ENSG00000160460	SPTBN4	100.36	16.26	2.626	3.57E-27
ENSG00000088899	LZTS3	436.07	138.93	1.6541	3.95E-24
ENSG00000095794	CREM	1760.64	355.27	2.3089	2.31E-23
ENSG00000144749	LRIG1	1648.48	755.18	1.127	2.46E-22
ENSG00000171793	CTPS1	1093.29	384.58	1.5064	1.08E-21
ENSG00000176641	RNF152	182.81	8.89	4.3723	5.64E-20
ENSG00000260196	-	163.81	50.63	1.6952	6.05E-19

Displayed as the top 10 genes with up-regulated expression.

Quantitative analysis of DIA proteomics

DIA-based quantitative proteomics was conducted for the SZ and HC groups. The total proteins and peptides identified in each sample are represented in a bar chart (Figure 2A). Differential analysis was performed to identify differentially expressed proteins (DEPs) between the groups, applying thresholds of fold change > 1.5 (upregulated) or < 0.67 (downregulated) with a P value < 0.05. This analysis identified 204 upregulated and 44 downregulated proteins (Figure 2B). The top 10 most significantly upregulated and downregulated DEPs are annotated in Figure 2C. Clustering analysis of DEP expression patterns effectively discriminated between SZ and HC groups, reflecting marked alterations in PBMC protein expression profiles in SZ patients (Figure 2D).

Figure 2 Analysis of quantitative results of data-independent acquisition proteomics. A: Histogram summarizing data-independent acquisition identification statistics for the schizophrenia (SZ) and healthy controls (HC) groups; B: Histogram of quantitative differences in protein expression between the SZ and HC groups; C: Volcano plot depicting quantitative protein differences between the SZ and HC groups; D: Cluster heatmap of differentially expressed proteins between the SZ and HC groups.

Integrative transcriptomics and proteomics analysis

Integration of transcriptomic and proteomic data revealed 6079 commonly expressed genes, including 25 DEGs differentiating SZ from HC groups (Figure 3A). Among these, haptoglobin (HP), lactotransferrin (LTF), and SERPING1 showed significant upregulation (Table 5). A Spearman correlation coefficient of 0.3928 was observed between significant DEGs and DEPs, indicating high data reliability (Figure 3B). Clustering analysis provided a visual representation of differential expression patterns for mRNAs and DEPs (Figure 3C). KEGG enrichment analysis highlighted the neuroactive ligand-receptor interaction pathway as potentially influential in SZ progression (Figure 3D). Based on these insights, subsequent studies prioritized the neuro-associated conduction protein pathway for further mechanistic exploration.

Figure 3 Integrated transcriptomics and proteomics analysis. A: Venn diagram showing overlapping transcriptomic and proteomic associations at the quantitative and differential expression levels (DE_Protein numbers represent genes corresponding to proteins); B: Correlation analysis (Spearman) between significantly differentially expressed genes and proteins; C: Cluster heatmap of transcriptomic and proteomic expression patterns for associated molecules; D: KEGG enrichment analysis of differentially expressed genes.

Table 5 The information of significantly differentially expressed genes.

Gene_id	DEGs	Protein_id	Protein_log2FC	Gene_log2FC
ENSG00000257017	HP	P00738	2.449962325	1.4294
ENSG00000012223	LTF	P02788	2.077758566	3.3348
ENSG00000091513	TF	P02787	1.587407166	-1.0837
ENSG00000149131	SERPING1	P05155	1.538266732	1.4989
ENSG00000125730	C3	P01024	1.428640888	1.0951
ENSG00000106804	C5	P01031	1.176272232	-1.0651
ENSG00000197249	SERPINA1	P01009	1.13095551	1.5216
ENSG00000177697	CD151	P48509	0.89951464	1.0137
ENSG00000122862	SRGN	P10124	0.77902409	1.4672
ENSG00000124762	CDKN1A	P38936	0.760160059	1.8984
ENSG00000079308	TNS1	Q9HBL0	0.745846947	2.6092
ENSG00000137507	LRRC32	Q14392	0.722864535	1.7123
ENSG00000064651	SLC12A2	P55011	0.716578449	1.4223
ENSG00000138448	ITGAV	P06756	0.706188096	1.4381
ENSG00000108405	P2RX1	P51575	0.662687029	1.6951
ENSG00000145685	LHFPL2	Q6ZUX7	0.662220906	3.638
ENSG00000151327	FAM177A1	Q8N128	0.640431398	1.0575
ENSG00000173083	HPSE	Q9Y251	0.620238989	1.8333
ENSG00000139970	RTN1	Q16799	0.601847201	-1.2624
ENSG00000235194	PPP1R3E	Q9H7J1	0.597450245	-1.0182
ENSG00000204390	HSPA1 L	P34931	0.593410635	-1.1346
ENSG00000108175	ZMIZ1	Q9ULJ6	-0.587965548	1.3241
ENSG00000166965	RCCD1	A6NED2	-0.613166445	-1.0529
ENSG00000139193	CD27	P26842	-0.730964915	-1.2033
ENSG00000168310	IRF2	P14316	-0.940903659	-1.239

DEGs: Differentially expressed genes.

Significantly higher expression levels of DEGs initially validated in small clinical samples

Protein expression levels of HP, LTF, and SERPING1 in PBMCs from clinical samples were analyzed via WB, revealing significant upregulation, consistent with findings from the integrated transcriptomics and proteomics analysis. Comparative analysis showed markedly higher expression of HP, LTF, and SERPING1 proteins in the SZ group relative to the HC group (Figure 4A and B). Similarly, RT-qPCR results confirmed a significant increase in mRNA expression of HP, LTF, and SERPING1 in the SZ group compared to the HC group (Figure 4C).

Figure 4 Expression of haptoglobin, lactotransferrin, and SERPING1 (n = 3). A and B: Detected by western blot, Immunoblotting image (A) and statistical bar chart of relative protein expression (B); C: Analyzed by real-time fluorescence quantitative PCR. ^aP < 0.001; HP: Haptoglobin; LTF: Lactotransferrin.

Significantly higher expression levels of the SERPING1 gene were verified by WB in all clinical participant samples

The study analyzed SERPING1 protein expression in 150 age- and gender-matched SZ patients and 150 HCs using WB, with representative results displayed in Figure 4A. Compared to the HC group, SERPING1 protein levels were markedly elevated in the SZ cohort (Figure 5A and B). Among the 150 SZ patients, the sample included 2 cases of the simple type, 56 of the undifferentiated type, and 92 of the paranoid type. A one-way ANOVA examining SERPING1 protein expression across these SZ subtypes showed no statistically significant differences (Figure 5C).

Figure 5 Expression of SERPING1 was detected by western blot (n = 150, including three samples of the results shown in Figure 4A; Here only a subset' western blot images are shown in Figure 5). A: Immunoblot image [schizophrenia (SZ) group samples numbered 1-150, healthy controls (HC) group samples numbered 301-450; representative western blot images are displayed]; B: Statistical bar chart of relative protein expression comparing SZ and HC groups; C: Statistical bar chart of relative protein expression across SZ subtypes. ^aP < 0.001, NS: Not significant.

Significant reduction in the expression level of SEEPING1 in neuronal cells in the CA1 region of rat hippocampus after lentivirus transfection

Lentiviral transfection with pre-constructed RNAi constructs effectively suppressed SERPING1 mRNA expression across all three silencing targets (si-SERPING1#1, si-SERPING1#2, and si-SERPING1#3) compared to the NC group (si-NC). Among the targets, si-SERPING1#1 demonstrated the greatest silencing efficiency (Figure 6) and was consequently chosen for further experimentation.

Figure 6 Analysis of SERPING1 expression in si-SERPING1 Lentiviral infection status by real-time fluorescence quantitative PCR (n = 3). ^aP < 0.01; ^bP < 0.001; NC: Negative control.

Significant increase in the expression level of the nerve-related conductive proteins calcium/calmodulin-dependent protein kinase II, cAMP response element-binding protein, and brain-derived neurotrophic factor after lentiviral transfection of neuronal cells in the CA1 region of the rat hippocampus

Transfection of target cells with the si-SERPING1#1 silencing lentivirus resulted in a marked upregulation of neuro-associated signaling proteins, including calcium/calmodulin-dependent protein kinase II (caMKII), cAMP response element-binding protein (CREB), and brain-derived neurotrophic factor (BDNF), when compared to the si-NC group (Figure 7).

Figure 7 Analysis of calmodulin-dependent protein kinase II, cAMP response element-binding protein, and brain-derived neurotrophic factor expression in lentiviral infection status by western blot (n = 3). A: Immunoblotting image; B: Statistical bar chart of relative protein expression. ^aP < 0.001; NC: Negative control.

Significant decrease in apoptosis and increase in cell viability of neuronal cells in the CA1 region of rat hippocampus after lentiviral transfection

Transfection of target cells with the si-SERPING1#1 silencing lentivirus resulted in a marked reduction in apoptosis rates and an enhancement in cell viability relative to the NC group (Figures 8 and 9). The data suggest that SERPING1 silencing enhances the proliferation of rat hippocampal neurons while suppressing apoptosis.

Figure 8 Apoptosis of rat neuronal cells which were treated with si-SERPING1 Lentiviral was detected by flow cytometry (n = 3). A: Flow cytometry scatter plot; B: Statistical bar chart of apoptosis rate. ^aP < 0.001; NC: Negative control.

Figure 9 The cell viability was analyzed by Cell Counting Kit-8 in rat neuronal cells treated with si-SERPING1 lentiviral (n = 3). ^aP < 0.001. NC: Negative control.

DISCUSSION

The pursuit of reliable biomarkers for early SZ diagnosis remains a significant challenge. To date, the identification of blood-based biomarkers for major psychiatric disorders has been limited and insufficient for clinical application[17]. Research has demonstrated that PBMCs and brain tissues exhibit similar expression profiles in certain signaling and metabolic pathways[18]. Furthermore, patients experiencing first-episode SZ have shown abnormal expression of synaptic genes, which are closely linked to the disorder, in PBMCs[19]. A study by Kozłowska et al[20] on a SZ case-control cohort revealed that PBMCs were actively involved in cytokine production when stimulated by both spontaneous mechanisms and phytohemagglutinin (PHA), suggesting that immune system dysfunction may underlie the pathophysiology of SZ. Petrikis et al[21] conducted a systematic phosphorylation analysis of Akt, GSK3-β, and S6 in freshly isolated PBMCs from first-episode psychiatric patients (FEP) and control subjects. The study identified two distinct signaling endophenotypes in FEP patients: a functional decline in GSK3-β, which exhibited a variable correlation with psychopathology and normalized after treatment, and a stable functional decline in mTORC1. These findings highlight peripheral dysfunction within the Akt/GSK3-β/mTORC1 pathway. To further investigate, transcriptomic and proteomic analyses were combined to examine PBMC samples from SZ patients. This integrated approach revealed upregulation of HP, LTF, and SERPING1 genes in SZ patients, confirmed through qRT-PCR and WB, which showed significantly increased mRNA and protein levels. Enrichment analysis implicated neuroactive ligand-receptor-related pathways in SZ progression. Validation with larger cohorts confirmed a significant elevation in SERPING1 expression in SZ patients. Silencing SERPING1 in vitro resulted in increased expression of neuronal pathway-related proteins CaMKII, CREB, and BDNF, reduced neuronal apoptosis, and enhanced cell viability, indicating a potential role of SERPING1 in SZ pathogenesis via the CaMKII-CREB-BDNF pathway.

The SERPING1 gene, located at 11q12-q13.1 on chromosome 11, encodes the plasma protease C1 inhibitor (C1-Inh), a key regulator of the classical complement and coagulation pathways. By targeting these systems, C1-Inh mitigates inflammatory and coagulation responses[22]. Research on SERPING1 has primarily centered on hereditary angioedema, where C1-Inh deficiency underpins the pathogenesis of types I and II hereditary angioedema[23]. Multiple variants of SERPING1 have been identified, including single nucleotide polymorphisms, small insertions or deletions, and larger structural changes such as insertions or duplications[24]. This study observed elevated SERPING1 expression in SZ but did not explore its structural variations, leaving the potential presence of pathogenic variants in SZ for future investigation. Cooper et al[25] performed a quantitative blood protein analysis in first-episode SZ patients, reporting significant upregulation of hemoglobin and C1-Inh, consistent with the proteomics findings presented here. Transcriptomic studies on post-mortem brain tissues from SZ patients revealed increased expression of immune-related and inflammatory genes, including SERPING1, implicating enhanced complement cascade activity in SZ pathophysiology[26]. Additionally, Allswede et al[27] analyzed PBMC samples from 129 adult SZ twins in Sweden, demonstrating that peripheral mRNA expression of complement genes, including SERPING1 and C5, uniquely correlated with variations in frontal cortical thickness.

The HP gene, located on chromosome 16 (16q22), is part of a multigene family comprising two alleles, HP1 and HP2, which encode the HP1 and HP2 proteins, respectively[28]. Research by Wan et al[29] identified altered HP protein expression and distinct genotype distributions in SZ patients within the Chinese Han population, suggesting a potential link between HP and SZ. Lee et al[30] reported increased HP gene expression in individuals with first-episode psychiatric disorders. The LTF gene, located on chromosome 3 (3p21) and also referred to as lactoferrin, belongs to the transferrin family. LTF encodes a multifunctional protein involved in antibacterial, antiviral, and anti-inflammatory processes, regulation of lipid metabolism, oxidative stress prevention, immune modulation, cell growth, and various enzymatic reactions[31]. Hällgren et al[32] observed significantly elevated serum LTF levels in SZ patients, independent of antipsychotic treatment. In this study, LTF exhibited substantial upregulation in both transcriptomic and proteomic analyses, highlighting its potential significance in SZ. However, the limited research on LTF in SZ emphasizes the need for further investigation, which may provide valuable insights into its role in SZ pathophysiology and guide future research directions.

The BDNF gene, located on chromosome 11p13 and spanning 70 kb with 11 exons[33], exhibits reduced expression in both first-episode and chronic SZ patients, suggesting its potential as a biomarker for cognitive impairment in SZ[34]. Polymorphisms in the BDNF gene have been linked to negative symptoms of SZ in studies conducted on the Han Chinese population[35]. Neurotrophic factors, including BDNF, are integral to neurodevelopment and synaptic plasticity[36,37]. The cAMP response CREB, a key regulator of neurotrophic responses, binds to specific sequences in the BDNF promoter in its phosphorylated form, thereby modulating transcription[38]. Guo et al[39] reported a correlation between increased BDNF mRNA expression and elevated phosphorylated CREB levels. In the hippocampus, calcium/CaMKII is critical for learning and memory consolidation. Evidence suggests that diminished CaMKII activity underlies widespread brain dysfunction, contributing to neuropsychiatric conditions such as addiction, SZ, depression, epilepsy, and other neurodevelopmental disorders, likely due to disrupted glutamate signaling and impaired neural plasticity[40]. Research on CaMKIIα heterozygous knockout mice has identified EEG and behavioral changes characteristic of SZ subtypes and intellectual disability[41]. Rodent studies have highlighted the therapeutic potential of enhancing CaMKII activity to ameliorate cognitive deficits associated with SZ[42]. Silencing SERPING1 significantly elevated the expression of neuronal pathway proteins, including CaMKII, CREB, and BDNF, suggesting its involvement in SZ through the CaMKII-CREB-BDNF signaling cascade and its potential as a novel target for SZ diagnosis and treatment. Supporting this mechanism, Lee et al[43] demonstrated that the atypical antipsychotic olanzapine promoted BDNF gene transcription by activating CREB via the PKA, PI3K, PKC, and CaMKII pathways, conferring neuroprotective effects. Cognitive impairments in SZ patients are closely associated with synaptic dysfunction. A study simulating synaptic damage in rats linked neuronal injury to the CaM/CaMKII/CREB/BDNF pathway, implicating postsynaptic ion channels and synaptic plasticity-related proteins[44]. Furthermore, theobromine treatment improved working memory in rats by upregulating the CaMKII/CREB/BDNF pathway in the medial prefrontal cortex[45]. While most research on this signaling cascade centers on depression, studies specifically addressing its role in SZ remain limited.

WB was employed to analyze SERPING1 protein expression in PBMCs from 150 SZ patients and 150 HCs, with comparisons across distinct disease subtypes. The analysis revealed significantly elevated SERPING1 protein levels in SZ patients relative to HCs. However, no notable differences were observed among the subtypes, possibly due to the limited sample size. SZ, a heterogeneous disorder influenced by polygenic and multifactorial contributions, does not adhere to Mendelian inheritance patterns. The observed disconnect between genotype and phenotype may account for the absence of a significant association between SERPING1 protein levels and clinical subtypes. An RNAi-silenced lentivirus targeting the SERPING1 gene was constructed and transfected into rat neuronal cells, resulting in markedly enhanced neuronal cell proliferation and reduced apoptosis. This intervention also increased the expression of channel proteins CaMKII, CREB, and BDNF, highlighting the neuroprotective role of the CaMKII-CREB-BDNF signaling pathway. Few studies have explored the influence of the SERPING1 gene on neuronal proliferation and apoptosis, placing this research within a nascent area of investigation. Study limitations include the exclusive enrollment of recurrent SZ patients undergoing long-term antipsychotic therapy, which may have introduced confounding variables into the peripheral blood omics analysis. Additionally, the relatively small cohort size may have led to selection bias, restricting the broader applicability of the findings to the general SZ population.

CONCLUSION

This study employed integrated transcriptomic and proteomic analyses to reveal significant upregulation of HP, LTF, and SERPING1 in PBMCs from SZ patients. Clinical sample validation further confirmed elevated SERPING1 expression levels. Silencing SERPING1 led to a pronounced increase in neural pathway-associated channel proteins, including CaMKII, CREB, and BDNF. Functional assays indicated that SERPING1 exerted anti-proliferative and pro-apoptotic effects on neuronal cells. These results suggest that SERPING1 may regulate neuronal cell proliferation via the CaMKII-CREB-BDNF pathway, potentially contributing to the development and progression of SZ. This discovery offers a foundation for further research and highlights potential therapeutic targets for the disease.

ACKNOWLEDGEMENTS

Acknowledgment is extended to the Hainan Province Clinical Medical Center for its support and to the reviewers for their valuable insights.