Background

Epilepsy, a central neurological disorder characterized by recurrent, spontaneous seizures, affects over 68 million people globally [1]. The primary causes of epilepsy are abnormal discharges in the hippocampus or cerebral cortex [2, 3]. While approximately 70% of seizures can be effectively managed with approved antiepileptic drugs [4], the remaining 30% are resistant to pharmacotherapy, resulting in significant psychological and physical burdens for patients [5].

The genetic determinants of epilepsy are largely unknown. There is evidence suggesting that genetic factors contribute to both generalized and focal epilepsies [6]. Mutations in certain genes have been associated with epilepsy [6, 7], but the role of common polymorphisms in epilepsy is still unclear [8, 9]. Several recent studies have identified new epilepsy loci [10,11,12,13,14,15], and further expanded analysis has revealed additional new loci for epilepsy [16]. Moreover, gene therapy has been reported as a potential treatment for refractory focal epilepsy [17].

Bioinformatics-based studies using microarray analysis are crucial for identifying disease-related gene expression patterns [18,19,20,21]. In this study, we analyzed gene expression between epileptic rat models and healthy samples to investigate their potential association with epilepsy. We further conducted integrative analyses with gene co-expression, expression quantitative trait loci (eQTL), protein–protein interaction (PPI) networks, and drug–gene interaction to clarify the potential role of newly discovered epilepsy-associated genes in the diagnosis and treatment of epilepsy.

Materials and methods

Gene Expression Omnibus datasets

In this study, we utilized two independent datasets from the Gene Expression Omnibus (GEO). These datasets, GSE44031 [22] and GSE1834 [23], were analyzed using the GPL9207 platform (Duke Operon Rat 27 k V3.0 printed oligonucleotide array) and the GPL85 platform ([RG_U34A] Affymetrix Rat Genome U34 Array), respectively. For detailed information on these datasets, please refer to the GEO database (http://www.ncbi.nlm.nih.gov/geo).

Differentially expressed genes (DEGs) analysis

Differentially expressed genes (DEGs) were identified using R software. The gene expression levels were extracted from the GSE44031 dataset, which includes microarray data from eight hippocampal tissue samples from epileptic rats induced by kainic acid injection, and four samples from rats injected with phosphate buffer saline. These differentially expressed genes were further validated in the GSE1834 dataset, which contains data from 15 hippocampal tissue samples from epileptic rats induced by kainic acid injection, and 15 samples from rats injected with phosphate buffer saline. A statistical significance threshold was set at an adjusted p value ≤ 0.05 and |logFold change|≥ 1. The overlapping DEGs were selected for further analysis using a Venn map, which was created using the Venn online tool (http://bioinformatics.psb.ugent.be/webtools/Venn/).

Gene co-expression analysis

Gene co-expression analysis is a method used to discover the regulatory relationships between genes and subsequently identify potential disease candidate genes. Given that gene expression is tissue-specific, we investigated the co-expression relationships between DEGs and known epilepsy genes in human brain tissues. The gene co-expression database for brain tissue serves as an effective tool for analyzing the co-expression relationships of genes associated with brain diseases [24].

eQTL analysis

To determine the impact of epilepsy risk single-nucleotide polymorphisms (SNPs) on the expression of DEGs in human brain tissue, we performed eQTL analysis. This analysis was conducted using BRAINEAC (http://www.Braineac.org/) [25], a tool designed to investigate the association between SNPs and the expression of their target genes in human brain tissue.

Evaluation of DEGs in epilepsy treatment

To identify new antiepileptic drug targets and facilitate the translation of these findings into clinical therapy, we first identified approved antiepileptic drug therapeutic targets using DrugBank5.0 and the Therapeutic Target Database 2020 [26, 27]. We then conducted PPI analysis to explore the relationship between newly discovered epilepsy-related genes and approved antiepileptic drug target genes. This PPI analysis was performed using the STRING database (https://string-db.org/cgi/input.pl) [28], and a PPI network was constructed using Cytoscape software [29]. The findings from the PPI analysis were further validated by a drug–gene interaction analysis using the Drug–Gene Interaction Database [30].

Result

Analysis of DEGs in epilepsy rat models and controls

Based on the selection criteria for DEGs outlined in the “Methods” section, 91 DEGs were selected from the GSE44031 dataset (see Additional file 1: Table S1), and 425 DEGs were selected from the GSE1834 dataset (see Additional file 2: Table S2). As depicted in Fig. 1, five genes overlapped between the two datasets (see Additional file 3: Table S3). Specifically, among these five DEGs, the expression regulation of one gene was inconsistent across the two datasets, leading to its removal. The expression of the remaining four genes was upregulated in both datasets (Fig. 2), with two of them (GFAP and S100A4) previously reported to be associated with epilepsy [31, 32]. Consequently, we selected two newly discovered candidate genes, the protein tyrosine phosphatase receptor-type O gene (PTPRO) and the growth arrest and DNA damage inducible alpha gene (GADD45A), for further analysis. The Volcano map of DEGs was created using the online volcano plotting tool (http://sangerbox.com/AllTools?tool_id=9699135).

Fig. 1
figure 1

DEGs were identified from GSE44031 and GSE1834 gene expression profiling datasets based on adjusted p value < 0.05 and |logfold change|≥ 1. The two datasets share 5 overlapping DEGs

Full size image
Fig. 2
figure 2

The volcano plot illustrates DEGs. The volcano plot illustrates DEGs between epilepsy and control after analysis of the GSE44031 gene expression profiling datasets

Full size image

Gene co-expression analysis

As genetic research advances, an increasing number of epilepsy-related genes have been identified [33]. In this study, we utilized these reported epilepsy-related genes along with the newly discovered genes for gene co-expression analysis. The analysis revealed that PTPRO was co-expressed with RBP4, NDN, PAK3, FOXG1, and IDS in human brain tissue. Similarly, GADD45A was found to be co-expressed with LRRK2 in human brain tissue (Fig. 3).

Fig. 3
figure 3

Analysis of gene co-expression in human brain. PTPRO was co-expressed with RBP4, NDN, PAK3, FOXG1, IDS, and IDS, and GADD45A was co-expressed with LRRK2 in human brain tissue

Full size image

eQTL analysis

In this study, we focused on SNPs that have been previously reported to be associated with epilepsy for eQTL analysis. Specifically, we explored whether the SNPs rs4794333, rs68082256, rs11943905, rs12185644, rs6432877, rs2212656, rs1402398, rs11890028, rs887696, rs1044352, rs13200150, and rs4671319, which have been associated with epilepsy in a multicenter study [16], regulate the expression levels of the newly discovered genes in human brain tissue. Interestingly, our findings suggest that these SNPs do indeed regulate the expression levels of PTPRO and GADD45A in human brain tissue (Table 1).

Table 1 Epilepsy risk SNPs regulate GADD45A and PTPRO expression in human brain tissue
Full size table

Evaluation of PTPRO and GADD45A in epilepsy treatment

To explore the potential role of PTPRO and GADD45A in epilepsy treatment, we first conducted an interaction analysis between the proteins encoded by PTPRO and GADD45A and the targets of approved antiepileptic drugs (Table 2). We found that the protein encoded by GADD45A had direct interactions with the targets of approved antiepileptic drugs, including PPARA and PPARG (Fig. 4). Further drug–gene interaction analysis revealed that GADD45A interacts with genistein and cisplatin.

Table 2 115 genes encoding proteins targeted by approved antiepileptic drugs
Full size table
Fig. 4
figure 4

The PPI network of PTPRO and GADD45A and the genes whose encoded protein targeted by antiepileptic drugs. The proteins connected by the red line mean that they have direct interactions

Full size image

Discussion

In this comprehensive analysis of epilepsy, we identified two new potential genes, PTPRO and GADD45A, and highlighted their crucial roles in epilepsy. Our results suggest that GADD45A might serve as a potential therapeutic target for epilepsy. The Consortium on Complex Epilepsies has made significant strides in identifying several loci with genome-wide significance for epilepsy [15, 16]. Research aimed at understanding the role of genes in epilepsy will contribute to a deeper understanding of the genetic architecture of epilepsy. To the best of our knowledge, the association of PTPRO and GADD45A with epilepsy has not been evaluated previously.

In this study, we initially identified five DEGs that overlapped in the GSE44031 and GSE1834 gene expression profiles. Further analysis revealed that two of these DEGs, PTPRO and GADD45A, had not been previously reported to be associated with epilepsy. Gene co-expression analysis uncovered specific regulatory relationships between these two newly discovered epilepsy-associated genes and known epilepsy genes (Fig. 3). These findings suggest that these potential genes are functionally associated with the reported epilepsy genes, further underscoring the role of these newly discovered genes in epilepsy. This could facilitate the diagnosis and treatment of epilepsy and may provide a new direction in the understanding of the disease.

SNPs located in noncoding regions can influence disease risk by regulating the expression of their target genes [34, 35]. Our eQTL analysis revealed that several noncoding SNPs associated with epilepsy risk alter the expression of PTPRO and GADD45A in brain tissue (Table 1). Moreover, these SNPs have been reported to be associated with epilepsy [16]. Therefore, we hypothesize that these SNPs may contribute to epilepsy risk by regulating PTPRO and GADD45A expression in human brain tissue. These findings further elucidate the mechanisms of PTPRO and GADD45A in the pathogenesis of epilepsy.

PTPRO, located on chromosome 12, encodes the receptor-type tyrosine-protein phosphatase O. It is highly expressed in the brain and promotes the formation of excitatory synapses [36]. PTPRO also plays a role in regulating the development and function of the sensory nervous system [37]. A genome-wide study has shown that PTPRO is associated with learning and memory [38]. On the other hand, GADD45A has been found to influence cortical evolution and diversity depending upon its expression levels [39, 40]. Therefore, both PTPRO and GADD45A may play significant roles in the pathogenesis of brain diseases.

PPI networks play a crucial role in drug target discovery and present drug discovery process [41]. Our PPI analysis revealed that the protein encoded by GADD45A directly interacts with the targets of approved antiepileptic drugs, PPARA and PPARG (Fig. 4). By integrating data from DrugBank5.0 and the Therapeutic Target Database 2020, we found that PPARA and PPARG are targets of Valproate. Furthermore, we discovered that GADD45A interacts with genistein and cisplatin. Genistein has been implicated in antiepileptic effects [42, 43], while cisplatin has been reported to induce seizures [44, 45]. Therefore, the results of the PPI networks and drug–gene interaction further underscore the significant role of GADD45A in epilepsy therapy. While recent studies have discovered new epilepsy-related genes [46, 47], our study not only identified new epilepsy-related genes but also explored the potential mechanisms of these new genes in the pathogenesis and treatment of epilepsy, which could be more conducive to the transformation of clinical application.

While our study integrated data from gene expression, gene co-expression, eQTL, PPI network, and drug–gene interaction to uncover the role of PTPRO and GADD45A in epilepsy diagnosis and therapy, there are still some limitations. First, further validation of our findings in independent populations is necessary to strengthen our conclusions. Second, additional functional characterization would help to better understand the mechanisms of PTPRO and GADD45A in the pathogenesis and treatment of epilepsy.

Conclusion

In summary, our findings underscore the potential of PTPRO and GADD45A as promising targets for the diagnosis and treatment of epilepsy.