1. Introduction
Multiple sclerosis (MS) is a chronic autoimmune disease that affects the central nervous system (CNS). Due to complex pathogenesis and lack of specific biomarkers, diagnosis of MS is a challenge. Recent data show that the incidence of MS is increasing globally. Early diagnosis of MS reduces the burden of disability-adjusted life years and associated health care costs (Solomon et al., 2023).
Extensive exploration has demonstrated that MS is thought to be caused by systemic immune activation of autoimmune mechanisms against CNS components. In MS, inflammation is regulated by interactions between several immune cells, such as T and B cells, macrophages, and CNS glial cells (microglia and astrocytes), as well as antigens that react against myelin, especially myelin basic protein and myelin oligodendrocyte glycoprotein (Pietrasik et al., 2021). Additionally, Th1 and Th17 cells can cross the blood-brain barrier (BBB) and migrate to the CNS, then activate microglia and secretion of inflammatory cytokines, such as tumor necrosis factor-α (l), interleukin (IL)-1β and IL-6 (Murphy et al., 2010). Diagnosis of MS is based on clinical and radiological assessment, and current effective treatments target the peripheral immune system (Healy et al., 2022a; Luo et al., 2017; Inojosa et al., 2021). The diagnostic criteria for MS have limitations of sensitivity and specificity (Ian McDonald et al., 2001). If a specific test is lacking (Tavazzi et al., 2020), some patients may be misdiagnosed or not diagnosed  (Ian McDonald et al., 2001;  Tavazzi et al., 2020). 
Noncoding RNAs (ncRNAs) are RNA molecules that do not encode proteins but play essential roles in gene expression regulation. Investigations indicate that ncRNAs are involved in the pathogenesis and progression of MS (Jalaiei et al., 2021; Zheleznyakova et al., 2021). The roles of ncRNAs in regulating immune cells and pathways, as well as their involvement in the neurodegeneration process and MS progression, are highlighted by researchers (Elkhodiry & El Tayebi 2021; Yousuf & Qurashi, 2021; Piket et al., 2019). In addition to their diagnostic potential, ncRNAs may serve as therapeutic targets for MS treatment. It may be effective in personalizing medicine (Gupta et al., 2019; Piket et al., 2019; Zheleznyakova et al., 2021; Chen et al., 2022). MicroRNAs (miRNAs) are small noncoding RNA molecules that regulate various cellular processes, including inflammation and immune response. They have emerged as potential biomarkers for MS, as they are stable, detectable, and quantifiable in various biological fluids, such as blood and cerebrospinal fluid. Several studies have shown that the expression levels of miRNAs are altered in MS patients compared to healthy controls or other neurological diseases. Moreover, some miRNAs have been associated with clinical features, such as disease duration, disability score, relapse rate, lesion load, inflammation, neurodegeneration, and treatment response (Gao et al., 2021; Yang et al., 2018; Slota et al., 2019; Yang et al., 2018; Gao et al., 2021; Minutti-Zanella., 2022). 
The application of bioinformatics in the interpretation of genomics outcomes has attracted researchers’ attention. Network analysis is a bioinformatics tool suitable for interpreting genomics data (Ni et al., 2014). There are several documents about the application of network analysis in exploring MS’s molecular aspects (Hao et al., 2022). Protein-protein interaction (PPI) network analysis is a computational method to identify interactions between proteins and understand their functional relationships. In a published study, researchers used PPI network analysis to identify hub-long ncRNAs and potential drugs for MS. They constructed a PPI network using differentially expressed mRNAs and identified four modules enriched in immune-related pathways. They identified three key long ncRNAs (LINC00649, TP73-AS1, and MALAT1) associated with MS (Chakraborty et al., 2022; Gabrielli & Verderio, 2023; Khan et al., 2016; Yadav et al., 2023; Wang et al., 2022). The present study investigates the possible role of miRNAs and their related protein in diagnosing MS disorder via network analysis using data from the GEO database.

2. Materials and Methods 

Data collection
MS microRNA data had been selected from the GEO database (GSE124900). Data were produced using the NGS approach. The 16 peripheral blood-extracted microRNA profiles from relapsing-remitting MS patients and 8 healthy controls (Baulina et al., 2019) were selected for more analysis. UMAP analysis revealed that the samples were not separated via the performed assessment. So, the suitable samples, including 4 patients and 4 controls, were candidates for evaluation. In addition to the microRNA analysis, related proteins have been investigated in the literature.

Pre-evaluation analysis
The GEO2R program was applied to evaluate data. Visualization tools such as Boxplot, Venn diagram, and volcano plot were employed to gain insights into potential microRNA expression signatures and identify differentially expressed microRNAs. These analytical tools are crucial in minimizing bias in microRNA expression levels and identifying potential diagnostic targets (microRNA) for future interventions. These visualization tools visually represent the data, enabling researchers to identify patterns and trends that may not be immediately apparent from raw data. The significant differentially expressed miRNAs with a P adj<0.05 were selected for further analysis.

PPI network analysis 
Actions (activation, expression, catalysis, and post-translation) between the explored proteins were assessed via an action map using Cytoscape software, version 3.7.2. Furthermore, PPI network analysis has been employed to analyze the related proteins. The associated proteins were included in an interactome via the STRING database using Cytoscape software, version 3.7.2 (Shannon et al., 2003). The nodes were connected via undirected edges. The utilization of PPI network analysis has provided a comprehensive understanding of the binding relationship among proteins and has facilitated the identification of potential key proteins involved in diagnosing MS.
Integrating multiple analytical approaches, including miRNA profiling, related protein investigation, and PPI network analysis, has enabled a more comprehensive understanding of the molecular mechanisms underlying MS.
Statistical analysis
P adj<0.05 was applied to find the significant differentially expressed microRNAs. A confidence score of 0.4 was considered to form a PPI network.

3. Results
The extracted significant differentially expressed miRNAs are presented in Table 1.



As shown in Table 1, 18 miRNAs are eligible for more investigation. Box plot (Figure 1) has been utilized to maintain that our selected miRNAs are trustworthy for further investigation because their median line was in the same position. The volcano plot was used to identify the significantly differentially expressed (DE) miRNAs between the two groups.



Figure 2 shows several significant DE miRNAs discriminate patient samples from controls.



Venn diagram visualizes 18 significant DE miRNAs that differentiate the compare groups (Figure 3).



Furthermore, 31 proteins related to 18 microRNAs, including FBL, NCL, NHP2L1, PTBP1, YBX1, NOP58, PRPF8, SNRNP70, U2AF2, STAT3, SP1, SNRPA, TLR4, PIK3CA, BOP1, DUSP1, TPD52, ZNF580, NECAP2, ELF3, GARS, BCL2, TRMT10A, NR2E3, KLF3, ZNF143, HCFC1, EEFA1, SF3B1, NHP2, and NPM1 were identified from literature. An action map was constructed to illustrate the connections among the 31 related proteins. Each color in the diagram represents a different protein-protein function: Green, yellow, black, and purple, which consists of activation, expression, catalysis, and post-translation. It should be noted that functional interactions for 12 proteins have remained unknown. A thorough investigation determined that NPM1 is the most highly connected protein in this PPI diagram. NPM1 is connected with FBL, YBX1, NOP58, PRPF8, STAT3, BOP1, PIK3CA, and DUSP1 (Figure 4).



The PPI network, including 31 proteins, is represented in Figure 5. As is detected in Figure 5, 27 proteins are connected in the subnetwork, while the remaining individuals are isolated nodes. Nodes of the network are visualized based on degree value.



Considering the results of the action map and PPI network analysis, the carousal proteins were selected for more study. A list of the carousal proteins and their biological descriptions are presented in Table 2.



4. Discussion
MS is a complex autoimmune disease that affects the CNS. The exact mechanisms underlying MS pathogenesis are still elusive. miRNAs, RNA molecules that do not encode proteins but regulate gene expression, have been suggested to play a role in MS. Moreover, miRNAs are stable biomarkers in peripheral blood that make them appropriate candidates for diagnosing MS. They can also reflect the status of immune cells and inflammatory pathways that are relevant to MS pathogenesis, distinguish MS patients from healthy controls and other disease controls, as well as different MS subtypes and stages, and predict the response to treatment and the risk of relapse or progression in MS patients (Gandhi 2015; LoPresti 2022; Ziemssen et al., 2019). Therefore, by analyzing the related proteins of miRNAs, we can find critical pathways that can be the key to finding new treatments and diagnosing methods for MS.
The present study used data from the GEO database to identify miRNAs differentially expressed in MS patients compared to healthy controls. A total of 18 significantly dysregulated miRNAs were selected, and 31 related proteins involved in MS disease were determined for analysis. Pre-evaluation of data indicates that the analysis is valid. An action map was used as a tool to screen data. Based on action map findings, NCL, U2AF2, SNRNP70, YBX1, and STAT3 have considerable relationships with others. Activation of SNRNP70 by U2AF2 is an important event because it has a role in neuromuscular synapses, and U2AF2 also catalyzes and has an effect on post-translation of PTBP1 which has been investigated in literature as an MS biomarker (Masaki et al., 2020). STAT3 can be a potential candidate because of its relationship with IL-6, which was upregulated in MS patients (Razia et al., 2023).
SNRNP70 plays an important role in regulating motor axonal growth, nerve-dependent acetylcholine receptor clustering, neuromuscular synaptogenesis, and a protective role for a limited number of axonal transcripts, preventing them from degradation. Moreover, non-nuclear SNRNP70 can locally regulate splice variants of transcripts such as agrin, thereby locally controlling the formation of synapses (Lubelsky et al., 2021; Nakaya, 2020; Nikolaou et al., 2020).
On the other hand, YBX1 affects both PTBP1 and U2AF2 and also expresses PRPF8, which has a role in the post-transcription of U2AF2. It has a role in neuroinflammation signaling (Kloetgen et al., 2020). The relationship and effect of STAT3 (which has been investigated as an MS biomarker) and TLR4 are remarkable; they both have positive regulation of each other. It is essential to mention that TLR4 is a drug target because of its main role in inflammation and T cell and macrophage activities (Ahuja et al., 2020; Li et al., 2019; Mukherjee et al., 2019). TLR4 and BTK are interconnected in the context of immune signaling and inflammation, and both have been implicated in the pathogenesis of MS; TLR4 also has an impact on BTK, which is an MS drug target (Geladariset al., 2022; Krämer et al., 2023). The other protein, PIK3K, can be a potential biomarker as it has a relationship with mir-21 in activating the PI3K/AKT pathway as a possible therapy for treating MS (Manian et al., 2021; Yin et al., 2022). It is connected with STAT3 by its post-translation modification and expression.
PPI network analysis was performed to identify the most relevant proteins interacting with these microRNAs. Based on degree value, the results indicated that several proteins, including FBL, NCL, PTBP1, SNRNP70, YBX1, PRPF8, and NPM1, are important ones. Except for NPM1, the other vital nodes of the PPI network are highlighted in the action map. Other studies have also reported these proteins. For instance, PTBP1 has been shown to regulate alternative splicing of genes involved in immune response and inflammation (Babenko et al., 2022; Hensel et al., 2022), key pathways in MS pathogenesis (Hecker et al., 2019; Masaki et al., 2020). NPM1 has been implicated in regulating microglia activation, a hallmark of MS (Healy et al., 2022b; Jäntti et al., 2022; Sen et al., 2022).
PTB1 has been shown to regulate the expression of BDNF (Notaras & van den Buuse, 2019), which has been shown to promote neuronal growth and survival in the CNS, synaptic plasticity, and mitochondrial biogenesis, making it a promising biomarker in neurodegenerative conditions (Ahmed et al., 2021; Pisani et al., 2023). 
SNRNP70 has been shown to regulate alternative splicing of genes involved in neuronal function and synaptic plasticity, which are disrupted in MS (Ksiazek-Winiarek et al., 2015; Nakaya, 2020; Nikolaou et al., 2020). YBX1 is identified as a common high central gene in PPI networks corresponding to both type 1 diabetes and MS (Safari-Alighiarloo et al., 2020). As discussed before, NOP58 is another candidate with a high degree and is also activated by BOP1 and NCL. NOP58 has a connection with PTBP1 and PRPF8. 

5. Conclusion
Moreover, as PTBP1, STAT3, and TLR4 have been suggested in the literature as potential biomarkers for MS, based on our investigation with bioinformatics approaches, they have roles in regulating the expression and function of NCL, NOP58, SNRNP70, U2AF2, YBX1, PRPF8, BOP1, and PIK3K proteins. These proteins involve various cellular processes relevant to MS pathogenesis, such as RNA splicing, transcription, inflammation, and apoptosis. Therefore, we propose a new biomarker panel consisting of PTBP1, STAT3, and TLR4, as well as SNRNP70, YBX1, PTB1, and PI3K as candidate drug targets. In addition, we suggest that monitoring the effect of targeting TLR4 on STAT3 activity could be a useful strategy for developing novel therapeutics for MS. Further experimental and theoretical investigations are suggested to validate the proposed biomarker panel and drug targets. Limited sources of databases and samples are limitations of this study.

Ethical Considerations

Compliance with ethical guidelines
This study was approved by the Research Ethics Committee of Shahid Beheshti University of Medical Sciences, Tehran, Iran (Code: IR.SBMU.RETECH.REC.1402.727).

Funding
This study was funded by the Student Research Committee, Shahid Beheshti University of Medical Sciences, Tehran, Iran (Project No.: 1402-63026).

Authors' contributions
 All authors contributed equally to the conception and design of the study, data collection and analysis, interception of the results and drafting of the manuscript. Each author approved the final version of the manuscript  for submission.

Conflict of interest
The authors declared no conflict of interest.

Acknowledgments
The authors appreciate the “Student Research Committee” and “Research & Technology Chancellor” at Shahid Beheshti University of Medical Sciences, Tehran, Iran  for their financial support of this study. 




References