This study investigates the diagnostic and prognostic potential of non-coding RNAs, specifically miR-382-5p and lncRNA Ftx, besides NRG1 protein expression, in distinguishing MS, NMO, and healthy controls. The patients in this study were under treatment. B cell therapy has been administered to patients with SPMS, and PRMS. B cell therapy, also known as B cell ablative therapy, targets cells called B cells that damage the nerves in the brain and spinal cord. Our findings highlight the differential expression of these biomarkers across groups. miR-382-5p was significantly upregulated in NMO compared to MS subtypes and healthy controls, while lncRNA Ftx exhibited a marked decrease in NMO. However, it demonstrated the potential to distinguish both MS and NMO from healthy controls. indicating its potential as a biomarker for identifying these conditions. The study further demonstrated that miR-382-5p plays a role in distinguishing NMO from MS, with robust diagnostic performance in differentiating MS from healthy controls. However, its ability to differentiate NMO from healthy controls was limited. LncRNA Ftx, while showing potential in distinguishing NMO from both MS and healthy controls, had limited diagnostic and prognostic utility in MS subtypes, with its expression levels being particularly low in NMO. Correlation analysis revealed complex, independent regulatory interactions between these biomarkers, with miR-382-5p and NRG1 showing significant negative correlation specifically in the MS group.

The findings underscore the potential of miR-382-5p and lncRNA Ftx as biomarkers for distinguishing NMO from MS, but their diagnostic utility is not without limitations. While miR-382-5p demonstrated high sensitivity for distinguishing MS from healthy controls, its specificity for NMO was suboptimal. LncRNA Ftx, although demonstrating high specificity in certain comparisons, exhibited low sensitivity for both MS and NMO. The results also suggest that combining these biomarkers could enhance diagnostic accuracy, though further studies with larger cohorts are necessary to validate their clinical utility.

Our findings partially align with those of Kataria et al., who demonstrated dysregulation of NRG1, particularly the NRG1β1 isoform, in MS pathology. However, we observed significant differences in total NRG1 levels between healthy controls and patients with NMO, MS (RRMS and SPMS), while Kataria et al. did not report such differences between MS patients and healthy controls. These discrepancies may stem from differences in the isoforms measured (total NRG1 vs. NRG1β1), study populations, and sample sizes, as well as methodological variations in quantification and processing techniques [19].

Additionally, prior studies have highlighted the protective, anti-inflammatory, and proliferative roles of NRG1 in neuroinflammatory diseases and various conditions, these include traumatic injury [25], LPC-induced focal demyelinating lesions [18], schizophrenia [26], psychosis [27], Stroke [28], autism [29], symptomatic epilepsy [30] and Alzheimer's disease [31], also, different tumors including lung cancer [32], breast cancer [33], esophageal squamous cell carcinoma [34]. These findings suggest that NRG1 is not only a marker of disease progression but also a potential therapeutic target for modulating inflammatory and neurodegenerative processes, as demonstrated in both MS and NMO by our current study.

Previous efforts to identify serum biomarkers for distinguishing NMO from MS have proposed candidates such as VEGF, MPIF-1, and NrCAM. However, these studies have largely relied on the presence of AQP4-IgG titers to confirm NMO diagnosis [35]. Additionally, analyses of cytokines, chemokines, and soluble receptors were conducted in MS patients and AQP4-IgG-positive NMO patients, but not in AQP4-IgG-negative NMO patients [36]. This approach is limited by the fact that a significant number of NMO patients are seronegative for AQP4-IgG, increasing the risk of false-negative results [35]. Our study advances this field by focusing on serum NRG1 levels in both AQP4-IgG-positive and negative NMO patients. We observed significantly lower NRG1 concentrations in NMO patients compared to MS patients, suggesting that NRG1 may serve as a more reliable biomarker for distinguishing between these conditions. This could address the limitations of current diagnostic approaches and improve diagnostic precision. Moreover, NRG1’s potential as a biomarker could pave the way for more personalized treatment strategies and highlight its promise as a therapeutic target in both MS and NMO. However, the efficacy of NRG1 as a prognostic biomarker was limited, particularly for distinguishing between RRMS and SPMS.

Non-coding RNAs, including lncRNAs and miRNAs, can both positively and negatively, directly and indirectly, influence protein expression, such as NRG1, positioning them as key players in the pathogenesis of neurodegenerative diseases, including MS [37]. For instance, numerous miRNAs have been identified as significantly contributing to the pathogenesis of MS and NMO, including hsa-miR-16, hsa-miR-15a, hsa-miR-124, hsa-miR-26b, hsa-miR-181, hsa-miR-633, hsa-miR-155, and hsa-miR-129–2-3p [38,39,40,41,42].

Notably, miR-382-5p has also been shown to target NRG1 in spinal cord injury [22] and its involvement extends to the pathogenesis of various neurological disorders, including migraine [43], schizophrenia [44], and cerebral ischemia–reperfusion injury [45]. Beyond its role in CNS disorders such as spinal cord injury [22, 46], glioma cell tumorigenesis [47], and depression [48], miR-382-5p also impacts other conditions, including liver regeneration, inflammation and various cancers [49,50,51,52,53].

Our study highlights the potential of miR-382-5p as a valuable biomarker for distinguishing between MS subtypes and healthy controls, with significant downregulation observed in patients with MS, SPMS, and RRMS. This downregulation suggests that miR-382-5p could serve as an effective marker for accurately differentiating MS subtypes from healthy individuals.

Notably, miR-382-5p also demonstrated the ability to differentiate between NMO and other MS subtypes, with significant upregulation in the NMO group compared to various MS subtypes. This upregulation underscores its potential as a diagnostic marker for NMO. The contrasting expression patterns—overexpression of miR-382-5p in NMO and downregulation in MS—may indicate distinct roles for this miRNA in the pathogenesis of MS and NMO, highlighting the complexity of these disorders. Furthermore, miR-382-5p was significantly different between RRMS and SPMS, so it could also serve as a prognostic marker for progression to SPMS in MS patients, further enhancing its diagnostic and prognostic utility.

Long non-coding RNAs (lncRNAs), which can act as competitive endogenous RNAs (ceRNAs) or enhance miRNA activity, play a crucial role in modulating gene expression. This dual functionality likely contributes to the complex regulatory mechanisms observed in MS and NMO pathogenesis, particularly in relation to miRNA interactions [54, 55]. Several lncRNAs have been implicated in the pathogenesis of MS and NMO, including TUG, LRRC75A-AS1, LINC00293, BDNF-AS, GAS5, lncDDIT4, and MALAT1 [56, 57].

LncRNA Ftx has been implicated in various disorders, including neurological conditions such as spinal cord injury (SCI) [22] and stroke [58]. Beyond the CNS, it plays a role in malignancies including gastric cancer [59], colorectal cancer [60], glioma [61], and pancreatic cancer [62]. Furthermore, it is involved in non-cancerous diseases like asthma [63] and endometriosis [64]. Notably, LncRNA Ftx has also been shown to target miR-382-5p linking it to regulatory pathways relevant to both neurological injury and inflammation [22]. Moreover, lncRNA Ftx has been implicated in the pathogenesis of various neurological disorders, including epilepsy [65], cerebral ischemia–reperfusion injury [66], and autism [67].

In our study, although LncRNA Ftx showed no significant diagnostic or prognostic value in distinguishing between healthy controls and other MS subtypes, it notably demonstrated ability to differentiate NMO from both healthy controls and MS subtypes. The significant downregulation observed in the NMO group suggests its potential as a differential diagnostic marker for NMO. However, LncRNA Ftx did not exhibit prognostic utility for predicting progression to SPMS in MS patients.

Interestingly, the direct inverse relationship between NRG1 and miR-382-5p expression, previously observed in spinal cord injury (SCI) [22], was seen in MS group in our study but not reach significant level in Spearman correlation analysis in the rest of the groups. Additionally, the expected direct positive relationship between NRG1 and lncRNA Ftx expression, previously identified in spinal cord injury (SCI) [22], was not replicated in our study, possibly need larger sample size These results emphasize the complex regulatory dynamics of non-coding RNAs in the pathogenesis of MS and NMO, further highlighting the intricate molecular mechanisms.

The expected inverse relationship between lncRNA Ftx and miR-382-5p was not observed in any of the studied groups, highlighting the complex and multifaceted regulatory mechanisms underlying these diseases. Similar interactions have been reported in other conditions, such as oral squamous cell carcinoma [68], gastric cancer [69], and lung carcinoma [70], where non-coding RNAs and their interactions influence various molecular processes. While some correlations in our study were statistically significant, the interpretation of these results should consider the potential for non-coding RNAs to impact each other's regulatory roles.

Recent studies have highlighted the potential of various serum biomarkers for diagnosing and prognosticating Multiple Sclerosis (MS). Notably, an autoantibody signature has been proposed as a predictive biomarker for MS, demonstrating strong potential for early detection and diagnosis [71]. Additionally, the study by Ovchinnikova et al. [72] suggested that the high heterogeneity of cross-reactive immunoglobulins in MS supports the combination of B-cell epitopes for diagnostic purposes. Further research has combined serum markers with optical coherence tomography angiography to improve diagnostic accuracy for both MS and Neuromyelitis Optica (NMO) [73]. Moreover, serum markers such as GFAP and NfL levels have been shown to differentiate between subsequent progression and disease activity in patients with progressive MS [74]. These findings, along with our current study on the inflammatory axis of LncRNA Ftx/miR-382-5p/NRG1, emphasize the growing potential of biomarkers in improving diagnostic precision and differentiating MS from other conditions, such as NMO.

This study has several limitations. Its retrospective design and challenges within the Egyptian healthcare system mean that patients may have previously received corticosteroids, potentially limiting the analysis of treatment-naïve MS patients. The inclusion of potentially misdiagnosed cases adds complexity to the findings, and the absence of treatment-naïve MS and NMO patients precluded analysis of the lncRNA Ftx/miR-382-5p/NRG1 axis across treatment groups. The focus on total NRG1 rather than isoform-specific analysis, due to logistical constraints, limited deeper insights into the distinct roles of NRG1 and NRG1β1. Additionally, the small, single-center cohort, lack of regular frequent cerebrospinal fluid samplying, and uncollected Epstein-Barr virus (EBV) status further restrict generalizability. Future studies with larger, more diverse cohorts, inclusion of treatment-naïve patients, and isoform-specific analyses are essential to validate and expand upon these findings. Furthermore, previous studies have shown that different DMTs variably impact immune cell subsets, including T cells and IL-4 expression, with some having no effect and others modulating specific immune pathways. While our study did not specifically analyze the effects of individual DMTs on biomarker expression, their potential influence should be considered when interpreting our results. Future studies with larger, more diverse cohorts, inclusion of treatment-naïve patients, and isoform-specific analyses are essential to validate and expand upon these findings. Additionally, further research should explore how DMTs influence the regulatory networks of lncRNAs, miRNAs, and cytokines in MS pathogenesis.

Our study highlights the potential of NRG1, miR-382-5p and lncRNA Ftx as transformative biomarkers in distinguishing between MS and NMO, offering new avenues for more precise diagnostics and personalized treatments. The significant dysregulation of NRG1 in MS, along with its differential levels between MS and NMO, underscores its importance in advancing neuroimmunology. While we have made substantial progress, continued research is essential to fully leverage these insights for improved therapeutic strategies and patient outcomes.