A Review on the Role of Non-Coding RNAs in the Pathogenesis of Myasthenia Gravis

Myasthenia gravis (MG) is an autoimmune condition related to autoantibodies against certain proteins in the postsynaptic membranes in the neuromuscular junction. This disorder has a multifactorial inheritance. The connection between environmental and genetic factors can be established by epigenetic factors, such as microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). XLOC_003810, SNHG16, IFNG-AS1, and MALAT-1 are among the lncRNAs with a possible role in the pathoetiology of MG. Moreover, miR-150-5p, miR-155, miR-146a-5p, miR-20b, miR-21-5p, miR-126, let-7a-5p, and let-7f-5p are among miRNAs whose roles in the pathogenesis of MG has been assessed. In the current review, we summarize the impact of miRNAs and lncRNAs in the development or progression of MG.


Introduction
Myasthenia gravis (MG) is an autoimmune condition caused by the presence of autoantibodies against certain proteins in the postsynaptic membranes in the neuromuscular junction [1]. The main detected autoantibodies are targeted against the muscle acetylcholine receptor (AChR) [2]. Other targets of autoantibodies are MuSK [3] and LRP4 [4]. The observed 35% concordance rate in MG occurrence among monozygotic twins shows the impact of environmental factors in the pathoetiology of this disorder [5]. From a clinical point of view, MG is characterized by fluctuating fatigability and weakness in a number of muscles, such as ocular, bulbar, and limb muscles. MG subtypes include ocular myasthenia and generalized myasthenia, which can have mild, moderate, or severe presentations [6]. Genetic studies have revealed the role of HLA loci as well as a number of other genes such as TNIP1 and PTPN22 in the pathogenesis of MG [5]. Moreover, pathways regulating differentiation of regulatory T cells as well as NF-κB signaling are implicated in this disorder. Significant heterogeneity has been detected in the course of MG in terms of epitopes targeted by autoantibodies, age of disease onset, and thymus histopathology [7]. While early onset MG cases are predominantly females with hyperplastic thymus histology, late onset cases are mainly males having normal or atrophic thymus [8].
Taken together, MG is a multifactorial disorder with both genetic and environmental etiologies. Meanwhile, the connection between environmental and genetic factors can be established by epigenetic factors [5], such as microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). In the current review, we summarize the impact of miRNAs and lncRNAs in the development or progression of MG.

LncRNAs and MG
LncRNAs are transcripts with sizes ranging from two hundred nucleotides to thousands of nucleotides. While they share several features with mRNAs, they lack the ability to be served as templates for proteins. Instead, they regulate expression of protein-coding genes at different levels [9]. Dysregulation of lncRNAs have been reported in MG. For instance, experiments in thymus specimens of patients with MG and MG-thymoma (MG-T) have shown upregulation of XLOC_003810 lncRNAs in these patients parallel with increased frequency of CD4+ T cells and release of proinflammatory cytokines in these patients. Functionally, XLOC_003810 upregulation has enhanced frequency of CD4+ T cells, increased synthesis of inflammatory cytokines, and reduced CD4+ PD-1+ T cells and CD14+ PD-L1+ monocytes in the mononuclear cells of thymus. XLOC_003810 silencing has resulted in opposite effects. Cumulatively, XLOC_003810 is an lncRNA that enhances T cells activation and suppresses PD-1/PD-L1 signaling in MG-T patients [10]. Another study has shown higher levels of XLOC_003810 in thymic CD4+ T cells obtained from MG-T patients compared with the control group. Additionally, MG-T thymic CD4+ T cells have exhibited a higher Th17/Treg ratio, higher frequency of Th17 cells and increased expression of Th17-associated proteins, while exhibiting lower amounts of Treg cells and downregulation of Treg-associated proteins. Notably, upregulation of XLOC_003810 has aggravated the imbalance between Th17 and Treg cells in MG-T thymic CD4+ T cells. Thus, XLOC_003810 could influence the balance between Th17 and Treg cells in the context of MG-T [11].
SNHG16 is another lncRNA that partakes in the pathogenesis of MG. This lncRNA acts as a competing endogenous RNA (ceRNA). Expression of SNHG16 has been found to be increased in peripheral blood mononuclear cells (PBMCs) of MG cases compared with controls. Mechanistically, SNHG16 increases expression of IL-10 through acting as a ceRNA for let-7c-5p. Moreover, SNHG16 could inhibit apoptotic processes in Jurkat cells and increase their proliferation through sponging this miRNA [12].
In MG patients, IFNG-AS1 is another dysregulated lncRNA whose expression is associated with the specific quantitative scoring system for MG, i.e., QMG as well as the presence of anti-AchR Ab antibody. Experiments in an animal model of MG have shown that this lncRNA affects proliferation of Th1/Treg cells and regulates expression of Th1/Treg-related transcription factors. Moreover, IFNG-AS1 has been found to decrease expressions of HLA-DRB and HLA-DOB. Moreover, it can affect the expression of CD40L and activity of CD4+ T cells through influencing HLA-DRB1 expression. Thus, IFNG-AS1 has a possible role in regulation of CD4+ T cell-associated immune response in MG [13].
Another study has reported downregulation of MALAT-1 in MG. This lncRNA has been found to compete with MSL2 for binding to miR-338-3p. Thus, MALAT-1/miR-338-3p/MSL2 axis is a new interaction network in MG [14].
A high throughput study in MG has shown upregulation of more than 1500 lncRNAs and downregulation of more than 1000 lncRNAs, parallel with dysregulation of several mRNAs in these patients. Dysregulated genes have been involved in MG-related cellular processes such as nucleic acid transcription factor activity, inflammatory responses, leukocytes activation, and lymphocytes proliferation. Moreover, they have been enriched in pathways related to immune function such as cytokine-cytokine receptor interaction, intestinal immune network for IgA synthesis, NOD-like receptor signaling, and cell adhesion, as well as NF-κB and TNF signaling pathways [15].
Another high throughput study has revealed differential expression of several lncR-NAs in MG patients versus MG-T patients-among them has been lncRNA oebiotech_11933, whose dysregulation has been confirmed by real-time PCR. Cell responses to IFN-γ, platelet degranulation, chemokine receptor binding, and cytokine interactions have been identified as processes being involved in the pathoetiology of MG. Transcription factors such as CTCF, TAF1, and MYC as trans-regulatory mechanism for modulation of lncRNAs and genes, showing an important transcription factor-lncRNA-target gene network in the context MG [16]. Table 1 summarizes the role of lncRNAs in MG. Oebiotech_11933 lncRNA was the most upregulated transcript in patients with thymoma and associated with cellular response to interferon-γ, platelet degranulation, chemokine receptor binding, and cytokine interactions terms that are important in MG pathogenesis. It also had regulatory role in TF-lncRNA-target gene network. [16]  A total of 46% of deregulated genes were associated with infectious disease and inflammatory response. miR-612, miR-3654, miR-3651, and pre-miR-3651 were upregulated in AchR-EOMG. There were no pattern differences between preand post-thymectomy patients.

miRNAs and MG
miRNAs are small regulatory molecules acting at post-transcriptional level to finely modulate the expression of genes. miRNAs can participate in the evolution of MG through different mechanisms. A system biology approach in the context of MG has led to construction of transcription factor/miRNA/gene network. Subsequent analyses have resulted in extraction of 5 genes, 3 transcription factors, and 13 miRNAs. Notably, MYC has been identified as the key transcription factor in MG. The identified genes and miRNAs were principally enriched in cancer-and infection-associated pathways. In addition, the composite feed-forward loop motif-specific subnetwork has shown the potential beneficial impact of estradiol, estramustine, raloxifene, and tamoxifen in treatment of MG [20].
An integrative bioinformatics approach and literature search has led to identification of 41 MG-associated signaling pathways and 105 medications that can influence these pathways. Notably, MG-associated miRNAs and drugs can affect key MG-associated pathways, including cytokine-cytokine receptor interaction. This approach has also shown that rituximab, adalimumab, sunitinib, and muromonab might influence MG course, potentiating these drugs as novel treatments for MG [21].
Expression profiling of samples obtained from orbicularis oculi muscle (OOM) and paralyzed extraocular muscle (EOM) of patients with ophthalmoplegic MG has shown similar expression profiles of transcripts among OOM and EOM samples. Ophthalmoplegic MG cases have exhibited downregulation of eight genes in OOM samples compared with controls. The mitochondrial transcription factor TFAM has been among these genes. Notably, numerous miRNAs known to be upregulated in EOM samples have been predicted to affect expression of a number of these genes [22].
Additionally, MG risk pathways such as T cell receptor and Toll-like receptor pathways as well as natural killer cell-mediated cytotoxicity have been shown to be targeted by several miRNAs. Notably, a number of miRSNPs "switches" have been found that affect miRNA regulation in the MG-associated pathways. These miRSNPs can affect gene expressions and pathway activities. An example of these SNPS is rs28457673 (miR-15/16/195/424/497 family), which affects IGF1R expression [23]. Another bioinformatics study has predicted the effects of 18 MG-associated miRNAs on expression of MAPK1, SMAD4, SMAD2, and BCL2 and their subsequent impact on cellular pathways associated with adherens, junctions, apoptosis, or cancer-related features. These important genes negatively regulate T cell differentiation [24].
Another miRNA profiling study has shown differential expression of 41 miRNAs among MG patients who respond to immunosuppressive treatments versus non-responders. Three miRNAs clustered on 14q32.31-namely, miR-323b-3p, miR-409-3p, and miR-485-3p have been validated to be expressed at lower levels in non-responder compared with responders, while miR-181d-5p and miR-340-3p have exhibited the opposite trend. miR-323b-3p, miR-409-3p, and miR-485-3p have been suggested as markers for the prediction of response to immunosuppressive treatment in MG. Five miRNAs have been predicted to be associated with immune functions and drug metabolism. In brief, miR-323b-3p, miR-409-3p, miR-485-3p, miR-181d-5p, and miR-340-3p expression profiles are correlated with therapeutic response in MG patients [25]. Another investigation has reported downregulation of miR-320a in MG patients compared with control subjects, parallel with upregulation of pro-inflammatory cytokines in these patients. MAPK1 has been identified as a direct target of miR-320a. miR-320a downregulation has led to upregulation of pro-inflammatory cytokines via increasing expression of COX-2. This process was regulated by ERK and NF-κB signaling pathways [26]. Table 2 shows the role of miRNAs in MG.      Their sequences were highly conserved in human genome. These miRNAs expressions also correlated with genes related to inflammatory immune response such as AchR-AB, IL-6, and FOXP3.
[ Thymic biopsies, PBMC, serum/RT-PCR Expression of miR-146a was significantly higher in corticosteroid-treated patients than corticosteroid-naïve samples, while its expression in naïve group was lower than controls.
Serum results represented downregulation of miR-146a in MG patients compared with controls, which associated with TLR activation, inflammation process, and thymic hyperplastic changes. Its expression was negatively correlated with mRNA targets (IRAK1, c-REL, and ICOS). Additionally, results led to new insight in the possible mechanism of corticosteroid function in MG. [52]

Discussion
MG is an autoimmune disorder with multifactorial inheritance. Epigenetic factors such as regulatory non-coding RNAs have been found to affect pathogenesis of this disorder, possibly linking between environmental and genetic factors. LncRNAs can affect immune cells phenotypes, modulate balance between Th17 and Treg cells, and regulate expression of proinflammatory cytokines. miRNAs have been more investigated in the context of MG compared with lncRNAs. A number of lncRNAs that have been found to contribute to MG pathogenesis act as ceRNAs for miRNAs, further highlighting the impact of miRNAs in MG. Examples of lncRNA/miRNA/mRNA axes in participating in the pathogenesis of MG are SNHG16/let-7c-5p/IL-10 and MALAT-1/miR-338-3p/MSL2. In addition, high throughput studies have established some transcription factor/lncRNA/target gene networks in the context of MG, adding a layer of complexity in regulation of function of ncRNAs in this context.
Several of the lncRNAs and miRNAs that affect pathogenesis of MG converge on NF-κB and TNF signaling pathways. Moreover, expression profiles of these transcripts have been changed in response to immunomodulatory therapies, particularly corticosteroids.
A number of miRNAs have differential expression between different classes of MG, particularly those with and without thymoma. Moreover, the expression profile of some ncRNAs such as miR-106a-5p, miR-23b, miR-27a-3p, and IFNG-AS1 has been correlated with disease severity and QMG score, suggesting their participation in the pathoetiology of MG.
NcRNAs, particularly miRNAs, have the potential to be used as markers for the predication of response of MG patients to prescribed drugs and for the stratification of patients based on this issue to design patient-specific therapeutic regimens. Moreover, expression of these transcripts might be different during distinct stages of MG, potentiating these transcripts as biomarkers for differentiation of disease status. In spite of extensive research in this field, it is not clear how the expression profile of these transcripts can affect the course of disorder or define the muscle groups that are affected during disease course.
Due to the complex nature of participation of genetic factors in the pathogenesis of MG, high throughput studies are needed to find factors acting at upstream and downstream of lncRNAs and miRNAs. This type of study will reveal novel transcription factor/lncRNA/miRNA/mRNA axes with putative roles in the pathophysiology of MG. These molecular axes represent therapeutic targets for MG. Moreover, system biology approaches have potential for the discovery of novel drugs for the treatment of MG.
The impact of miRNAs on drug response has been suggested through both system biology and experimental studies, potentiating these transcripts as targets for therapeutic interventions. Taken together, miRNAs and lncRNAs partake in the pathoetiology of MG, disease course, and response to immunosuppressive treatments. Thus, these transcripts can be used as markers for prediction of these aspects.

Conclusions and Future Perspectives
As an autoimmune disorder, MG is associated with dysregulation of miRNAs and lncRNAs. However, the therapeutic implications of these transcripts in MG are not clear. The biomarker role of non-coding RNAs in MG has been investigated. Nonetheless, there is no clear evidence of whether the expression profile of these transcripts can determine disease course or involvement of certain groups of muscles. Future high throughput sequencing experiments should unravel differential expression of ncRNAs during different stages of MG.