Abstract
Seronegative myasthenia gravis (SNMG) is a clinically heterogeneous subtype of myasthenia gravis (MG) that lacks the classical pathogenic autoantibodies used in routine diagnosis, leaving the underlying mechanisms poorly defined. Integrative genetic and single-cell transcriptomic analyses identified VARS2 as an MG-associated candidate gene with particular relevance to SNMG. VARS2 dysregulation was predominantly observed in CD4+ T cells, especially in the SNMG subgroup. To evaluate its in vivo relevance to SNMG, we established a zebrafish vars2 knockdown model. vars2 deficiency induced a myasthenia-like phenotype characterized by impaired locomotor activity, abnormal skeletal muscle development, defective motor neuron formation, and reduced expression of neuromuscular junction-related genes. Furthermore, the loss of vars2 disrupted mitochondrial homeostasis, evidenced by mitochondrial structural damage, membrane potential loss, increased reactive oxygen species accumulation, and suppression of metabolic pathways. These changes were accompanied by immune dysregulation, marked by CD4+ T cell/Th17-associated inflammatory activation, reduced expression of Treg-related genes, and increased neutrophil and macrophage recruitment. Crucially, pharmacological interventions partially reversed these defects through mitochondrial support (CoQ10), acetylcholinesterase inhibition (Neostigmine), and targeted immune suppression (ML604440 for Th17 differentiation, GSK805 for Th17 function, and UNBS5162 for CXCL-related signaling). Together, these findings suggest that VARS2 links mitochondrial dysfunction to immune imbalance and neuromuscular injury in SNMG. Our study highlights mitochondrial immunometabolism as a candidate mechanistic axis and identifies VARS2 as a potential biomarker for SNMG-related immune remodeling.
1 Introduction
Myasthenia gravis (MG) is an acquired autoimmune disorder characterized by impairment of neuromuscular junction (NMJ) transmission (, ). Based on the spectrum of pathogenic autoantibodies, MG comprises immunologically distinct subtypes (–). While most patients harbor autoantibodies against the acetylcholine receptor (AChR) or muscle-specific kinase (MuSK), a subset lacks these classical serological markers and is classified as having seronegative myasthenia gravis (SNMG) (, ). Notably, up to 15% of patients with generalized MG and 50% of those with ocular MG are negative for AChR and MuSK antibodies on routine testing (). A critical clinical challenge in managing SNMG is the absence of definitive serological evidence, which is compounded by profound clinical heterogeneity. This diagnostic ambiguity is particularly evident in early-stage disease or ocular involvement, where SNMG is frequently misdiagnosed as mitochondrial myopathy or other neuromuscular disorders (, ). Consequently, there is an urgent need to elucidate the molecular basis of SNMG and to identify mechanism-driven biomarkers for precise stratification and targeted intervention.
Beyond humoral autoimmunity, CD4+ T cell-mediated cellular immune dysregulation is a pivotal driver of MG immunopathology (). CD4+ T cells not only participate in B-cell activation and antibody production, but also influence disease progression by secreting inflammatory cytokines and regulating the immune microenvironment (). Studies have shown that MG patients typically exhibit enhanced Th17 cell-associated inflammatory responses, together with reduced numbers or impaired function of regulatory T cells (Treg cells). The imbalance between Th17 and Treg cells is considered an important mechanism for immune tolerance breakdown and chronic inflammation maintenance (–). In some patients with SNMG, enhanced CD4+ T cell activation and abnormal Th17/Tfh-related cytokine profiles have also been observed, suggesting that despite the absence of classical autoantibodies, SNMG may still be characterized by a distinct basis of cellular immune imbalance (). Abnormal immunometabolism may further contribute to immune dysregulation in autoimmune diseases, with mitochondria serving as central regulators of energy metabolism, redox homeostasis, cellular stress, and immune responses (, ). In CD4+ T cells, mitochondrial dysfunction may promote pro-inflammatory immune responses and weaken immunosuppressive function, thereby driving the initiation and persistence of autoimmune responses (–). Concurrently, the neuromuscular system, as a highly energy-demanding tissue, is particularly vulnerable to mitochondrial defects, and impaired mitochondrial translation or oxidative phosphorylation (OXPHOS) can directly lead to neuromuscular injury (–). Therefore, mitochondrial dysfunction may represent an upstream mechanism linking immune imbalance and neuromuscular damage in SNMG, although the precise molecular drivers remain unclear.
The mitochondrial valyl-tRNA synthetase, encoded by the VARS2 gene, is an essential component of the mitochondrial translation machinery. It catalyzes the attachment of valine to mitochondrial tRNAVal, thereby maintaining the synthesis of mitochondrially encoded proteins and the function of respiratory chain complexes (, ). Mutations or dysregulation of VARS2 have been linked to mitochondrial translation defects, OXPHOS deficiency, disrupted fatty acid oxidation, and severe multisystem mitochondrial disorders (–). Furthermore, members of the aminoacyl-tRNA synthetase family possess noncanonical functions extending beyond protein synthesis, including the regulation of inflammatory and immune responses (, ). Given these dual roles, we hypothesized that VARS2 may act as a critical molecule linking mitochondrial dysfunction, immunometabolic imbalance, and downstream tissue injury (, ). However, the spatial expression pattern of VARS2 in SNMG and its specific contribution to CD4+ T cell dysfunction and subsequent NMJ impairment remain unexplored.
In this study, by integrating genome-wide association study (GWAS) data, summary-data-based Mendelian randomization (SMR), and colocalization analyses, we identified VARS2 as a genetically supported candidate driver of MG. Single-cell transcriptomic profiling further revealed that VARS2 dysregulation was primarily localized to CD4+ T cells, particularly within the SNMG patient subgroup. Utilizing an in vivo zebrafish model, we systematically demonstrated that vars2 deficiency precipitates an SNMG-like phenotype, disrupting mitochondrial homeostasis, driving immune-related abnormalities, and impairing neuromuscular development. Our findings unveil the pathogenic contribution of VARS2 to SNMG and provide novel mechanistic insights into the interplay between immunometabolism and NMJ integrity.
2 Results
2.1 Integrative Mendelian randomization, colocalization and single-cell transcriptomic analyses identify VARS2 as a MG-associated candidate gene
To identify candidate causal genes for MG, we performed a SMR analysis using whole-blood cis-expression quantitative trait locus (cis-eQTL) data from the eQTLGen. Information on the included GWAS, eQTL, and single-cell datasets is summarized in Supplementary Table 1. Four genes, VARS2, HLA-DRB5, PGAP3 and CTSH, were significantly associated with MG (Figure 1A; Supplementary Table 2), among which VARS2 showed the strongest association with MG. Forest plot analysis showed that higher expression of HLA-DRB5 and CTSH was associated with increased MG risk, whereas lower expression of VARS2 and PGAP3 was associated with increased MG risk (Figure 1B). Further colocalization analysis showed that only VARS2 demonstrated strong colocalization evidence with MG risk in whole blood (PPH4 = 0.996, P < 0.001, Figure 1C).
Figure 1
Given the genetic implication of VARS2 in whole blood data, we next profiled its cellular origin and expression dynamics in the MG immune microenvironment by single-cell transcriptomics of peripheral immune cells from healthy controls (HC), SNMG, and AChR-positive MG patients. Unsupervised clustering and cell-type annotation identified diverse immune populations across cohorts (Figure 1D). Uniform Manifold Approximation and Projection (UMAP) showed heterogeneous VARS2 expression across cell types (Figure 1E), with the highest average level in naïve CD4+ T cells (Figure 1F).
To delineate the disease- and subtype-specific expression pattern of VARS2, we compared its transcript abundance across the HC, SNMG, and AChR-positive MG groups. VARS2 dysregulation was most evident in naïve CD4+ T cells, where expression was significantly higher in SNMG than in HC or AChR-positive MG (Figures 1G, H; Supplementary Table 3). These results identify VARS2 as a high-confidence MG-associated risk gene and localize its most prominent transcriptional alteration to naïve CD4+ T cells in SNMG.
2.2 vars2 depletion compromises skeletal muscle myogenesis and spinal motor neuron development in zebrafish
Having identified VARS2 as a genetically prioritized candidate gene, we next examined its function in vivo using a zebrafish model. VARS2 is highly conserved across vertebrates and exhibits a similar predicted overall structure in humans, mice, and zebrafish (Supplementary Figures 1A, B), supporting the use of zebrafish for functional analysis. We therefore established a zebrafish vars2 knockdown model utilizing a vars2 -targeted morpholino (MO). RT-qPCR confirmed that vars2 expression was reduced by approximately 85% in vars2 morphants (Supplementary Figure 1C).
Morphological assessment revealed concurrent developmental defects in skeletal muscle and spinal motor neurons following vars2 depletion. Polarized light microscopy showed that while the standard control MO (STD MO) group displayed regularly arranged muscle fibers with distinct somite borders, whereas vars2 morphants exhibited disorganized muscle fibers, a reduced somite count, and indistinct segment boundaries (Figure 2A). Analysis in Tg(mylpfa:RFP) transgenic zebrafish showed that the regular alignment and continuous fluorescent signals of fast muscle fibers observed in the STD MO group were disrupted in vars2 morphants, which displayed reduced fluorescence intensity and disorganized muscle fibers (Figure 2B). In parallel, spinal motor neuron development was also impaired. Whole-mount in situ hybridization for isl1 showed well-formed spinal motor neurons with identifiable ganglia in the STD MO group at 72 hours post fertilization (hpf), whereas vars2 morphants presented with disorganized motor neuron arrangements and fewer distinct ganglia (Figure 2C). Tg(mnx1:EGFP) transgenic larvae showed that the regular motor neuron branching present in the STD MO group was replaced by disorganized branching patterns and restricted axonal extension in the morphants (Figure 2D).
Figure 2
RT-qPCR analysis further showed broad downregulation of myogenic markers (myod1, myog, and acta1a) and NMJ-related genes (musk, chrne, and chrnb1) in vars2 morphants (Figure 2E). Concurrently, essential neural and neurogenesis markers, including pax2a, isl1, ache, neurog1, and neurod1, were also significantly downregulated in the vars2 morphants (Figure 2F). Together, these findings indicate that vars2 is required for normal skeletal muscle and spinal motor neuron development in zebrafish, suggesting that its loss may impair neuromuscular function.
2.3 vars2 deficiency induces an SNMG-like motor phenotype in zebrafish
Given the neuromuscular defects observed in vars2 morphants, we next examined whether vars2 deficiency also impaired motor function. At 5 day post fertilization (dpf), vars2 morphants exhibited locomotor deficits, characterized by shortened swimming trajectories under basal conditions compared with the STD MO group (Figures 3A–C). Quantitative analysis showed that cumulative active movement time was reduced to 36% of control levels, whereas average swimming velocity decreased to 17% of control values (approximately 0.1 mm/s). In contrast, resting time was increased nearly four-fold in vars2 morphants (Figures 3D–F).
Figure 3
To determine whether this motor impairment was related to myasthenic dysfunction, we treated larvae with neostigmine, a canonical acetylcholinesterase inhibitor used in clinical MG management. Neostigmine significantly improved the locomotor abnormalities of vars2 morphants, increasing cumulative movement time, swimming speed, and total travel distance, while reducing resting time (Figures 3B−F). These results indicate that vars2 depletion induces an MG-like motor phenotype in zebrafish that is responsive to acetylcholinesterase inhibition.
2.4 Loss of vars2 induces mitochondrial structural damage and oxidative stress in zebrafish embryos
To elucidate the cellular mechanisms underlying the observed motor deficits, and given that VARS2 encodes a mitochondrial valyl-tRNA synthetase, we next examined whether the motor abnormalities in vars2 morphants were accompanied by mitochondrial defects. MitoTracker staining showed an approximately 27% reduction in mitochondrial abundance in vars2 morphants compared with the STD MO group (Figures 4A, B). Consistent with this, genes involved in mitochondrial fusion/fission dynamics (mfn2, fis1, opa1) and membrane transport (timm23a, tomm20) were significantly downregulated (Figure 4C).
Figure 4
To resolve these defects at the subcellular level, we performed transmission electron microscopy (TEM) on skeletal muscle at 5 dpf. While the STD MO group muscle exhibited intact mitochondria with well-organized cristae, vars2 morphants displayed pronounced ultrastructural abnormalities, including cristae disruption, matrix swelling, and irregular organelle morphology (Figure 4D). In parallel, genes involved in mitochondrial calcium handling (vdac1, mcu, tspo) and ion homeostasis (letm1, bcl2l13) were markedly dysregulated in vars2 morphants (Figure 4E), further supporting impaired mitochondrial integrity.
Beyond structural abnormalities, vars2 deficiency impaired mitochondrial functional and metabolic capacity. JC-1 staining showed a loss of mitochondrial membrane potential (Δψm) in vars2 morphants (Figure 4F), and this depolarization was confirmed by the quantification of the red/green fluorescence ratio (Figure 4G). vars2 depletion suppressed energy-generating pathways, indicated by the downregulation of genes involved in glycolysis (hk2, pfkpa, pkma), the tricarboxylic acid (TCA) cycle (cs, idh3a, mdh2) and OXPHOS (ndufs1, sdhb, cox4i1) (Figure 4H). Consistent with mitochondrial dysfunction, reactive oxygen species (ROS) accumulation was significantly increased in vars2 morphants (Figure 4I), with quantification confirming this elevation (Figure 4J). Moreover, mitochondrial proteostasis genes (clpp, clpxa) were downregulated, whereas ROS-response genes (sod2, prdx3, txn2, gpx4a) were upregulated (Figure 4K). Together, these findings indicate that vars2 is required to maintain mitochondrial structural and metabolic homeostasis in zebrafish.
2.5 vars2 depletion induces global transcriptomic remodeling and immune dysregulation
To delineate the systemic molecular alterations driven by vars2 deficiency, we performed RNA-seq on zebrafish larvae. GO and Reactome pathway analyses of the upregulated genes revealed significant enrichment in immune- and stress-related processes, including inflammatory activation, extracellular matrix remodeling, neutrophil degranulation, antigen presentation, and JAK-STAT signaling (Figures 5A, A’; Supplementary Tables 11, 12). In contrast, downregulated transcripts were mainly enriched in pathways associated with synaptic and metabolic functions. Functional annotation of these repressed genes highlighted chemical synaptic transmission, synapse organization, monoatomic ion transport and channel activity, neuromuscular junction- and postsynaptic membrane-related components, as well as fatty acid metabolic and peroxisomal processes (Figures 5B, B’; Supplementary Tables 11, 12). Together, these transcriptomic changes indicate broad disruption of neuro-immune and metabolic homeostasis in vars2 morphants.
Figure 5
Targeted analysis of immune-related differentially expressed genes further refined the nature of this inflammatory shift. Immune-associated transcripts displayed subtype-related alterations, with Th17-associated genes including il1b, il21, and il17f markedly upregulated in vars2 morphants, whereas several transcripts related to CD8+ T cell function were downregulated (Figure 5C). This immune expression pattern supports a vars2 deficiency-induced immunometabolic shift characterized by enhanced pro-inflammatory signaling and Th17-skewed immune activation, providing a molecular basis for the immune abnormalities observed in the SNMG-like model.
2.6 vars2 deficiency drives immune imbalance characterized by T cell subset dysregulation and myeloid cell recruitment in zebrafish
To investigate the transcriptomic shift towards immune activation in vivo, we evaluated the immunological state following vars2 depletion. Analysis of Tg(rag2:mCherry) larvae revealed thymic expansion and redistribution of rag2-positive T cell precursors in vars2 morphants compared with the STD MO group (Figure 6A), suggesting altered T cell homeostasis. RT-qPCR analysis showed that CD4+ T cell-related genes (cd4-1, lck, trac) were significantly upregulated in vars2 morphants (Figure 6B), whereas CD8+ T cell-related markers (cd8a, ifng1, gzm3) remained unaltered between groups (Figure 6C).
Figure 6
Subsequent evaluation of CD4+ T cell subset polarization revealed that the Th17-associated cytokines (il17a/f1, il17a/f2, il17a/f3, il21, and il22) were comprehensively upregulated (Figure 6D), whereas the Treg marker foxp3 exhibited no significant change, and the immunosuppressive cytokine il10 was significantly downregulated (Figure 6E).
We further examined the dynamic changes of neutrophils and macrophages, given that myeloid immune cells are important effector cells in local inflammatory responses and tissue damage. In vivo tracking utilizing Tg(mpx1:GFP) transgenic larvae demonstrated enhanced recruitment and the spatial distribution of neutrophils in vars2 morphants relative to the STD MO group (Figure 6F). Similarly, Tg(mpeg:mCherry) transgenic larvae analysis showed macrophages evenly distributed in the STD MO group skeletal muscle, but markedly accumulated in structurally disorganized tissue regions in vars2 morphants (Figure 6G). These morphological observations were further supported by RT-qPCR analysis showing significant upregulation of neutrophil markers (mpx, lyz, and nspn) (Figure 6H) and macrophage markers (mpeg1, macro, and csf1ra) (Figure 6I) in the vars2 morphants. Together, these results indicate that vars2 deficiency establishes a highly inflammatory in vivo microenvironment, linking T cell imbalance and local accumulation of activated myeloid cells.
2.7 Pharmacological interventions targeting mitochondrial, neuromuscular, and immune axes ameliorate vars2 depletion-induced deficits
To further investigate the functional relationships among mitochondrial dysfunction, immune dysregulation, and neuromuscular impairment, we evaluated mechanism-specific pharmacological interventions in vars2 morphants. For upstream targets, we administered CoQ10 (to restore mitochondrial function) and neostigmine (to enhance neuromuscular transmission). Locomotor trajectory analysis and kinematic heatmaps showed that the severely reduced swimming activity in vars2 morphants was partially restored by individual treatment with either CoQ10 or neostigmine (Figures 7A, B). Notably, co-administration of both agents yielded an enhanced therapeutic efficacy. RT-qPCR analysis showed that the elevated expression of CD4+/Th17-associated transcripts (cd4-1, lck, trac, il17a/f1, il17a/f2, il17a/f3, il21, and il22) and myeloid-associated markers (mpx, lyz, nspn, mpeg1, macro, and csf1ra), alongside the suppressed Treg marker il10, were partially reversed by treatment with CoQ10 or neostigmine alone, and the recovery was more marked under the combined treatment, while CD8+ T cell markers (cd8a, ifng1, and gzm3) remained relatively stable across conditions (Figures 7C−H).
Figure 7
These experiments were designed to assess whether interventions targeting upstream mitochondrial or neuromuscular abnormalities could also alleviate the associated immune dysregulation. Conversely, we next asked whether directly suppressing the downstream immune abnormalities could improve the neuromuscular phenotype. We subsequently investigated whether directly targeting the downstream immune dysregulation could mitigate the disease phenotype utilizing specific immunomodulatory agents ML604440 (proteasome β1i/LMP2 inhibitor, blocking naïve T cell differentiation into Th17 cells), GSK805 (RORγt inhibitor, suppressing Th17 cell function), and UNBS5162 (broad-spectrum CXCL antagonist, inhibiting excessive myeloid cell attack). All three immune-targeted treatments effectively ameliorated the locomotor deficits in vars2 morphants (Figures 7I, J). Consistent with these behavioral improvements, the abnormal expression of critical myogenic (myod1, mylpfa, acta1a, chrne, and lpin1) and neurogenic (ache, isl1, neurod1, neurog1, and pax2a) markers was substantially restored across the treatment groups (Figures 7K, L). Collectively, these pharmacological rescue experiments demonstrate that intervening at the level of mitochondrial function, neuromuscular transmission, or downstream immune signaling can effectively alleviate the vars2 deficiency-induced myasthenic pathology, thereby functionally validating the integrated neuro-immune-metabolic mechanism underlying SNMG.
3 Discussion
By integrating genetic association analysis, single-cell transcriptomics, and in vivo functional evaluation, this study identifies VARS2 as a candidate contributor to SNMG and suggests that its pathogenic relevance lies in linking mitochondrial dysfunction to immune and neuromuscular abnormalities. Beyond its established role in mitochondrial translation, VARS2 may participate in an SNMG-related immunometabolic axis centered on CD4+ T cell dysregulation. In zebrafish, vars2 deficiency caused coordinated defects in mitochondrial integrity, immune homeostasis, and neuromuscular development, resulting in a myasthenia-like motor phenotype that was partially rescued by mitochondrial support, acetylcholinesterase inhibition, and immune modulation. Across the RT-qPCR datasets, 2−ΔCt-based analyses further contextualized the observed fold changes by revealing substantial differences in basal expression abundance among the examined genes (Supplementary Figures 1–S5). The comparable β-actin Ct values between STD MO and vars2 MO groups across experiments further supported its stability as the reference gene (Supplementary Figure 6). Based on these findings, we present a mechanistic framework in which VARS2 insufficiency promotes mitochondrial dysfunction, CD4+ T cell-associated immune imbalance, and downstream neuromuscular injury, thereby contributing to SNMG pathogenesis (Figure 8).
Figure 8
One important issue raised by our data is the apparent discrepancy between the genetic association analysis and the single-cell expression results. The SMR analysis captures the relationship between genetically regulated VARS2 expression and overall MG susceptibility, whereas the single-cell data reflect disease-state transcriptional changes within specific immune-cell subsets. These readouts are therefore not directly equivalent. Moreover, the single-cell findings should be interpreted in the context of the source cohort: the SNMG patients had refractory early-onset generalized MG, received long-term immunosuppressive treatment, and were sampled before IVIg rather than before any immunotherapy (). Seronegative status was also defined without live cell-based assays, so low-titer or low-affinity autoantibodies cannot be excluded (). These factors limit subtype-specific and causal interpretation of the transcriptomic differences. Nevertheless, the increased VARS2 expression in naïve CD4+ T cells may reflect disease-state or compensatory transcriptional remodeling rather than the primary direction of pathogenic effect (, ). Consistent with this interpretation, in vivo vars2 knockdown induced locomotor and neuromuscular abnormalities that were partially rescued by neostigmine. Together with the genetic association results, these functional data support a protective role for VARS2 in maintaining neuromuscular integrity, while its specific relevance to rigorously defined SNMG warrants further validation in independent, treatment-matched, and appropriately serotyped cohorts.
Our findings also extend current views of MG immunopathology. Previous studies have identified broad immune dysregulation in MG beyond autoantibody production, including altered CD4+ T cell activation, Th17/Treg imbalance, inflammatory signaling abnormalities, and disrupted Tfh/Tfr-associated immune organization (–). However, these alterations have mainly been described at the level of immune phenotypes and cytokine imbalance, with limited evidence linking them to defined mitochondrial mechanisms. Our data suggest that, in SNMG, immune dysregulation may be mechanistically coupled to mitochondrial dysfunction. VARS2 dysregulation was most evident in CD4+ T cells from SNMG patients, and vars2 deficiency in vivo was associated with mitochondrial dysfunction, ROS accumulation, metabolic suppression, and a Th17-dominant inflammatory profile with reduced il10 expression. Together, these findings support mitochondrial dysfunction as a potential upstream driver of immune imbalance in SNMG.
This interpretation is biologically plausible in light of the known role of VARS2 and the metabolic sensitivity of T cell fate decisions. VARS2 encodes the mitochondrial valyl-tRNA synthetase and is required for mitochondrial translation, respiratory-chain assembly, and OXPHOS homeostasis (, ). Previous studies of VARS2 have mainly focused on primary mitochondrial disorders and multisystem energetic failure (, , ), whereas its role in immune regulation has remained largely unexplored. In parallel, CD4+ T cell activation and differentiation are highly dependent on mitochondrial bioenergetics, redox control, and metabolic reprogramming, with the Th17/Treg axis being especially sensitive to such perturbations (, –). Our data connect these previously separate lines of evidence by suggesting that VARS2 dysfunction may promote SNMG-related immune remodeling through disruption of mitochondrial homeostasis.
Beyond adaptive immunity, our results also implicate innate immune recruitment in the pathological consequences of vars2 deficiency. The increase in neutrophils and macrophages in the zebrafish model is consistent with growing evidence that innate immune pathways contribute to chronic inflammatory remodeling, thymic microenvironmental alteration, and tissue injury in MG (, , ). These findings indicate that VARS2 dysfunction not only perturbs adaptive immune regulation but also promotes innate immune recruitment, thereby reinforcing local inflammatory injury at the tissue level. Our data therefore support an integrated pathogenic framework in which mitochondrial stress, adaptive immune deviation, and innate immune activation are mechanistically linked rather than operating as isolated processes.
The intervention experiments further corroborate this framework. In vivo modulation of mitochondrial dysfunction, neuromuscular transmission, and downstream inflammatory pathways each partially alleviated the vars2-deficient phenotype. In particular, immune-targeted interventions directed at naïve T cell differentiation into Th17 cells, Th17-associated effector responses, and chemokine-related signaling improved locomotor phenotypes and corrected neural and muscular molecular abnormalities, consistent with prior evidence implicating inflammatory and CXCL-associated pathways in MG immune remodeling (–, , ). In parallel, upstream intervention with CoQ10 and neostigmine also ameliorated the defects induced by vars2 deficiency, suggesting that mitochondrial support and restoration of cholinergic neuromuscular transmission can independently modify the disease phenotype. However, because the pharmacological experiments were designed to interrogate different levels of the proposed pathogenic pathway, reciprocal assessment of both immune and neuromuscular markers was not performed for every intervention. Future studies incorporating such reciprocal analyses will be important to more comprehensively define the relationships among these pathological processes. Taken together, these findings support a multilayer pathogenic model in which mitochondrial impairment, immune dysregulation, and neuromuscular dysfunction act in concert to drive an SNMG-relevant myasthenic phenotype.
In summary, our study identifies a potential link between VARS2 deficiency, mitochondrial dysfunction, and CD4+ T cell-related immune disturbance in SNMG, offering a new perspective on the interplay between immune imbalance and neuromuscular injury. Although the present findings support a role for VARS2 in SNMG, several important questions remain to be resolved. Further studies in independent SNMG cohorts and complementary experimental models will be needed to validate the disease relevance of VARS2 and to define its role in long-term pathological progression. In particular, it remains unclear whether VARS2 directly regulates metabolic programming and lineage differentiation within CD4+ T cells, or whether its effects occur indirectly through broader mitochondrial stress and changes in the immune microenvironment. Future studies combining purified immune cell subsets, cell type-specific models, and patient-derived samples will be needed to address this question. Such work may clarify how mitochondrial dysfunction contributes to immune imbalance and neuromuscular injury in SNMG, and may also inform the development of mechanism-based therapies.
4 Materials and methods
4.1 Data source
Publicly available datasets were used in this study. Whole-blood cis-eQTL data were obtained from the eQTLGen Consortium, which included 31,684 blood samples (). GTEx v8 data were used as an additional eQTL resource where needed (). GWAS summary statistics for MG were obtained from two public datasets: the study by Chia R et al. (1,873 cases and 36,370 controls) and the FinnGen R10 release (461 cases and 408,430 controls) (, ).
Single-cell RNA-sequencing (scRNA-seq) data were obtained from the public dataset GSE227835, generated on the DNBSEQ-G400 platform (GPL28038) (). This dataset included healthy controls, AChR-positive MG patients, and SNMG patients. In the present study, only healthy controls, AChR-positive MG patients, and SNMG patients were included. Detailed information on the included studies and consortia is provided in Supplementary Table 1.
4.2 Genetic and single-cell analyses
MG GWAS summary statistics were quality controlled by removing variants with missing values, invalid statistics, duplicated SNP identifiers, or minor allele frequency < 0.5%. Meta-analysis was performed using METAL based on the standard error scheme (). SMR analysis was conducted using SMR software (version 1.3.1) with a cis-window of 1 Mb. Genome-wide significant cis-eQTLs (P < 5 × 10-8) were retained as instrumental variants, and linkage disequilibrium clumping was performed using the European reference panel from the 1000 Genomes Project (). Multiple testing correction was carried out using the p.adjust function in R, and genes with FDR-adjusted P < 0.05 were considered significant (). The full SMR results are provided in Supplementary Table 2. HEIDI tests were used to assess heterogeneity, and colocalization analysis was performed using the coloc R package (version 5.2.3), with PPH4 ≥ 0.8 considered strong evidence of colocalization (, ). Candidate genes shared by genetic and single-cell transcriptomic analyses are listed in Supplementary Table 8.
For single-cell analysis, genes detected in fewer than three cells were excluded. Cells with fewer than 200 or more than 4,000 detected genes, or with mitochondrial gene content >20%, were excluded from downstream analysis. After filtering, 94,920 cells and 73,917 genes were retained (). Data were analyzed using Seurat (). Expression data were normalized and scaled, followed by principal component analysis and Harmony-based batch correction. Cell clustering and visualization were performed using standard Seurat workflows (), and cell types were annotated using the ScType database together with canonical immune markers (). Marker genes distinguishing naïve CD4+ T cells from other immune cell populations are provided in Supplementary Table 3. Expression of target genes was compared among healthy controls, AChR-positive MG, and SNMG samples. Differential expression results for immune cells and naïve CD4+ T cells in the pairwise comparisons of SNMG versus healthy controls and SNMG versus AChR-positive MG are provided in Supplementary Tables 4–7.
4.3 Cross-species conservation analysis of VARS2
The full-length amino acid sequences of the VARS2 protein from seven vertebrate species were retrieved from the NCBI GenBank database. Multiple sequence alignment was performed using MultAlin. Structural models of VARS2 in human, mouse, and zebrafish were generated using AlphaFold3 to assess sequence and structural conservation.
4.4 Zebrafish maintenance and vars2 knockdown
Adult zebrafish (AB strain) and the transgenic lines Tg(mylpfa:RFP), Tg(mnx1:EGFP), Tg(rag2:mCherry), Tg(mpx1:GFP), and Tg(mpeg:mCherry) were maintained at 28.5 °C under a 14 h light/10 h dark cycle. Embryos were obtained by natural mating and staged according to hours post fertilization or days post fertilization (). All animal procedures were approved by the Experimental Animal Management and Ethics Committee of Northwest University (NWU-AWC-20241107Z).
Antisense morpholinos targeting zebrafish vars2 (splice-donor morpholino, vars2 MO) and a standard control morpholino (STD MO) were purchased from Gene Tools, LLC (USA) (Supplementary Table 9). Morpholinos were injected at the 1-cell stage to generate transient vars2 knockdown embryos. Knockdown efficiency was assessed by RT-qPCR at 24 hpf.
4.5 Behavioral and neuromuscular analyses
Locomotor behavior was assessed using the ZebraMonitor 2.0 system and ZebraEthogram 2.0 software (Huaxin Microfish, China). Larvae were acclimated at 28.5 °C for 30 min before recording. Locomotor activity was monitored for 30 min with a sampling interval of 10 s. Parameters including movement trajectory, total movement duration, resting time, swimming speed, and total swimming distance were analyzed. Heatmaps were generated to visualize activity intensity ().
For morphological evaluation of the neuromuscular system, embryos at 72 hpf were examined by birefringence using an SMZ25 stereomicroscope with a DS-Ri2 digital camera (Nikon, Japan) to assess muscle integrity, including somite number and muscle fiber organization. Tg(mylpfa:RFP) and Tg(mnx1:EGFP) larvae were imaged using the same system to evaluate fast muscle fibers and motor neuron morphology. Whole-mount in situ hybridization was performed using a digoxigenin-labeled isl1 probe (Roche Diagnostics, USA) to detect spinal motor neurons, with image acquisition performed as described above. The expression of key myogenic and neurogenic marker genes was quantified via RT-qPCR.
4.6 Assessment of mitochondrial structure and function
For ultrastructural examination of mitochondria, embryos at 48 hpf were fixed overnight at 4 °C in 2.5% glutaraldehyde prepared in 0.1 M phosphate buffer (pH 7.4), followed by post-fixation in 1% osmium tetroxide. Samples were dehydrated, embedded in Epon 812 resin, sectioned at 70 nm, and examined by TEM (Hitachi, Tokyo, Japan). The integrity of mitochondrial cristae and structural morphology were evaluated in randomly selected microscopic fields.
Mitochondrial staining intensity, ROS, and mitochondrial membrane potential were assessed by fluorescent staining. Embryos were stained with 1 μM MitoTracker™ Red CMXRos (Invitrogen, USA) for mitochondrial staining intensity, 10 μM DCFH-DA (Beyotime, China) for ROS detection, and 5 μM JC-1 (Sigma-Aldrich, USA) for mitochondrial membrane potential analysis. Fluorescence images were acquired using an SMZ25 stereomicroscope with a DS-Ri2 digital camera (Nikon, Japan) and quantified with ImageJ (version 1.8.0) (). The transcript levels of genes regulating mitochondrial dynamics and oxidative stress responses were also measured.
4.7 RT-qPCR analysis
Total RNA was isolated from zebrafish embryos in each group using TRIzol reagent (Ambion, USA). cDNA was synthesized using SuperScript III (Invitrogen, USA) according to the manufacturer’s instructions. RT-qPCR was performed using the ChamQ Blue Universal SYBR RT-qPCR Master Mix (Vazyme, China) on a QuantGene 9600 Real-Time PCR System (Bioer Technology, China). Primer sequences are listed in Supplementary Table 10. β-actin was used as the internal reference, and relative expression levels were calculated using the 2-ΔΔCt method ().
4.8 RNA library preparation, sequencing and transcriptomic analysis in zebrafish
Total RNA was extracted from 5 dpf larvae in the STD MO and vars2 MO groups using TRIzol reagent (Ambion, USA). RNA integrity was assessed using an Agilent 2100 Bioanalyzer, and samples with an RNA integrity number > 8.0 were used for library construction. Libraries were prepared using the TruSeq Stranded mRNA Kit (Illumina) and sequenced on the NovaSeq 6000 platform.
Raw reads were filtered using Trimmomatic (version 0.39) and aligned to the danRer11 reference genome using HISAT2 (version 2.2.1). Gene expression levels were quantified using featureCounts (version 2.0.1). Differentially expressed genes were identified using DESeq2 with |log2 fold change| > 1 and adjusted P < 0.05 as the cutoff values. Gene Ontology and Reactome enrichment analyses were performed using clusterProfiler (version 4.2.2).
4.9 Immune analyses in zebrafish
To assess immune alterations in vars2 deficient zebrafish, Tg(rag2:mCherry), Tg(mpx1:GFP), and Tg(mpeg:mCherry) larvae were imaged using an SMZ25 stereomicroscope with a DS-Ri2 digital camera (Nikon, Japan) to evaluate T cell precursors, neutrophils, and macrophages, respectively.
4.10 Pharmacological intervention experiments
To assess functional rescue, vars2 MO larvae were treated with selected pharmacological agents targeting specific mechanistic axes. To evaluate mitochondrial and neuromuscular pathways, larvae were exposed to 50 μM coenzyme Q10 and 10 μM neostigmine, administered either individually or in combination. For immune intervention, larvae were treated with ML604440, GSK805, or UNBS5162 (TargetMol, China). Untreated STD MO larvae and vehicle-treated vars2 MO larvae served as appropriate controls. Pharmacological exposure was initiated at 3 dpf, and samples were collected or analyzed at 5 dpf.
4.11 Statistical analysis
Sample sizes and biological replicates are specified in the corresponding figure legends. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per replicate; for behavioral analysis, n = 6 larvae per group. Data are presented as the mean ± SD. For RT-qPCR data, a formal normality test was not performed because each group contained only three biological replicates, for which normality testing has very limited power. Statistical comparisons of RT-qPCR data were performed on ΔCt values using Welch’s two-tailed unpaired t-test, whereas relative expression was calculated using the 2^−ΔΔCt method for visualization. For other analyses with an adequate sample size, data distribution was assessed using the Shapiro–Wilk test. Multiple-group comparisons were performed using one-way ANOVA followed by Duncan’s multiple range test, or two-way ANOVA where appropriate. Results based on n = 3 biological replicates should be interpreted with appropriate caution. P < 0.05 was considered statistically significant.
Statements
Data availability statement
The raw sequence data reported in this paper have been deposited in the Genome Sequence Archive in National Genomics Data Center, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA041728) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa/browse/CRA041728.
Ethics statement
Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements. The animal study was approved by the Experimental Animal Management and Ethics Committee of Northwest University (NWU-AWC-20241107Z). The study was conducted in accordance with the local legislation and institutional requirements.
Author contributions
ZL: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. BY: Conceptualization, Data curation, Software, Writing – original draft. YZ: Formal analysis, Validation, Writing – original draft. MS: Validation, Writing – original draft. ZW: Conceptualization, Supervision, Writing – original draft. SS: Validation, Writing – original draft. XZ: Funding acquisition, Writing – review & editing. JT: Conceptualization, Data curation, Funding acquisition, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Natural Science Foundation of China (32170618), the Fundamental Research Funds for the Central Universities (KLSB2023KF-02), and the Natural Science Basic Research Program of Shaanxi Province (2026JC-YBQN-0998). The funders had no role in the study design, data collection and analysis, or manuscript preparation.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. ChatGPT and Doubao AI for language editing, wording refinement, and drafting assistance.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1861435/full#supplementary-material
Supplementary Figure 1VARS2 is conserved across vertebrates and efficiently knocked down in zebrafish. (A) Multiple sequence alignment of VARS2 protein sequences from representative vertebrate species. (B) Predicted protein structures of VARS2 in human, mouse, and zebrafish. (C) RT-qPCR analysis of vars2 expression in zebrafish embryos injected with STD MO or vars2 MO. (C′) Relative quantification of vars2 expression in the same samples, calculated as 2^−ΔCT and plotted on the indicated scale. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per biological replicate. Bar plots show individual biological replicates. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
Supplementary Figure 2Relative quantification of muscle- and neural-related gene expression in vars2-depleted zebrafish. (A) Relative quantification (2^−ΔCt) of muscle- and neuromuscular junction-related genes in zebrafish embryos injected with STD MO or vars2 MO. (B) Relative quantification (2^−ΔCt) of neural and neurogenesis-related genes in the STD MO and vars2 MO groups. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per biological replicate. Bar plots show individual biological replicates. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
Supplementary Figure 3Relative quantification of mitochondrial and metabolic gene expression in vars2-deficient zebrafish. (A) Relative quantification (2^−ΔCt) of genes related to mitochondrial structure in zebrafish embryos injected with STD MO or vars2 MO. (B) Relative quantification (2^−ΔCt) of genes related to the mitochondrial swelling response in the STD MO and vars2MO groups. (C) Relative quantification (2^−ΔCt) of genes related to mitochondrial energetics in the STD MO and vars2 MO groups. (D) Relative quantification (2^−ΔCt) of genes related to the mitochondrial stress response in the STD MO and vars2 MO groups. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per biological replicate. Bar plots show individual biological replicates. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
Supplementary Figure 4Relative quantification of immune-related gene expression in vars2-deficient zebrafish. (A) Relative quantification (2^−ΔCt) of CD4+ T cell marker genes in zebrafish embryos injected with STD MO or vars2 MO. (B) Relative quantification (2^−ΔCt) of CD8+ T cell marker genes in the STD MO and vars2 MO groups. (C) Relative quantification (2^−ΔCt) of Th17-related genes in the STD MO and vars2 MO groups. (D) Relative quantification (2^−ΔCt) of Treg-related genes in the STD MO and vars2 MO groups. (E) Relative quantification (2^−ΔCt) of neutrophil marker genes in the STD MO and vars2 MO groups. (F) Relative quantification (2^−ΔCt) of macrophage marker genes in the STD MO and vars2 MO groups. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per biological replicate. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
Supplementary Figure 5Relative quantification of immune-, muscle-, and neural-related gene expression following pharmacological interventions in vars2-deficient zebrafish. (A–F) Relative quantification (2^−ΔCt) of (A) CD4+ T cell, (B) CD8+ T cell, (C) Th17, (D) Treg, (E) neutrophil, and (F) macrophage marker genes in the STD MO, vars2 MO, vars2 MO + CoQ10, vars2 MO + neostigmine, and vars2 MO + CoQ10 + neostigmine groups. (G, H) Relative quantification (2^−ΔCt) of (G) muscle-related and (H) neural-related genes in the STD MO, vars2 MO, vars2 MO + ML604440, vars2 MO + GSK805, and vars2 MO + UNBS5162 groups. For RT-qPCR analysis, n = 3 biological replicates, with 20 embryos per biological replicate. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
Supplementary Figure 6Stability of β-actin expression across RT-qPCR in STD MO and vars2 MO groups. (A) Raw β-actin Ct values across all RT-qPCR experiments in the STD MO and vars2 MO groups. Individual values and mean ± SD are shown (P = 0.803). (B) Paired comparison of mean β-actin Ct values (paired t-test, P = 0.681). ns, not significant.
References
1
GilhusNE. Myasthenia gravis. N Engl J Med. (2016) 375:2570–81. doi: 10.1056/nejmra1602678
2
GilhusNEVerschuurenJJ. Myasthenia gravis: subgroup classification and therapeutic strategies. Lancet Neurol. (2015) 14:1023–36. doi: 10.1016/s1474-4422(15)00145-3
3
VinciguerraCBevilacquaLLupicaAGinanneschiFPiscosquitoGRiniNet al. Diagnosis and management of seronegative myasthenia gravis: lights and shadows. Brain Sci. (2023) 13:1286. doi: 10.3390/brainsci13091286
4
EvoliAPalaceJSpagniGCheliMRuiterAVerschuurenJet al. 275th ENMC international workshop: seronegative myasthenia gravis: an update paradigm for diagnosis and management, 9–11 February 2024, Hoofddorp, the Netherlands. Neuromuscul Disord. (2024) 44:104468. doi: 10.1016/j.nmd.2024.104468
5
UzawaAKuwabaraSSuzukiSImaiTMuraiHOzawaYet al. Roles of cytokines and T cells in the pathogenesis of myasthenia gravis. Clin Exp Immunol. (2021) 203:366–74. doi: 10.1111/cei.13546
6
MasudaMMatsumotoMTanakaSNakajimaKYamadaNIdoNet al. Clinical implication of peripheral CD4+CD25+ regulatory T cells and Th17 cells in myasthenia gravis patients. J Neuroimmunol. (2010) 225:123–31. doi: 10.1016/j.jneuroim.2010.03.016
7
XieYLiHFJiangBLiYKaminskiHJKusnerLLet al. Elevated plasma interleukin-17A in a subgroup of myasthenia gravis patients. Cytokine. (2016) 78:44–6. doi: 10.1016/j.cyto.2015.06.011
8
ÇebiMDurmusHAysalFÖzkanBGülGEÇakarAet al. CD4+ T cells of myasthenia gravis patients are characterized by increased IL-21, IL-4, and IL-17A productions and higher presence of PD-1 and ICOS. Front Immunol. (2020) 11:809. doi: 10.3389/fimmu.2020.00809
9
JiaLZhangLLiuMJiHWenZWangCet al. Mitochondrial control for healthy and autoimmune T cells. Cells. (2023) 12:1800. doi: 10.3390/cells12131800
10
JiaoYYanZYangA. Mitochondria in innate immunity signaling and its therapeutic implications in autoimmune diseases. Front Immunol. (2023) 14:1160035. doi: 10.3389/fimmu.2023.1160035
11
ShinBBenavidesGAGengJKoralovSBHuHDarley-UsmarVMet al. Mitochondrial oxidative phosphorylation regulates the fate decision between pathogenic Th17 and regulatory T cells. Cell Rep. (2020) 30:1898–1909.e4. doi: 10.1016/j.celrep.2020.01.022
12
SisslerMGonzález-SerranoLEWesthofE. Recent advances in mitochondrial aminoacyl-tRNA synthetases and disease. Trends Mol Med. (2017) 23:693–708. doi: 10.1016/j.molmed.2017.06.002
13
KonovalovaSTyynismaaH. Mitochondrial aminoacyl-tRNA synthetases in human disease. Mol Genet Metab. (2013) 108:206–11. doi: 10.1016/j.ymgme.2013.01.010
14
KayvanpourEWisdomMLacknerMKSedaghat-HamedaniFBoeckelJNMüllerMet al. VARS2 depletion leads to activation of the integrated stress response and disruptions in mitochondrial fatty acid oxidation. Int J Mol Sci. (2022) 23:7327. doi: 10.3390/ijms23137327
15
BruniFDi MeoIBellacchioEWebbBDMcFarlandRChrzanowska-LightowlersZMAet al. Clinical, biochemical, and genetic features associated with VARS2-related mitochondrial disease. Hum Mutat. (2018) 39:563–78. doi: 10.1002/humu.23398
16
NieASunBFuZYuD. Roles of aminoacyl-tRNA synthetases in immune regulation and immune diseases. Cell Death Dis. (2019) 10:901. doi: 10.1038/s41419-019-2145-5
17
OgnjenovićJSimonovićM. Human aminoacyl-tRNA synthetases in diseases of the nervous system. RNA Biol. (2018) 15:623–34. doi: 10.1080/15476286.2017.1330245
18
OkuzonoYMiyakawaSItouTSagaraMIwataMIshizuchiKet al. B-cell immune dysregulation with low soluble CD22 levels in refractory seronegative myasthenia gravis. Front Immunol. (2024) 15:1382320. doi: 10.3389/fimmu.2024.1382320
19
LeiteMIJacobSViegasSCossinsJCloverLMorganBPet al. IgG1 antibodies to acetylcholine receptors in ‘seronegative’ myasthenia gravis. Brain. (2008) 131:1940–52. doi: 10.1093/brain/awn092
20
PorcuESadlerMCLepikKAuwerxCWoodARWeihsAet al. Differentially expressed genes reflect disease-induced rather than disease-causing changes in the transcriptome. Nat Commun. (2021) 12:5647. doi: 10.1038/s41467-021-25805-y
21
NathanAAsgariSIshigakiKValenciaCAmariutaTLuoYet al. Single-cell eQTL models reveal dynamic T cell state dependence of disease loci. Nature. (2022) 606:120–28. doi: 10.1038/s41586-022-04713-1
22
HudaRH. Inflammation and autoimmune myasthenia gravis. Front Immunol. (2023) 14:1110499. doi: 10.3389/fimmu.2023.1110499
23
LeeGR. The balance of Th17 versus Treg cells in autoimmunity. Int J Mol Sci. (2018) 19:730. doi: 10.3390/ijms19030730
24
WenYYangBLuJZhangJYangHLiJet al. Imbalance of circulating CD4+CXCR5+FOXP3+ Tfr-like cells and CD4+CXCR5+FOXP3-2011; Tfh-like cells in myasthenia gravis. Clin Exp Med. (2016) 16:565–73. doi: 10.1016/j.neulet.2016.07.049
25
AshidaSOchiHHamataniMFujiiCKimuraKOkadaYet al. Immune skew of circulating follicular helper T cells associates with myasthenia gravis severity. Neurol Neuroimmunol Neuroinflamm. (2021) 8:e945. doi: 10.1212/NXI.0000000000000945
26
WuYWLuoJLGardenOA. Immunoregulatory cells in myasthenia gravis. Front Neurol. (2020) 11:593431. doi: 10.3389/fneur.2020.593431
27
MaitiAKimJI. The alterations in and the role of the Th17/Treg balance in metabolic diseases. Front Immunol. (2021) 12:678355. doi: 10.3389/fimmu.2021.678355
28
TyynismaaH. Disease models of mitochondrial aminoacyl-tRNA synthetase defects. J Inherit Metab Dis. (2023) 46:406–21. doi: 10.1002/jimd.12652
29
ZhaoRLuoSZhaoC. The role of innate immunity in myasthenia gravis. Autoimmun Rev. (2021) 20:102800. doi: 10.1016/j.autrev.2021.102800
30
DangEVBarbiJYangHYJinasenaDYuHZhengYet al. Control of TH17/Treg balance by hypoxia-inducible factor 1. Cell. (2011) 146:772–84. doi: 10.1016/j.cell.2011.07.033
31
PayetCAYouAFayetOMHemeryETruffaultFBondetVet al. Central role of macrophages and nucleic acid release in myasthenia gravis thymus. Ann Neurol. (2023) 93:643–54. doi: 10.1002/ana.26590
32
VerdierJFayetOMHemeryETruffaultFPinzónNDemeretSet al. Single-cell mass cytometry on peripheral cells in myasthenia gravis identifies dysregulation of innate immune cells. Front Immunol. (2023) 14:1083218. doi: 10.3389/fimmu.2023.1083218
33
MeraounaACizeron-ClairacGLe PanseRBismuthJTruffaultFTallaksenCet al. The chemokine CXCL13 is a key molecule in autoimmune myasthenia gravis. Blood. (2006) 108:432–40. doi: 10.1182/blood-2005-06-2383
34
ShiaoYMLeeCCHsuYHHuangSFLinCYLiLHet al. Ectopic and high CXCL13 chemokine expression in myasthenia gravis with thymic lymphoid hyperplasia. J Neuroimmunol. (2010) 221:101–6. doi: 10.1016/j.jneuroim.2010.02.013
35
VõsaUClaringbouldAWestraHJBonderMJDeelenPZengBet al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet. (2021) 53:1300–10. doi: 10.1038/s41588-021-00913-z
36
GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. (2020) 369:1318–30. doi: 10.1126/science.aaz1776
37
ChiaRSaez-AtienzarSMurphyNChioABlauwendraatCRodaRHet al. Identification of genetic risk loci and prioritization of genes and pathways for myasthenia gravis: a genome-wide association study. Proc Natl Acad Sci USA. (2022) 119(5):e2108672119. doi: 10.1073/pnas.2108672119
38
KurkiMIKarjalainenJPaltaPSipilaTPKristianssonKDonnerKMet al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. (2023) 613:508–18. doi: 10.1038/s41586-022-05473-8
39
WillerCJLiYAbecasisGR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. (2010) 26:2190–1. doi: 10.1093/bioinformatics/btq340
40
ZhuZZhangFHuHBakshiARobinsonMRPowellJEet al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. (2016) 48:481–7. doi: 10.1038/ng.3538
41
ChenJXuFRuanXSunJZhangYZhangHet al. Therapeutic targets for inflammatory bowel disease: proteome-wide Mendelian randomization and colocalization analyses. EBioMedicine. (2023) 89:104494. doi: 10.1016/j.ebiom.2023.104494
42
WallaceC. A more accurate method for colocalisation analysis allowing for multiple causal variants. PLoS Genet. (2021) 17:e1009440. doi: 10.1371/journal.pgen.1009440
43
ChenSHuangWWanQTangZLiXZengFet al. Investigation of the acute pathogenesis of spondyloarthritis/HLA-B27-associated anterior uveitis based on genome-wide association analysis and single-cell transcriptomics. J Transl Med. (2024) 22:271. doi: 10.1186/s12967-024-05077-y
44
HaoYStuartTKowalskiMHChoudharySHoffmanPHartmanAet al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol. (2024) 42:293–304. doi: 10.1038/s41587-023-01767-y
45
IanevskiAGiriAKAittokallioT. Fully-automated and ultra-fast cell-type identification using specific marker combinations from single-cell transcriptomic data. Nat Commun. (2022) 13. doi: 10.1038/s41467-022-28803-w
46
LyuZKouYFuYXieYYangBZhuHet al. Comparative transcriptomics revealed neurodevelopmental impairments and ferroptosis induced by extremely small iron oxide nanoparticles. Front Genet. (2024) 15:1402771. doi: 10.3389/fgene.2024.1402771
47
LuSLyuZWangZKouYLiuCLiSet al. Lipin 1 deficiency causes adult-onset myasthenia with motor neuron dysfunction in humans and neuromuscular junction defects in zebrafish. Theranostics. (2021) 11:2788–805. doi: 10.7150/thno.53330
48
ZhuHWangZWangWLuYHeYWTianJ. Bacterial quorum-sensing signal DSF inhibits LPS-induced inflammations by suppressing toll-like receptor signaling and preventing lysosome-mediated apoptosis in zebrafish. Int J Mol Sci. (2022) 23:7110. doi: 10.3390/ijms23137110
49
LyuZDaYLiuHWangZZhuYTianJ. Chsy1 deficiency reduces extracellular matrix productions and aggravates cartilage injury in osteoarthritis. Gene. (2022) 827:146466. doi: 10.1016/j.gene.2022.146466
Summary
Keywords
immunometabolism, mitochondrial dysfunction, myasthenia gravis (MG), neuromuscular junction, seronegative myasthenia gravis (SNMG), VARS2, zebrafish model
Citation
Lyu Z, Yang B, Zhou Y, Shi M, Wang Z, Song S, Zhang X and Tian J (2026) VARS2 deficiency drives seronegative myasthenia gravis pathology via mitochondrial, immune, and neuromuscular dysregulation. Front. Immunol. 17:1861435. doi: 10.3389/fimmu.2026.1861435
Received
21 April 2026
Revised
18 August 2026
Accepted
21 August 2026
Published
18 September 2026
Volume
17 - 2026
Edited by
Marc De Baets, Maastricht University, Netherlands
Reviewed by
Amol K. Bhandage, Uppsala University, Sweden
Shigeaki Suzuki, Keio University, Japan
Bo Wang, The First Affiliated Hospital of China Medical University, China
Updates
Copyright
© 2026 Lyu, Yang, Zhou, Shi, Wang, Song, Zhang and Tian.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Jing Tian, tianjing@nwu.edu.cn; Xinwen Zhang, xianxinwen@163.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.