Abstract

Introduction

   The comorbidity of optic neuritis with multiple sclerosis has been well
   recognized. However, the causal association between multiple sclerosis
   and optic neuritis, as well as other eye disorders, remains
   incompletely understood. To address these gaps, we investigated the
   genetically relationship between multiple sclerosis and eye disorders,
   and explored potential drugs.

Methods

   In order to elucidate the genetic susceptibility and causal links
   between multiple sclerosis and eye disorders, we performed two-sample
   Mendelian randomization analyses to examine the causality between
   multiple sclerosis and eye disorders. Additionally, causal
   single-nucleotide polymorphisms were annotated and searched for
   expression quantitative trait loci data. Pathway enrichment analysis
   was performed to identify the possible mechanisms responsible for the
   eye disorders coexisting with multiple sclerosis. Potential therapeutic
   chemicals were also explored using the Cytoscape.

Results

   Mendelian randomization analysis revealed that multiple sclerosis
   increased the incidence of optic neuritis while reducing the likelihood
   of concurrent of cataract and macular degeneration. Gene Ontology
   enrichment analysis implicated that lymphocyte proliferation,
   activation and antigen processing as potential contributors to the
   pathogenesis of eye disorders coexisting with multiple sclerosis.
   Furthermore, pharmaceutical agents traditionally employed for allograft
   rejection exhibited promising therapeutic potential for the eye
   disorders coexisting with multiple sclerosis.

Discussion

   Multiple sclerosis genetically contributes to the development of optic
   neuritis while mitigating the concurrent occurrence of cataract and
   macular degeneration. Further research is needed to validate these
   findings and explore additional mechanisms underlying the comorbidity
   of multiple sclerosis and eye disorders.

   Keywords: Mendelian randomization, genetic susceptibility, causal
   pathway, multiple sclerosis, eye disorder, innate immunity

1. Introduction

   Multiple sclerosis (MS) is an autoimmune disease characterized by
   inflammation, demyelination, and axonal loss in the central nervous
   system (CNS). Over the years, the prevalence and incidence rates of MS
   have been continuously increasing in both developed and developing
   countries ([27]1, [28]2). Ocular pathology is a common occurrence in up
   to 80% of MS patients throughout the course of their disease ([29]3).
   Among which, acute optic neuritis affects approximately 30% to 70% of
   individuals with MS at some point ([30]4). Unfortunately, like MS,
   there is no specific cure for ocular pathology coexisting with MS. For
   optic neuritis coexisting with MS, the most commonly used
   corticotherapy may accelerate the visual recovery, but it has no
   long-term visual benefit ([31]5). In addition, it is not the best
   treatment option for other complications of MS ([32]6). Therefore, it
   is important to target prevention and new therapy of ocular pathology
   coexisting with MS by means of establishing causal links.

   Furthermore, studies utilizing the 25-Item National Eye Institute
   Visual Functioning Questionnaire and a 10-Item Neuro-Ophthalmic
   Supplement have identified additional eye-related symptoms in MS
   patients, including decreased visual acuity, contrast sensitivity,
   binocular vision defects, visual field abnormalities, reduced color
   vision, blurred vision, and diplopia ([33]3). The presence of these
   clinical symptoms raises questions about the potential involvement of
   other eye disorders in the progression of MS, aside from optic
   neuritis. In clinical practice, decreased visual acuity and blurred
   vision are common symptoms among patients with eye disorders ([34]7).
   These disorders encompass refractive errors (myopia, hyperopia, and
   astigmatism), macular degeneration (dry and wet), glaucoma, and
   cataracts (senile and other types). However, the causal relationship
   between these eye disorders and MS still remains elusive.

   Mendelian randomization (MR) analysis, which follows the same design
   principles as a randomized controlled trial (RCT), utilizes
   single-nucleotide polymorphisms (SNPs) as instrumental variables (IVs)
   to discover a causal relationship between exposures and outcomes
   ([35]8). Given the clinical coexistence of MS and eye disorders, we
   employ MR analysis to examine the genetic relationships between these
   two conditions. Furthermore, the extracted causal SNPs were annotated
   and the targeted gene expression was identified using the expression
   quantitative trait loci (eQTL) database. The biological function
   analysis of the targeted genes revealed that lymphocyte proliferation,
   activation and antigen processing might be involved in the pathogenesis
   of eye disorders coexisting with MS. Additionally, we also predicted
   that the drugs used for allograft rejection could potentially serve as
   a viable therapeutic intervention for patients simultaneously suffering
   from ocular disorders and MS. Overall, our study provides a new insight
   into the pathogenesis of MS coexisting with eye disorders.

2. Materials and methods

2.1. GWAS summary statistic data

   We investigated the genetic causality between MS and five commonly
   occurring eye disorders that might be misdiagnosed in the early phase
   of MS, namely optic neuritis, glaucoma, cataract, refraction disorder
   and macular degeneration. Although the symptoms of these eye disorders
   are all related to vision loss, the etiological factors vary among
   them. Optic neuritis is caused by the immune system that mistakenly
   targets the optic nerve, resulting in optic nerve inflammation.
   Glaucoma is a group of eye conditions (raised intraocular pressure,
   inflammation etc.) that damage the optic nerve. Cataract is a clouding
   of the lens, which can be caused by increasing aging (senile cataract),
   eye injury and inflammation etc. (other cataract). Refraction disorder
   occurs when the shape alteration of the lens prevents light from
   focusing directly on the retina. There are two types of macular
   degeneration. Dry age-related macular degeneration is due to the
   breaking down of the inner layers of the macula, and the wet
   age-related macular degeneration is caused by blood vessels that leak
   fluid or blood into the macula. The comprehensive design of the current
   study is depicted in [36]Figure 1. We obtained GWAS summary statistic
   data from the GWAS catalog ([37]https://gwas.mrcieu.ac.uk/). Four MS
   traits, including finn-b-G6_MS, ukb-b-17670, ebi-a-GCST005531 and
   ieu-a-1025 ([38]9, [39]10), were selected to validate the results. For
   the eye disorders, we utilized the most recent traits with European
   ancestry and the largest sample size available. [40]Table 1 provides
   detailed information about these traits. Specifically, the traits of
   eye disorders contain finn-b-H7_OPTNEURITIS, finn-b-H7_GLAUCSECINFLAM,
   finn-b-H7_CATARACTSENILE, finn-b-H7_CATARACTOTHER, finn-b-H7_OCUMUSCLE,
   finn-b-DRY_AMD and finn-b-WET_AMD.

Figure 1.

   [41]Figure 1
   [42]Open in a new tab

   Flowchart for Mendelian randomization analysis. (A) Directed acyclic
   graph representing the Mendelian randomization framework for
   investigating the causal relationship between MS and eye disorders.
   Assumptions for instrumental variables: (1) The instruments must be
   associated with the exposures; (2) The instruments must influence eye
   disorders only through their impact on MS; and (3) The instruments must
   remain independent of both observed and unobserved confounding factors
   affecting the relationship between MS and eye disorders. (B) The
   work-flow of this study. Summary data of GWAS traits were extracted for
   detecting causal relationships between MS and eye disorders (optic
   neuritis, cataract, glaucoma, macular degeneration and refraction
   disorder). Subsequently, the causal SNPs were then annotated and mapped
   to the plasma eQTL. Finally, the identified putatively causal genes
   were used for enrichment analysis.

Table 1.

   Detailed information of the GWAS datasets explored and prioritized
   instruments used for the Mendelian randomization analysis.
   GWAS ID Exposure Year Sample size NO. of SNPs P value
   finn-b-G6_MS Multiple sclerosis 2021 1048 cases/217141 controls
   16380460 5×10^-8
   ukb-b-17670 Multiple sclerosis 2018 1679 cases/461251 controls 9851867
   5×10^-8
   ebi-a-GCST005531 Multiple sclerosis 2013 14498 cases/24091 controls
   132089 5×10^-8
   ieu-a-1025 Multiple sclerosis 2013 14498 cases/24091 controls 156632
   5×10^-8
   finn-b-H7_OPTNEURITIS Optic neuritis 2021 582 cases/217491 controls
   16380463 5×10^-6
   finn-b-H7_GLAUCSECINFLAM Glaucoma secondary to
   eye inflammation 2021 96 cases/210201 controls 16380446 5×10^-6
   finn-b-H7_CATARACTSENILE Senile cataract 2021 26758 cases/189604
   controls 16380461 5×10^-6
   finn-b-H7_CATARACTOTHER Other cataract 2021 7873 cases/189604 controls
   16380403 5×10^-6
   finn-b-H7_OCUMUSCLE Refraction disorder 2021 7861cases/210931 controls
   16380466 5×10^-6
   finn-b-DRY_AMD Dry age-related macular degeneration 2021 2469
   cases/206221 controls 16380423 5×10^-6
   finn-b-WET_AMD Wet age-related macular degeneration 2021 2114
   cases/206601 controls 16380422 5×10^-6
   [43]Open in a new tab

2.2. Selection of genetic instruments

   The SNPs used as exposure IVs for MS were selected with a p value less
   than 5 × 10^-8. However, in the eye disorder traits, the p-value
   cut-off was set as 5 × 10^-6 since there was a lack of enough IVs for
   evaluation at p value of 5 × 10^-8. In addition, the linkage
   disequilibrium in the selected SNPs was clumped with r^2 <0.1 based on
   the 1000 Genomes Project. The F-statistics of each selected IV were
   calculated using the mRnd method ([44]11). The IVs with F-statistics
   less than 10 were excluded to ensure the strength of instruments.

2.3. Two-sample MR analysis

   Two-sample MR analyses were conducted using three models:
   inverse-variance weighted (IVW), MR-Egger (ME), and weighted median
   (WM). The palindromic SNPs with intermediate allele frequencies were
   excluded from the MR analysis. The odds ratio (OR) and corresponding
   95% confidence interval (CI) were calculated to show the relationship
   between exposure variables and outcome variables. We also performed
   Cochran’s Q test to assess the horizontal heterogeneity. A Cochrane’s Q
   value in the MR–Egger regression and IVW tests was evaluated.
   Heterogeneity was considered statistically significant when the P value
   of the Q value (P[Q-value]) was less than 0.05. In addition, MR-Egger
   regression was used to evaluate the pleiotropic effects of IVs. All
   analyses were performed using the TwoSampleMR package (version 0.5.6)
   in R software (version 4.1.2).

2.4. Functional evaluation of causal SNPs

   The basic information about the causal SNPs, including rsID, location,
   and effect allele, was extracted using the ‘mr_singlesnp’ function in
   the TwoSampleMR R package. To explore whether these causal SNPs affect
   gene expression, we analyzed eQTL-targeted genes in whole blood and
   identified motif alterations using the VannoPortal database ([45]12).
   Additionally, we conducted Gene Ontology (GO) enrichment analysis and
   network cluster analysis of targeted genes of eQTL using STRING to
   elucidate the affected biological processes. Furthermore, curated
   pathway was explored using Cytoscape to identify potential therapeutic
   targets and pharmaceutical agents for the eye disorders coexisting with
   multiple sclerosis.

3. Results

3.1. The presence of MS facilitates optic neuritis pathology

   To explore the genetic relationship between MS and optic neuritis, a
   two-sample MR was performed. Four different MS datasets (GWAS ID:
   finn-b-G6_MS, ukb-b-17670, ebi-a-GCST005531, ieu-a-1025) and one optic
   neuritis dataset (GWAS ID: finn-b-H7_OPTNEURITIS) were selected
   ([46]Table 1). The MR results from the IVW method showed that MS was
   causally associated with a higher risk of optic neuritis (finn-b-G6_MS:
   OR = 1.92, 95% CI = 1.60 - 2.30, P = 1.04× 10^-12; ukb-b-17670: OR =
   59.65, 95% CI = 27.43 - 91.88, P = 2.85× 10^-4; ebi-a-GCST005531: OR=
   1.77, 95% CI = 1.58 - 1.98, P = 3.18× 10^-23; ieu-a-1025: OR = 1.68,
   95% CI = 1.48 - 1.92, P = 2.12× 10^-15, [47]Figures 2, [48]3).
   Considering potential horizontal heterogeneity or pleiotropy in the SNP
   effects ([49]Table 2), the MR-Egger and Weighted median methods were
   used to validate the causal effect. The results confirmed the robust
   and significant causal relationship between MS and optic neuritis
   ([50]Figures 2, [51]3).

Figure 2.

   [52]Figure 2
   [53]Open in a new tab

   Causal relationship between multiple sclerosis and the risk of eye
   disorders. Data were presented as OR with 95% CI. Three GWAS traits of
   multiple sclerosis, including finn-b-G6_MS, ebi-a-GCST005531 and
   ieu-a-1025, were utilized as exposure variables respectively. Optic
   neuritis, senile cataract, other cataract, dry age-related macular
   degeneration and wet age-related macular degeneration were considered
   as outcomes. Three Mendelian randomization analysis methods were
   applied: the IVW, MR Egger and weighted median method. The nSNP refers
   to the number of SNP.

Figure 3.

   [54]Figure 3
   [55]Open in a new tab

   Mendelian randomization analysis assessing the causal link between
   multiple sclerosis (dataset: ubk-b-17670) and the risk of eye
   disorders. Data were presented as log-transformed odds ratios (log-OR)
   along with 95% CI. The genetic traits associated with multiple
   sclerosis from the ukb-b-17670 were employed as the exposure of
   interest. The considered eye disorders encompass optic neuritis, senile
   cataract, other cataract, dry age-related macular degeneration, and wet
   age-related macular degeneration as the outcome variables. Three
   distinct models, including the IVW, MR Egger and weighted median models
   were utilized for Mendelian randomization analysis. The nSNP refers to
   the number of SNP.

Table 2.

   Heterogeneity and pleiotropy tests of the Mendelian randomization
   analysis between MS and eye disorders.
   Exposure Outcome Q value P[Q-value] Intercept P[intercept]
   MR Egger Inverse- variance weighted MR Egger Inverse- variance weighted
    finn-b-G6_MS Optic neuritis 6.16 9.22 0.10 0.06 -0.21 0.31
   Senile cataract 3.47 7.86 0.32 0.10 -0.05 0.15
   Other cataract 9.39 10.82 0.02 0.03 -0.04 0.55
   Dry age-related macular degeneration 0.82 1.96 0.84 0.74 0.07 0.36
   Wet age-related macular degeneration 6.31 9.11 0.10 0.06 0.11 0.33
    ebi-a-
   GCST005531 Optic neuritis 42.58 42.58 0.53 0.58 5.81E-04 0.98
   Senile cataract 87.48 88.95 1.05E-03 1.03E-03 4.47E-03 0.39
   Other cataract 53.35 60.05 0.16 0.07 0.01 0.02
   Dry age-related macular degeneration 40.46 41.91 0.62 0.60 0.01 0.24
   Wet age-related macular degeneration 40.76 49.46 0.36 0.30 0.02 0.12
    ieu-a-1025 Optic neuritis 50.96 52.03 0.32 0.32 0.02 0.33
   Senile cataract 80.08 80.92 1.86E-03 2.07E-03 3.68E-03 0.49
   Other cataract 61.15 62.74 0.08 0.07 7.63E-03 0.27
   Dry age-related macular degeneration 41.04 41.32 0.72 0.74 0.01 0.60
   Wet age-related macular degeneration 53.89 55.67 0.23 0.21 0.02 0.22
    ukb-b-17670 Optic neuritis 6.64 18.30 0.08 1.08E-03 -0.18 0.11
   Senile cataract 8.66 9.60 0.03 0.05 9.66E-03 0.61
   Other cataract 8.81 10.70 0.03 0.03 0.02 0.48
   Dry age-related macular degeneration 7.05 7.79 0.07 0.10 -0.02 0.62
   Wet age-related macular degeneration 3.01 3.08 0.39 0.54 0.01 0.80
   [56]Open in a new tab

   The reversal MR analyses (IVW method) showed that the causal link
   between optic neuritis and MS was only observed in the two selected
   datasets (finn-b-G6_MS: OR = 1.76, 95% CI = 1.27–2.44, P = 7.16× 10^-4;
   ukb-b-17670: OR = 1.00, 95% CI = 1.00–1.00, P = 3.90× 10^-5,
   [57]Figure 4). The outcome datasets ebi-a-GCST005531 and ieu-a-1025
   lacked an adequate number of IVs to conduct a MR analysis, indicating
   that optic neuritis is not strongly associated with the risk of MS.
   Collectively, the presence of MS is postulated to potentially increase
   the risk of optic neuritis but not vice versa. This finding is
   consistent with the widely accepted notion that optic neuritis is one
   of the most common complications of MS in clinics, and provide the
   genetic information to support the theory.

Figure 4.

   [58]Figure 4
   [59]Open in a new tab

   Reverse Mendelian randomization estimating the causal relationship
   between eye disorders and multiple sclerosis. The data are expressed in
   the form of OR accompanied by 95% CI. The different eye disorders
   served as the exposure variables, while MS was the designated outcomes.
   Only datasets for MS that contained a sufficient number of SNPs were
   included in the analysis. Three distinct Mendelian Randomization models
   were employed, including the IVW method, MR Egger, and the Weighted
   Median method. The nSNP refers to the number of SNP.

3.2. MS is genetically associated with a decreased risk of non-senile
cataract wet age-related macular degeneration

   We then explored the causal relationship between MS and other eye
   disorders, including cataract, macular degeneration, glaucoma and
   refraction disorder, since patients with these diseases also having
   similar vision problems as MS patients. There was a lack of genetic
   causality between MS and senile cataract and dry macular degeneration
   ([60]Figures 2, [61]3). However, the MR analysis (IVW method) revealed
   that MS was genetically related to a lower risk of other type of
   cataract (finn-b-G6_MS, OR = 0.89, 95% CI = 0.85–0.94, P = 5.22× 10^-5;
   ukb-b-17670, OR = -8.25, 95% CI = -15.36–1.14, P = 0.02;
   ebi-a-GCST005531, OR= 0.94, 95% CI = 0.90–0.97, P = 1.19× 10^-3;
   ieu-a-1025, OR = 0.96, 95% CI = 0.92–1.00, P = 0.05, [62]Figures 2,
   [63]3). The similar results were found using the MR-Egger
   (ebi-a-GCST005531: OR= 0.88, 95% CI = 0.82–0.94, P = 4.13× 10^-4,
   [64]Figures 2, [65]3) and weighted median methods (finn-b-G6_MS: OR =
   0.88, 95% CI = 0.85–0.92, P = 9.52× 10^-9; ukb-b-17670: OR = -10.56,
   95% CI = -15.61–5.51, P = 4.17× 10^-5, [66]Figures 2, [67]3),
   indicating a stable and strong causality between MS and other type of
   cataract. In addition, causality between MS and wet age-related macular
   degeneration was found in three MS datasets using the IVW method
   (finn-b-G6_MS, OR = 0.88, 95% CI = 0.79 –0.98, P = 0.01;
   ebi-a-GCST005531, OR = 0.90, 95% CI = 0.83 –0.96, P = 1.5× 10^-3;
   ieu-a-1025, OR = 0.91, 95% CI = 0.84 –0.99, P = 0.01) ([68]Figures 2,
   [69]3), indicating a weak causality between MS and wet age-related
   macular degeneration. However, we did not find any significant
   causality between MS and glaucoma or refraction disorder
   ([70]Supplementary Figure 1), and vice versa ([71]Figure 4;
   [72]Supplementary Figure 1). In summary, our findings suggest that MS
   was negatively associated with other types of cataracts and wet
   age-related macular degeneration, but not with senile cataract, dry
   age-related macular degeneration, glaucoma, or refraction disorders.

3.3. The causal association between MS and eye disorders could be mediated by
the lymphocyte proliferation, activation and antigen processing

   We collected a total of 69 causal SNPs from the exposure of four
   different MS datasets. Among them, 36 SNPs were shared by both
   ebi-a-GCST005531 and ieu-a-1025 datasets ([73]Figure 5A). The causal
   SNPs were annotated using the VannoPortal Index database ([74]Table 3),
   and it was found that the associated genes were related to JAK-STAT
   signaling pathway (IL12a, IL2Ra, IL20Ra, STAT3, STAT4, IL7R) and
   adaptive immunity (CD58, CD69, CD86, TNFSF14) by STRING clustering
   (Data not shown).

Figure 5.

   [75]Figure 5
   [76]Open in a new tab

   Enrichment analysis and the network of the eQTL-regulated genes from
   all causal SNPs. (A) A Venn diagram showing the shared SNPs among
   various datasets. (B) A Venn diagram illustrating the overlapping
   eQTL-regulated genes across distinct datasets. (C) Gene ontology
   enrichment analysis illustrating the biological processes enriched of
   eQTL-regulated genes from all causal SNPs (N = 69). (D) Gene ontology
   enrichment analysis illustrating the biological processes enriched of
   eQTL-regulated genes from the SNPs shared by ebi-a-GCST005531 and
   ieu-a-1025 (N = 36). (E) Network analysis of eQTL-regulated genes from
   all causal SNPs. (F) Network analysis of eQTL-regulated genes from the
   SNPs shared by ebi-a-GCST005531 and ieu-a-1025 (N = 36).

Table 3.

   eQTL-regulated genes of the causal SNPs from four MS datasets.
   Exposure SNP Effect Allele Nearest Gene eQTL-regulated genes
   finn-b-G6_MS rs141298848 T PDCD6IP ——
   specific (n = 5) rs149765808 A ZSCAN12P1 ——
     rs55782725 A MICA MICA, NCR3, PSORS1C1, C4A, PSORS1C2
     rs9264277 C HLA-C HCG27, XXbac-BPG181B23.7, PSORS1C3, MIR6891, HLA-L
     rs9271069 G HLA-DRB1 HLA-DRB5, HLA-DRB6, HLA-DQB1, HLA-DQB1-AS1,
   HLA-DQA2
    ukb-b-17670 rs28746956 G HLA-DQB1 HLA-DQB2, HLA-DQB1, HLA-DQB1-AS1,
   HLA-DQA2, HLA-DQA
    specific (n = 5) rs4899257 T ZFP36L1, RAD51B CTD-2325P2.4, ZFP36L1
     rs67382147 G HLA-DRB5 HLA-DRB5, HLA-DRB6, HLA-DQB1, HLA-DQB1-AS1,
   HLA-DRB1
     rs7759971 T AHI1 AHI1, RP3-388E23.2, LINC00271
     rs9378141 C MICD, HLA-W HLA-J, IFITM4P, HLA-K, HCG4, MICD
    ieu-a-1025 rs1131265 C TIMMDC1 POGLUT1, ARHGAP31, TIMMDC1, RPL10P7
   specific (n = 13) rs115266049 A TMPOP1, MICC ——
     rs115985474 T TRIM31, RNF39 RPL23AP1, HLA-F, HCG4P7, IFITM4P, RNF39
     rs12210359 T MCCD1P2, HCG4P3 HLA-J, HLA-K, HLA-A, MICD, SFTA2
     rs12296430 C LTBR, RP1-102E24.1 RP1-102E24.8, LTBR
     rs1359062 G RGS1, RGS21 RGS1, RP5-101101.2
     rs212405 T TAGAP, FNDC1 RP1-111C20.4, TAGAP
     rs2857700 C NCR3, UQCRHP1 HLA-DRB5, HLA-DRB6, HCG27, LY6G5B, HLA-C
     rs3129727 T MTCO3P1, HLA-DQB3 HLA-DQB1-AS1, HLA-DQA2, HLA-DQB1,
   HLA-DRB6, HLA-DQB
     rs9263823 T POU5F1 PSORS1C1, HLA-C, XXbac-BPG248L24.12, HCG22, POU5F1
     rs9277535 G HLA-DPB1 HLA-DPB2, RPL32P1, HLA-DPA1, HSD17B8, HLA-DPB
     rs9736016 A CXCR5, SETP16 DDX6, TREH, AP002954.4, JAML, TREHP1
     rs9989735 C SP140 ——
    ebi-a- rs1323292 A RGS1, RGS21 RGS1, RP5-1011O1.2, RGS2
    GCST005531 rs13426106 A SP140 SP140, [77]AC009950.2
    specific (n = 10) rs57271503 A CD80 POGLUT1, ARHGAP31, TIMMDC1,
   RPL10P7
     rs3115627 G MICF, HCG4P7 HLA-J, IFITM4P, HLA-K, HCG4, RPL23AP1
     rs9277641 A HLA-DPB2 HLA-DPA1, COL11A2, HLA-DPB1, WDR46, HLA-DPB2
     rs4947337 C C6orf10 ——
     rs212407 A TAGAP, FNDC1 RP1-111C20.4, TAGAP
     rs3130283 C AGPAT1 HLA-DRB5, HLA-DRB6, HLA-DRB1, HLA-DQB1-AS1,
   HLA-DQB1
     rs10892299 T CXCR5 ——
     rs2364482 G LTBR, RP1-102E24.1 LTBR, RP1-102E24.1
    ebi-a- rs1014486 C IL12A IL12A, IL12A-AS1
    GCST005531 rs1021156 C ZC2HC1A, PKIA RP11-578O24.2, IL7, ZC2HC1A,
   PKIA, THAP12P7
   and ieu-a-1025 rs10420809 T SLC44A2 SLC44A2, KRI1
   shared (n = 36) rs1077667 T TNFSF14 ——
     rs11052877 G CD69 CLECL1, RP11-705C15.2, DDX12P, CLEC2D,
   RP11-705C15.3
     rs11154801 A AHI1 AHI1, RP3-388E23.2, LINC00271
     rs11172342 T TSFM METTL21B, TSFM, ATP23, TSPAN31, RP11-571M6.17
     rs11554159 A IFI30, PIK3R2 MPV17L2, KIAA1683, PDE4C, MAST3, IL12RB1
     rs11865086 A MAPK3 BOLA2, TBX6, NPIPB12, INO80E, YPEL3
     rs12087340 T BCL10, DDAH1 BCL10
     rs12927355 T CLEC16A RP11-66H6.4, HNRNPCP4, RMI2
     rs17066096 G IL20RA ——
     rs1800693 C TNFRSF1A TNFRSF1A
     rs1813375 T EOMES ——
     rs2104286 C IL2RA IL2RA
    ebi-a- rs34383631 T CD6 CD6
    GCST005531 rs3748817 C MMEL1 TTC34, MMEL1, FAM213B, TNFRSF14,
   RP3-395M20.8
   and ieu-a-1025 rs41286801 T EVI5 EVI5, GFI1
   shared (n = 36) rs4410871 C MYC PVT1
     rs4780355 C RMI2 ——
     rs4796791 C STAT3 STAT3, PTRF, RAB5C, TUBG2, SPB9
     rs4944958 G NADSYN1 RP11-660L16.2, NADSYN1, KRTAP5-7, KRTAP5-10,
   KRTAP5-9
     rs4976646 C RGS14 RGS14, MXD3, FGFR4, LMAN2, F12
     rs60600003 G ELMO1 ——
     rs6677309 C CD58 CD58, NAP1L4P1, RP5-1086K13.1
     rs67297943 C TNFAIP3 ——
     rs6881706 T IL7R LMBRD2, IL7R
     rs706015 G SKAP2 HOTAIRM1, HOXA5, HOXA-AS3, HOXA-AS2, SKAP2
     rs71624119 A ANKRD55 IL6ST, ANKRD55
     rs74796499 A GALC GPR65, LINC01146
     rs7783 G CHERP, C19orf44 ——
     rs7923837 A EXOC6 EIF2S2P3, HHEX, TNKS2-AS1, FGFBP3
     rs8070345 C VMP1 RNFT1, RP11-178C3.2, DHX40, DHX40P1, RPS6KB1
     rs842639 A REL C2orf74, PUS10, AHSA2, RP11-493E12.2, PAPOLG
     rs9282641 A CD86 IQCB1, SLC15A2
     rs9967792 C STAT4 STAT4
   [78]Open in a new tab

   To further clarify the function of the causal SNPs, we searched for
   eQTL in whole blood ([79]Table 3). A Venn diagram was used to visualize
   the shared eQTL genes in these four datasets, revealing four human
   leukocyte antigen (HLA) genes, including HLA-DRB5, HLA-DQB1-AS1,
   HLA-DRB6 and HLA-DQB1 ([80]Figure 5B). GO enrichment analyses showed
   that lymphocyte proliferation and activation were enriched
   ([81]Figure 5C). Additionally, the interleukin-35-mediated signaling
   pathway was enriched based on the genes associated with the shared SNPs
   between ebi-a-GCST005531 and ieu-a-1025 datasets ([82]Figure 5D). Using
   the remaining SNPs (except for the shared ones), the
   interferon-gamma-mediated signaling pathway was identified
   ([83]Supplementary Figure 3). STRING Protein-Protein interaction
   networks functional enrichment analysis showed that all the eQTL
   associated proteins were categorized into three clusters, mainly
   focused on MS risk genes, IL-6-mediated signaling pathway and the
   G-protein signaling domain ([84]Figure 5E). The shared eQTL associated
   proteins were also categorized into three clusters, mainly enriched in
   cytokine signaling pathway ([85]Figure 5F).

3.4. Potential therapeutic chemicals for MS coexisting with eye disorders

   All the eQTL-regulated genes were used to explore the curated pathway
   using Cytoscope. Genes were mainly enriched in the “allograft
   rejection” pathway. Additionally, ten clinically used drugs were
   identified as inhibitors of this biological process. These chemicals
   targeted cytokine-mediated lymphocyte activation (sirolimus,
   tacrolimus, and ciclosporin for IL2-induced CD8^+ T cell activation;
   Daclizumab and basiliximab for IL2RA-induced Th1 cell activation;
   Fenofibrate for IFEN-induced CD4^+ T cell activation;
   Methylprednisolone, prednisone, and fenofibrate for Th17 cell
   activation), antigen processing (belatacept), and the complement system
   (eculizumab). A detailed list of the information and targets of these
   chemicals was shown in [86]Figure 6.

Figure 6.

   [87]Figure 6
   [88]Open in a new tab

   Potential drug discovering for MS coexisting with eye Disorders based
   on eQTL-regulated genes. Curated pathway analysis revealed that
   “allograft rejection” pathway was enriched. The associated
   eQTL-regulated genes were indicated by red boxes, and potential drugs
   to target related pathway were highlighted in green.

4. Discussion

   In this study, we analyzed the genetic relationships between MS and
   common eye disorders, such as optic neuritis, cataract, glaucoma,
   macular degeneration and refractive disorders, using MR analysis. Our
   data have revealed that (1) MS is genetically associated with optic
   neuritis pathology, but there is no evidence of a reciprocal
   relationship; (2) MS is causally related to reducing cataract and
   macular degeneration; (3) MS might influence the pathology of eye
   disorders through regulating the lymphocyte activation, proliferation
   and differentiation; (4) Drugs used for allograft rejection exhibit
   potential therapeutic effects on the MS coexisting with eye disorders.
   Our study provides genetic evidence regarding the incidence of and
   potential therapeutic targets for MS in conjunction with eye disorders.

4.1. MS and eye disorders

   It has been widely accepted in clinics that optic neuritis, as an
   inflammatory demyelinating disorder of the optic nerve, can occur as a
   symptom of MS ([89]13, [90]14). According to the North American Optic
   Neuritis Treatment Trial (ONTT), the risk of developing MS after 15
   years was higher in patients who had abnormal brain scans than in those
   who had normal brain scans ([91]15). However, patients with optic
   neuritis who lived in areas with low MS prevalence had lower rates of
   brain MRI abnormalities ([92]14), indicating that optic neuritis might
   not be an independent risk factor for MS. It was supported that optic
   neuritis, in isolation, does not inherently involve the widespread
   autoimmune response observed in MS ([93]16). Consistently, our
   investigation revealed a relatively limited causal relationship between
   optic neuritis and MS. However, the MR results showed that MS was a
   causal risk factor for optic neuritis. MS is a chronic inflammatory,
   demyelinating and neurodegenerative disease that affects the central
   nervous system (CNS). The demyelinating lesions in the optic nerve can
   impair the optic nerve in MS patients, leading to optic neuritis and
   even visual dysfunction ([94]1).

   We also identified that MS was causally related to reducing the
   incidence of other cataract and wet age-related macular degeneration.
   However, a population-based cohort study conducted using the UK General
   Practice Research Database from 1987 to 2009 reported that MS is not
   associated with the overall incidence of cataract or macular
   degeneration, but the risk of cataract is higher in MS patients younger
   than 50 years old, especially for men ([95]17). MS and cataract share
   some mechanisms of disease pathological processes ([96]17). The
   activity of proteolytic enzyme is markedly elevated in relapsing MS
   patients ([97]18), and over-activation of proteolytic enzyme is thought
   to cause the degradation of lens proteins and changes in the
   cytoskeletal architecture, resulting in the formation of cataract
   ([98]19). Additionally, prolonged exposure to corticosteroids,
   anticonvulsant and statin, commonly used in the treatment of MS, have
   been postulated as a potential risk factor for cataract development
   ([99]20–[100]23).

   Some studies suggest an increased prevalence of age-related macular
   degeneration among patients with MS ([101]24, [102]25). There is a
   potential interplay between the neurodegenerative processes occurring
   in MS and the retinal degeneration seen in age-related macular
   degeneration. Both conditions involve neuronal damage and degeneration,
   albeit in different anatomical locations ([103]26). We speculate that
   the shared disease pathological processes, medication use and
   aging-associated diseases in patients might induce the discordance
   between clinical findings and our MR results.

4.2. MS and adaptive immune response

   We have identified the causal relationship between MS and optic
   neuritis, cataract and macular degeneration using MR analysis. However,
   it still remains unclear how MS might influence the onset of eye
   disorders. After MS occurs, the adaptive immune system, especially T
   cells (Such as Th1 and Th17 cell) and B cells, becomes hyperactive and
   produces high levels of pro-inflammatory cytokines ([104]27). In the
   present study, the enrichment analysis of targeted eQTL of extracted
   SNPs, revealed that the adaptive immune response, including T
   lymphocyte proliferation, activation, antigen processing and
   cytokine-mediated signaling was involved in the pathogenesis of MS
   coexisting with eye disorders. It is consistent with a previous report
   that adaptive immune cells infiltrated into the parenchyma,
   perivascular spaces, and meninges in optic nerve tissues in MS
   patients, even without visual disturbances ([105]28). In the
   experimental autoimmune encephalomyelitis (EAE) model, the
   transcription of Th1 and Th17-specific cytokine genes interferon gamma
   (Ifng) and interleukin 17A (Il17a) was upregulated in optic nerve, and
   decreasing Ifng expression had a protective effect on optic neuritis in
   EAE mice ([106]29). Additionally, aqueous humor IFNγ levels also
   increased in patients with either aged cataract or age-related macular
   degeneration ([107]30, [108]31), indicating the involvement of Th cells
   in the pathogenesis of these two diseases. The above evidences support
   our findings that lymphocyte-induced immune response might play a role
   in MS coexisting with the eye disorders.

   We also annotated the nearest genes (total 80 genes) of the causal SNPs
   from four MS datasets. There are thirty-four genes that were directly
   associated with the immune response and inflammation, and thirty genes
   were not. Other sixteen genes are pseudogenes. The detailed biological
   functions of all genes were listed in the [109]Supplementary Table 2.
   Then we conducted functional enrichment analysis of the genes that were
   indirectly associated with inflammatory response, and found that the
   main enriched biological processes included asymmetric cell division,
   inositol lipid-mediated signaling, response to epidermal growth factor,
   process utilizing autophagic mechanism, and phagocytosis
   ([110]Supplementary Table 3). It is worth noting that the epidermal
   growth factor signaling pathway play important role in eye diseases
   through enhancing the anti-apoptotic and anti-inflammatory effects of
   corneal epithelial cells ([111]32). Anticancer drugs which target to
   epidermal growth factor receptor can inhibit corneal reparation and
   induce ocular toxicity ([112]33), indicating that MS may also affect
   the occurrence of optic neuritis through non-inflammatory ways.

   In the thirty-four inflammation associated genes, we identified several
   HLA family genes involved in the MS coexisting with eye disorders. HLA
   family genes are associated with adaptive immune response and have been
   reported to be related to the genetic susceptibility within the immune
   system, leading to MS ([113]34, [114]35). Treatments based on genetic
   information have proven to be a promising path for drug repurposing and
   drug development. By using Cytoscape, we enriched the drugs that used
   for allograft rejection, suggesting their potential as therapeutic
   options for MS patients with eye disorders.

   There are ten allograft rejection drugs predicted in our study. Among
   these drugs, some have already been the first-line treatment for MS
   (methylprednisolone and prednisone) ([115]36). Additionally, others,
   including eculizumab, sirolimus, ciclosporine, tacrolimus, fenofibrate
   and daclizumab, have also been tested in clinical trials for MS.
   Eculizumab, a humanized monoclonal antibody approved for the treatment
   of aquaporin-4-positive neuromyelitis optica spectrum disorders, has
   undergone investigation for its safety and efficacy in MS ([116]37).
   However, its efficacy as a disease-modifying therapy for MS patients
   still remains inconclusive due to study limitations ([117]37, [118]38).
   Sirolimus, as a type of rapamycin inhibiting the mTOR kinase, is
   currently approved for lymphangioleiomyomatosis therapy ([119]39).
   Recent studies have explored its potential in treating MS, with
   promising results suggesting its consideration as a therapeutic option
   with minimal side effects ([120]40, [121]41). Ciclosporine and
   tacrolimus, inhibitors of calcineurin (an enzyme involved in the
   transcription of IL-2 and T-cell activation), have exhibited potential
   therapeutic effects for MS ([122]42, [123]43). The PPAR alpha agonists
   such as fenofibrate, have been observed to ameliorate the clinical
   symptoms of MS through inhibiting interleukin-4 and interferon-γ
   secretion ([124]44). Daclizumab, a monoclonal antibody targeting the
   CD25 subunit of the interleukin-2 receptor, has shown efficacy in
   slowing the inflammatory process of MS ([125]45). Moreover, Afief and
   colleagues have reported that belatacept acted as a promising MS drug
   candidate using integration of genomic variants and bioinformatic-based
   approach ([126]46). These studies support our prediction that allograft
   rejection drugs are potential therapeutics for MS, even MS coexisting
   with eye disorders.

   The present study has several limitations. First, MS is more prevalent
   in women than in men, with a female-to-male ratio of approximately 3:1
   ([127]47, [128]48). However, the GWAS summary statistic datasets used
   in this study include both sexes, which could not distinguish the
   differences between women and men. The different gender ratio in the
   datasets could affect the MR results and introduce bias in the
   evaluation. Second, the prevalence of MS varies across different
   regions and ethnicities ([129]49). We only used the datasets from
   European populations, which may limit generalizability of our findings
   to other populations. Third, the drug prediction was based on the
   causal SNPs, which should be interpreted with caution, as they may not
   reflect the true causal variants or genes. Additional limitations in
   our study are as followings. The sample size of the five eye disorders
   traits is relatively small, so we have adopted a relaxed p value of
   1E-6 to screen for more exposure IVs, but this could compromise the
   reliability of our analysis. We have focused on the exposure of MS
   traits, which have robust sample size to perform the analysis. In
   addition, performing drug prediction by only using causal SNPs would
   leave out some important therapies, such as the widely used
   dihydroorotate-dehydrogenase inhibitor (teriflunomide).

   In conclusion, we analyzed the causal associations between MS and
   common eye disorders using MR to explore the mechanisms of MS
   coexisting with eye disorders. Our results reveal that MS is casually
   associated with a higher risk of optic neuritis and a lower risk of
   non-senile cataract and age-related macular degeneration.
   Bioinformatics analysis revealed that lymphocyte proliferation,
   activation and antigen processing are involved in the pathogenesis of
   MS coexisting with eye disorders. These findings provide a better
   understanding of the etiology of MS and potential therapeutic targets
   in MS patients with eye disorders.

Data availability statement

   The original contributions presented in the study are included in the
   article/[130]Supplementary Material. Further inquiries can be directed
   to the corresponding authors.

Author contributions

   XQ: Data curation, Formal analysis, Funding acquisition, Investigation,
   Writing – review & editing. MNH: Funding acquisition, Supervision,
   Writing – review & editing. SP: Conceptualization, Funding acquisition,
   Validation, Writing – original draft.

Acknowledgments