Mendelian randomization analysis suggests no associations of human herpes viruses with amyotrophic lateral sclerosis

Background The causal associations between infections with human herpes viruses (HHVs) and amyotrophic lateral sclerosis (ALS) has been disputed. This study investigated the causal associations between herpes simplex virus (HSV), varicella-zoster virus (VZV), Epstein–Barr virus (EBV), cytomegalovirus (CMV), HHV-6, and HHV-7 infections and ALS through a bidirectional Mendelian randomization (MR) method. Methods The genome-wide association studies (GWAS) database were analyzed by inverse variance weighted (IVW), MR-Egger, weighted median, simple mode, and weighted mode methods. MR-Egger intercept test, MR-PRESSO test, Cochran’s Q test, funnel plots, and leaveone-out analysis were used to verify the validity and robustness of the MR results. Results In the forward MR analysis of the IVW, genetically predicted HSV infections [odds ratio (OR) = 0.9917; 95% confidence interval (CI): 0.9685–1.0154; p = 0.4886], HSV keratitis and keratoconjunctivitis (OR = 0.9897; 95% CI: 0.9739–1.0059; p = 0.2107), anogenital HSV infection (OR = 1.0062; 95% CI: 0.9826–1.0304; p = 0.6081), VZV IgG (OR = 1.0003; 95% CI: 0.9849–1.0160; p = 0.9659), EBV IgG (OR = 0.9509; 95% CI: 0.8879–1.0183; p = 0.1497), CMV (OR = 0.9481; 95% CI: 0.8680–1.0357; p = 0.2374), HHV-6 IgG (OR = 0.9884; 95% CI: 0.9486–1.0298; p = 0.5765) and HHV-7 IgG (OR = 0.9991; 95% CI: 0.9693–1.0299; p = 0.9557) were not causally associated with ALS. The reverse MR analysis of the IVW revealed comparable findings, indicating no link between HHVs infections and ALS. The reliability and validity of the findings were verified by the sensitivity analysis. Conclusion According to the MR study, there is no evidence of causal associations between genetically predicted HHVs (HSV, VZV, EBV, CMV, HHV-6, and HHV-7) and ALS.


Introduction
Amyotrophic lateral sclerosis is a rapidly progressive and fatal neuronal disease characterized by progressive degeneration of motor neurons in the brain and spinal cord, ultimately leading to almost total skeletal muscle paralysis (van Es et al., 2017).Patients with ALS often present Zheng et al. 10.3389/fnins.2023.1299122Frontiers in Neuroscience 02 frontiersin.orgwith progressive weakness and atrophy of the extremities, gradually leading to an inability to walk, talk, swallow, and breathe (Taylor et al., 2016).ALS is a rare disease with an incidence of 3.1 per 100,000 person-years (Vasta et al., 2022), with familial ALS (fALS) accounting for 10% of cases and sporadic ALS (sALS) accounting for 90% of cases (Pham et al., 2020) and the average survival after diagnosis is in the range of 3-5 years (Brown and Al-Chalabi, 2017).The number of ALS cases is expected to increase by 69% from 2015 to 2040 due to global aging (Arthur et al., 2016).However, little is known despite the time and money spent investigating ALS's pathogenic mechanisms.ALS is believed to be caused by genetic and environmental interactions (Celeste and Miller, 2018;Schram et al., 2020;Julian et al., 2021;Motataianu et al., 2022;Goutman et al., 2023), with some investigators suggesting that viruses are an important environmental factor in ALS (Celeste and Miller, 2018;Castanedo-Vazquez et al., 2019).
Studies have shown that enteroviruses and herpesviruses are the two most common viruses infecting hospitalized patients' central nervous system (CNS) (Roshdy et al., 2023).The α/β/γ-herpesviruses are neurotoxic and neurotropic viruses and are often considered important risk factors for neurodegenerative disease (NDD) (Osorio et al., 2022).However, the relationship between HHVs and ALS is unknown.A study reported that herpesvirus infection and ALS can flare up simultaneously.Nevertheless, there is no way to determine if it is accidental or intentional (Ferri-de-Barros and Moreira, 2010).Some investigators have suggested that HSV-1 can latently infect the trigeminal ganglion and may be a causative factor in ALS (Feng et al., 2022).In addition, the pathogenic risk of ALS was slightly associated with HHV-6 seropositivity in a casecontrol study (Cermelli et al., 2003).In contrast, no significant correlation was found between ALS and IgG antibodies to HSV and CMV in an immunological evaluation of early ALS (Provinciali et al., 1988).These studies with conflicting conclusions may be due to methodological flaws, including confounders and reverse causality in observational studies.We have no method of determining the association, let alone the causality, between HHVs and ALS.
Random control trials (RCTs) are a type of experimental research methodology that evaluates the effect of a causative factor or a treatment regimen on a disease by randomizing study subjects into groups, implementing different interventions for different groups, and finally comparing the results.RCTs are the gold standard for clinical diagnosis and can determine the causal association between exposure and outcome.There are no reported RCTs on the association between HHVs and ALS, mainly due to the strict constraints of the design process and medical ethics and few reports of observational studies and their mixed conclusions, which makes it important to carry out MR analyses in this context.MR is an epidemiological investigation method based on instrumental variables (IVs) to analyze summary-level data from GWAS, which can greatly reduce confounding bias and consistently and reliably infer causality between exposures and outcomes due to the stochastic nature of the genetic variants and the fact that alleles are not affected by disease (Emdin et al., 2017).Two-sample Mendelian randomization (TSMR) refers to using genetic variants as IVs in both exposure and outcome samples to investigate the effects of modifiable risk factors for disease.Although lower than RCTs, the strength of evidence from MR analyses is stronger than observational studies (Davies et al., 2018).Particularly in rare diseases, MR can achieve more reliable results by analyzing and evaluating much larger sample sizes than in conventional clinical trials.Therefore, our study investigated whether there is a causal association between HHVs and ALS using a bidirectional TSMR method based on GWAS data.

Study design
This study strengthened epidemiological observational studies using Mendelian randomization (STROBE-MR) (Skrivankova et al., 2021).All data were obtained from the publicly available GWAS database without re-ethical approval.In MR, SNPs as IVs must fulfill the following three assumptions.(1) The relevance assumption: IVs are closely related to exposure; (2) the independence assumption: the IVs are not associated with the potential confounders; (3) the exclusion restriction assumption: IVs affect outcomes only through the exposure pathway (no directional pleiotropy) (Supplementary Figure S1).

GWAS data sources
The summary-level statistics for all cases and controls in this study are of European ancestry, and the study subjects were residents recruited from multiple research centers in Europe to minimize bias due to race-related confounding factors.GWAS data for HHV-8 of suitable European ancestry could not be found and were therefore not analyzed.

Outcome
The ALS dataset for GCST90027164 (27,205 cases,110,881 controls,and 10,461,755 SNPs) was searched in the GWAS database.After cleaning and conversion, we saved the data downloaded from the FinnGen database in GWAS database format.After comparing the sources of participants in the eight datasets for HHVs with the one dataset for ALS, we consider the samples for HHVs and ALS to be independent.We list these data in Supplementary Table S1.

Selection of instrumental variables
A critical step in MR analysis is to obtain valid IVs.We extracted SNPs (p < 5 × 10 −5 ) with significant correlation with HSV, VZV, EBV, CMV, HHV-6, and HHV-7 from the eight exposure datasets.Subsequently, we performed linkage disequilibrium (LD) analysis on the obtained SNPs with "r 2 < 0.001, 10,000 = kb" by using the "clump_ data" function to exclude the mutual linkage SNPs and to discard non-biallelic SNPs.We used the F-statistic to assess the strength of the association between the selected IVs and exposure.F-statistic is calculated as F = (β/se(β)) 2 (Zhang et al., 2023) when F > 10 indicates that IVs are strong instrumental variables, which avoids the bias caused by weak IVs.In addition, the summary set may produce errors if the effect alleles for the SNP effects are different in the GWAS data for exposure and outcome.Therefore, we used the "harmonise_data" function to test the causal direction of the screened SNPs in exposure and outcome, excluded palindromic alleles, and selected the SNPs with "TRUE" results for MR analysis.

TSMR analysis
The data in this study were analyzed based on the "TwoSampleMR" package of R version 4.2.3 software.Analyses included Forward MR with HHVs infection or IgG as the exposure and ALS as the outcome and reverse MR with ALS as the exposure and HHVs infection or IgG as the outcome.We chose MR Egger, weighted median, IVW, simple mode, and weighted mode methods to calculate the causal relationship between exposure and outcome, and IVW was the most valid and reliable of these methods.We then perform sensitivity analyses.Cochran's Q-statistic was used to test for heterogeneity (p < 0.05) between SNPs in MR-Egger and IVW analyses to assess the robustness of IVs (Bowden et al., 2019).Heterogeneity was additionally visualized by constructing a funnel plot of the IVs.MR-PRESSO can detect outliers that may bias the results and give a causal change in exposure and outcome after removing the outlier (Verbanck et al., 2018).Therefore, when MR-PRESSO detects an outlier, we exclude the SNPs and re-perform the MR analysis and evaluation.Pleiotropy refers to the fact that some IVs affect outcomes through pathways other than exposure, which would seriously affect the reliability of the causal association between exposure and outcome (Davey Smith and Hemani, 2014).We used the MR-Egger intercept for bias detection and effect estimation.When the "MR_pleiotropy_test" function calculates p < 0.05, it means that there is directional pleiotropy (Bowden et al., 2015).Leave-one-out analysis estimates the effect of the remaining SNPs on the outcome by sequentially removing individual SNPs and then performing the IVW analysis again, which can determine whether any single SNPs drive causality.

Discussion
Neurodegenerative diseases can damage and degenerate neurons in the CNS and peripheral nervous system, resulting in severe loss of memory, behavior, and sensory and motor functions due to the inability of the neurons to renew and regenerate.Classic NDDs include Parkinson's disease, Alzheimer's disease, and ALS, which impose an enormous global burden (Wilson et al., 2023).Although the pathogenic mechanism of ALS is unclear, it is believed that it may be due to oxidative stress, apoptosis, mitochondrial dysfunction, axonal degeneration, neuroinflammation, and viruses (Sever et al., 2022).
Among HHVs, the most well-known HSV can infect neurons and reach the CNS through retrograde axonal transport, ultimately leading to diseases such as encephalomyelitis, and several studies have shown that HSV is closely related to NDDs (Carneiro et al., 2022).Primary infection with VZV causes varicella and can latently infect neurons.When the body is immunocompromised or aged, VZV is reactivated and causes a zoster.In addition, VZV can cause neurological syndromes such as myelitis and segmental motor paralysis (Gershon et al., 2015).EBV is most commonly associated with infectious mononucleosis, while primary or latent infections are mostly associated with neurological disorders, such as encephalomyelitis (Andersen et al., 2023).CMV can exhibit tropism for neural stem cells and cause multiple spinal cord radiculitis (Kleinschmidt-DeMasters and Gilden, 2001; Kamte et al., 2021).The viral load of HHV-6 is associated with increased central nervous system demyelination (Lucas et al., 2023).The damage to the CNS system by HHV-7 has also been reported (Li et al., 2022).
However, do these HHVs increase the risk of ALS? Earlier studies reported that chronic viral infections play an important role in the pathogenesis of ALS, and antibody titers to HSV-1 were significantly increased in the sera of ALS patients (Irkeç et al., 1989).In contrast, studies have reported that significant elevations of HSV-1, HSV-2, and VZV antibodies were not detected in the sera of ALS patients (Catalano, 1972).In a mouse model of latent HSV-2 infection, it was found that HSV-2-induced spinal cord inflammation, although similar to that of ALS patients, was insufficient to induce the characteristic changes in the pathology of ALS (Cabrera et al., 2020).In addition, it has been reported that HSV-1, EBV, CMV, and HHV-6 can cause motor neuron degeneration by activating endogenous retroviruses, thereby inducing the expression of the envelope glycoprotein HERV-K in ALS (Medina et al., 2017;Mayer et al., 2018).However, a subsequent study questioned this view due to their failure to detect highly expressed HERV-K RNA in ALS (Garson et al., 2019).Therefore, a causal association between HHVs and ALS cannot be stated based on the available evidence, and our systematic MR analyses can contribute to related studies.Thus, studies on the relationship between human herpesviruses and ALS are conflicting, and there are few reports of a causal association between them.We concluded that there is no evidence to support a causal association between HHVs (HSV, VZV, EBV, CMV, HHV-6, and HHV-7) and ALS using bidirectional TSMR based on a sizable sample of GWAS data, implying that observational studies in which HHVs and ALS are associated may be due to confounding factors such as environment or shared genetic structure.
Our bidirectional TSMR study focused on the causal relationship between HHVs and ALS for the first time.This study has several strengths.Firstly, we are not limited to studying HSV and ALS but have expanded to study the causal relationship between multiple HHVs and ALS.Secondly, a bidirectional TSMR analysis reduces bias from confounding factors and excludes the effects caused by reverse causality.Of course, this study has some limitations.First, the significance threshold was relaxed from 5 × 10 −8 to 5 × 10 −5 because the number of IVs was so small.Distortion caused by weak instruments is possible, even though the resulting IVs were all defined as strong instrumental variables after calculating the F statistic (F > 10).Second, although sensitivity analyses of MR indicate robustness among SNPs, there is still the possibility of residual heterogeneity.Finally, we only analyzed GWAS data using European ancestry, and results should be interpreted with caution when applied to other populations.Forest plots of causal effect estimates in the forward MR analysis.SNP, single-nucleotide polymorphism; ALS, amyotrophic lateral sclerosis; IVW, inverse variance weighted; OR, odds ratio; 95% CI, 95% confidence interval.

Conclusion
We have shown no causal association between genetically predicted human herpes viruses (HSV, VZV, EBV, CMV, HHV-6, and HHV-7) and ALS based on Mendelian randomization analysis of currently relevant GWAS data.The associations observed in epidemiological studies may be partly attributable to shared genetic structure or environmental confounders, and we could devote more time and money to studies of other environmental factors associated with ALS and genetic structure.

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.

TABLE 1
The causal effect of human herpes viruses (HHVs) and amyotrophic lateral sclerosis (ALS) by two-sample Mendelian Randomization (TSMR) and the sensitivity analysis results.

TABLE 2
The causal effect of amyotrophic lateral sclerosis (ALS) and human herpes viruses (HHVs) by two-sample Mendelian Randomization (TSMR) and the sensitivity analysis results.