Role of the HLA System in the Association Between Multiple Sclerosis and Infectious Mononucleosis

Objective  To determine whether multiple sclerosis (MS) and infectious mononucleosis (IM) share common HLA associations.

Design  A prospective cohort study was conducted from October 1, 1999, through September 30, 2003.

Setting  University of Edinburgh Richard Verney Health Centre, Edinburgh, Scotland.

Patients  Participants included 179 individuals who underwent asymptomatic Epstein-Barr virus seroconversion and 175 patients who developed IM.

Intervention  Genotyping for 5 classical HLA loci (HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1).

Main Outcome Measure  Diagnosis of IM and allele frequency.

Results  Allelic analysis showed that HLA-DRB1*01:01 was significantly associated with the development of IM (odds ratio, 3.2; P < .001). Patients with IM and HLA-DRB1*01:01 had a lower Epstein-Barr virus viral load compared with those without the allele (median, 783 vs 7366 copies/10⁶ peripheral blood mononuclear cells; P = .03).

Conclusion  HLA-DRB1*01:01 is protective against developing MS; thus, a common genetic basis between IM and MS is not supported.

Multiple sclerosis (MS) is the most common disease of the central nervous system to cause permanent disability in young adults.¹ Based on strong circumstantial evidence, MS is considered to be an organ-specific autoimmune disorder,^1,2 but much remains to be understood about the initiation of the disease. It appears unlikely that MS results from a single causative event; rather, it is more probable that the disease develops in a genetically susceptible population as a result of environmental exposures.

There is a growing body of evidence in support of a role for Epstein-Barr virus (EBV) in determining susceptibility to MS. A key observation is the finding that individuals with a history of infectious mononucleosis (IM) have an increased risk of developing MS. A systematic review and meta-analysis³ reported a combined relative risk of MS after IM of 2.3 (95% confidence interval, 1.7-3.0). This risk has been confirmed in large population-based studies.^4,5 The mechanism by which IM is associated with MS is poorly understood. A suggested mechanism is common genetic factors that predispose to both diseases. A recent prospective study⁶ comparing individuals who developed asymptomatic EBV infection with those who developed IM demonstrated association of IM with microsatellite markers in the HLA class I region. However, that study made no attempt to correlate risk with specific HLA genes. Given the strong association of MS with the major histocompatibility complex, we used this prospective IM cohort^6-8 to investigate the effects of the HLA class I and II genes on the risk of developing IM.

Participants were recruited through the University of Edinburgh Richard Verney Health Centre. The study was approved by the Lothian Research Ethics Committee, and all participants provided written consent. The procedures followed were in accordance with the Helsinki Declaration of 1975, as revised in 1983.

Details on enrollment and serotyping have been published.^7,8 In brief, all students enrolling at the university health center from October 1, 1999, through September 30, 2000, were approached to take part in the study and, on recruitment, to provide a blood sample for EBV serologic determination. Those with EBV-seronegative results were then monitored for the development of IM while at the university (approximately 4 years). In addition, students who developed IM but were not enrolled in the original study were referred by the university health center and asked to participate. Upon diagnosis of IM (median, 10 days after onset of symptoms; range, 3-42 days), patients were examined and asked to provide a blood sample. Students with seronegative EBV test results on enrollment who did not report symptoms of IM were asked to return for further testing upon exit from the university. Students with seropositive EBV results were regarded as asymptomatic seroconverters and formed the EBV-seropositive group. Data on 179 individuals identified as asymptomatic seroconverters (controls) and 175 patients with IM were available for analysis. There was no variation in ethnicity between groups. Most control participants (78%) and all those diagnosed as having IM carried the type 1 EBV virus strain; therefore, infection with different EBV types was assumed to be minimal.⁹

Peripheral blood mononuclear cells were separated from whole blood by routine density gradient centrifugation, washed, and counted. DNA was extracted from approximately 5 ×10⁶ cells (Easy-DNA Kit; Invitrogen, Paisley, Scotland), according to the manufacturer's instructions, and stored at −70°C.

High-resolution sequence–based typing of HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1 was performed as described¹⁰ on 179 asymptomatic seroconverters and 144 patients with IM. Data on an additional 31 patients with IM who previously underwent genotyping for HLA-DRB1 alone were added to the analysis.

The EBV-DNA load was determined by quantitative competitive polymerase chain reaction as described by Stevens et al.¹¹ The amplification reaction contained 50mM potassium chloride, 1.5mM magnesium chloride, 10mM TRIS-hydrochloride buffer (pH, 8.5), 200μM deoxynucleoside triphosphates, 25 pmol of each primer (5′ primer labeled with biotin), and 1 U of Taq polymerase. Cycling conditions were 4 minutes at 95°C; 40 cycles at 95°C, 55°C, and 72°C for 1 minute each; and 3 minutes at 72°C. Products were captured on a streptavidin-coated plate and probed with digoxigenin-labeled wild-type and internal standard probes. Optical density was measured and used to calculate the number of copies.

A Fisher exact test was used to compare allele frequencies between cohorts at each of the 5 loci. All P values are presented as uncorrected for multiple testing. Viral copy number was compared across groups using the Mann-Whitney test. A P value <.05 was considered significant. To correct for multiple testing, permutation tests were performed (10⁶ permutations).

Haplotype analyses were performed (R package haplo.stats, Version 1.4.4; http://mayoresearch.mayo.edu/schaid_lab/software.cfm); cohorts were treated as either case or control. This approach, achieved using the haplo.cc function, adopts an additive-effect linear modeling method to assess association between the binary trait and the haplotypes. Various case-control groupings were assessed, with 2000 control simulations done for each. The minimum haplotype frequency for inclusion was set at 0.005, and the most frequent haplotype was treated as the baseline. Because sample numbers were small, some key haplotypes may not have been modeled.

Initial allele and haplotype frequencies were compared between the IM cohort (n = 144) and EBV-positive group (n = 179) for individuals who underwent genotyping for all 5 HLA loci.

Six different alleles were identified with a frequency difference yielding a nominal P value of <.05 (Table). HLA-DRB1*01:01 was the most significant allele in the IM vs EBV-positive cohort comparison, being significantly more frequent in the IM group (odds ratio, 2.49; 95% confidence interval, 1.29-5.01; P = .003). The frequency of only HLA-DRB1*01:01 remained significant after permutation testing (corrected P = .048).

In the haplotype analysis, HLA-A*03:01–HLA-B*35:01–HLA-C*04:01–HLA-DRB1*01:01–HLA-DQB1*05:01 was significantly associated with IM (P = .03), being overrepresented in the IM cohort. All other haplotypes were nonsignificant in all comparisons (P > .05).

When we included the data from 31 patients who underwent genotyping for HLA-DRB1 alone, the association of IM with HLA-DRB1*01:01 became stronger (IM vs EBV-positive comparison, P < .001; odds ratio, 3.2; 95% confidence interval, 1.72-6.27); the association with all other HLA-DRB1 alleles became nonsignificant (P > .05). There were no significant differences in HLA-DRB1 allele frequencies between the first (n =144) and second (n = 31) sets of patients with IM.

Viral load data were available for a small subset of patients with IM (n = 32). Comparison of EBV viral copy number demonstrated a significant difference (P = .03) between patients with IM carrying the HLA-DRB1*01:01 allele (median, 783 copies/10⁶ peripheral blood mononuclear cells; range, 26-33 783) and those without the allele (median, 7366 copies/10⁶ peripheral blood mononuclear cells; range, 483-48 283) (Figure).

We observed that HLA-DRB1*01:01 is the allele most significantly associated with IM, being overrepresented in individuals who develop this syndrome compared with individuals with asymptomatic EBV infection. The HLA-DRB1*01 allele is protective in MS,¹² and we found no association with the strong MS risk allele HLA-DRB1*15. Therefore, our study rules out a common genetic association between MS and IM.

The association of HLA class II with IM is functionally plausible. Epstein-Barr virus preferentially infects B lymphocytes through the binding of the major viral envelope glycoprotein Gp350 to the CD21 receptor on the surface of B cells and through the binding of a second glycoprotein, Gp42, to HLA class II molecules.¹³ Therefore, it may be that HLA-DRB1*01:01 is somehow more permissive for EBV entry, leading to enhanced proliferation during the initial stage of infection and effective inducement of the immune response. Interestingly, HLA-DRB1*01:01 has also been shown^14,15 to be a potent stimulator of CD8+ and CD4+ T-cell responses after infection with hepatitis B and mumps vaccination—both in terms of magnitude and quality of function. Furthermore, reports¹⁶ investigating the role of HLA class II molecules in the control of hepatitis C virus infection show that HLA-DRB1*01:01 possession correlates with good control and clearance of the virus, suggesting that the immune response is more effective in these individuals. Similarly, in the case of primary EBV infection, it may be that the HLA-DRB1*01:01 allele strongly binds an immunodominant EBV epitope or a selection of epitopes, leading to a more robust helper T-cell function and the subsequent heightened immune response that characterizes IM. This may in turn control initial virus replication more effectively and minimize the duration of the infection. Our viral load estimations on samples taken at similar time points (HLA-DRB1*01:01 positive: median, 11.5 days after onset of symptoms [range, 4-42 days]; HLA-DRB1*01:01 negative: median, 10 days [range, 3-28]) show a significant reduction in EBV copy number in individuals with HLA-DRB1*01:01, suggesting that immune control of virus replication is more effecive in these individuals. However, the remit of this study did not allow for a detailed analysis of the immune response to confirm this hypothesis. In addition, because of the low allele frequency of HLA-DRB1*01:01, the attributable risk may be low. It may also be that HLA-DRB1*01:01 leads to more severe IM rather than susceptibility to IM.

A previous investigation⁶ highlighted the role of microsatellites in the HLA class I region as being associated with IM; however, the present study suggests that the HLA class II region is the principal association, as HLA-DRB1*01:01 is the most significantly associated allele in the analysis. Nevertheless, the haplotype analysis highlights the fact that HLA-A*03:01–HLA-C*04:01–HLA-DRB1*01:01 is the haplotype associated with IM, which may reflect either linkage disequilibrium or epistatic effects to increase the risk of IM.

The cohorts that we studied were perhaps underpowered to fully disentangle the effects of HLA class I and II genes as well as to detect loci exerting a small effect on IM risk, and population stratification could be an issue despite our efforts to match cohorts for ethnicity. Replication of the results obtained is therefore key to understanding in detail the genetic basis of IM. However, a common genetic association between MS and IM would mean that HLA-DRB1*15 would exert effects of similar magnitude in both diseases, which we can exclude. These data suggest that EBV plays a causal role in MS pathogenesis. Further work is warranted to understand the mechanism by which EBV influences MS risk.

Correspondence: Karen A McAulay, PhD, Centre for Virus Research, Institute of Infection, Immunity, and Inflammation, University of Glasgow, Glasgow G61 1QH, Scotland (karen.mcaulay@glasgow.ac.uk).

Accepted for Publication: December 21, 2010.

Author Contributions:Study concept and design: Meier, Giovannoni, Crawford, and McAulay. Acquisition of data: Ramagopalan, Conacher, Giovannoni, Crawford, and McAulay. Analysis and interpretation of data: Ramagopalan, Meier, Conacher, Ebers, Giovannoni, and McAulay. Drafting of the manuscript: Ramagopalan Crawford, and McAulay. Critical revision of the manuscript for important intellectual content: Ramagopalan, Meier, Conacher, Ebers, Giovannoni, Crawford, and McAulay. Statistical analysis: Ramagopalan. Obtained funding: Giovannoni and Crawford. Administrative, technical, and material support: Conacher, Giovannoni, Crawford, and McAulay. Study supervision: Meier, Ebers, Giovannoni, Crawford, and McAulay.

Financial Disclosure: None reported.

Funding/Support: Funding for this study was provided by the Medical Research Council grant G0801976.

Role of the Sponsor: The study sponsor had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: The authors thank many colleagues and collaborators at Barts and The London School of Medicine, the University of Oxford, and the Edinburgh University Health Service and its participants. The authors also acknowledge the Histocompatibility and Immunogenetics Laboratory, Royal Infirmary of Edinburgh, for typing of some IM samples.