Advertisement
Research Article

Association of Multiple Sclerosis Susceptibility Variants and Early Attack Location in the CNS

  • Ellen M. Mowry mail,

    emowry1@jhmi.edu

    Affiliation: Multiple Sclerosis Center, Department of Neurology, Johns Hopkins University, Baltimore, Maryland, United States of America

    X
  • Robert F. Carey,

    Affiliation: Multiple Sclerosis Center, Department of Neurology, University of California San Francisco, San Francisco, California, United States of America

    X
  • Maria R. Blasco,

    Affiliation: Department of Neurology, Hospital Universitario Puerta de Hierro, Madrid, Spain

    X
  • Jean Pelletier,

    Affiliation: Pole de Neurosciences Cliniques, Service de Neurologie, Centre de Résonance Magnétique Biologique et Médicale, Centre Hospitalier Universitaire Timone, Aix Marseille Université, Marseille, France

    X
  • Pierre Duquette,

    Affiliation: Multiple Sclerosis Clinic, Centre Hospitalier de L'Universite de Montreal, Montreal, Canada

    X
  • Pablo Villoslada,

    Affiliation: Center of Neuroimmunology, Institute of Biomedical Research August Pi Sunyer-Hospital Clinic, Barcelona, Spain

    X
  • Irina Malikova,

    Affiliation: Pole de Neurosciences Cliniques, Service de Neurologie, Centre de Résonance Magnétique Biologique et Médicale, Centre Hospitalier Universitaire Timone, Aix Marseille Université, Marseille, France

    X
  • Elaine Roger,

    Affiliation: Multiple Sclerosis Clinic, Centre Hospitalier de L'Universite de Montreal, Montreal, Canada

    X
  • R. Phillip Kinkel,

    Affiliation: Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, United States of America

    X
  • Jamie McDonald,

    Affiliation: Multiple Sclerosis Center, Department of Neurology, University of California San Francisco, San Francisco, California, United States of America

    X
  • Peter Bacchetti,

    Affiliation: Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, California, United States of America

    X
  • Emmanuelle Waubant

    Affiliation: Multiple Sclerosis Center, Department of Neurology, University of California San Francisco, San Francisco, California, United States of America

    X
  • Published: October 10, 2013
  • DOI: 10.1371/journal.pone.0075565

Abstract

Objective

The anatomic location of subsequent relapses in early multiple sclerosis (MS) appears to be predicted by the first attack location. We sought to determine if genetic polymorphisms associated with MS susceptibility are associated with attack location.

Methods

17 genome-wide association study-identified MS susceptibility polymorphisms were genotyped in 503 white, non-Hispanic patients seen within a year of MS onset. Their association with the CNS location of the first two MS attacks was assessed in multivariate repeated measures analyses (generalized estimating equations with robust standard errors).

Results

The IL12A polymorphism was independently associated with increased odds of attacks involving the spinal cord (OR = 1.52, 95% CI 1.11, 2.07, p = 0.009), as was the IRF8 polymorphism (OR = 2.40, 95% CI [1.04, 5.50], p = 0.040). The IL7R polymorphism was associated with reduced odds of attacks involving the brainstem/cerebellum (OR = 0.46, 95% CI 0.22, 0.97, p = 0.041), as were the TNFRSF1A and IL12A polymorphisms. The CD6 polymorphism conferred reduced odds of optic neuritis as an attack location (OR = 0.69, 95% CI [0.49, 0.97], p = 0.034). Several other genes showed trends for association with attack location.

Conclusions

Some of the MS susceptibility genes may be associated with MS attack location. The IL12A polymorphism is of particular interest given that interferon beta therapy appears to influence IL12 levels. These findings may lead to improved understanding of MS pathogenesis and treatment.

Introduction

While the anatomic location of multiple sclerosis (MS) attacks differs from one person to the next, we have recently demonstrated that an individual patient's early attacks are likely to occur in the same location within the central nervous system (CNS) [1]. In other words, a person who has spinal cord involvement in the first attack is likely to have involvement thereof during the second attack, whereas a person with onset involving the optic nerve has greater odds of experiencing optic neuritis during the second attack. We hypothesized that genes known to be associated with MS susceptibility may also determine various features of its phenotype. In this investigation, we sought to determine if MS susceptibility genes are associated with the CNS location of early MS attacks.

Methods

Ethics statement

This study was approved by the University of California, San Francisco Committee on Human Research (CHR); each center contributed data from patients enrolled in Institutional Review Board-approved studies who had provided written informed consent.

Subjects and sites

This is a retrospective study of white patients with clinically isolated syndrome or relapsing-remitting MS. Race was restricted to reduce genetic heterogeneity. Subjects at five MS centers (UCSF, San Francisco, USA; MS Unit, Marseille, France; Centre Hospitalier de L'Université de Montreal, Montreal, Canada; University of Navarra, Spain; Hospital Puerta de Hierro Neuroimmunology Unit, Madrid, Spain); and from two clinical trials (Controlled High Risk Avonex Multiple Sclerosis Prevention Study in Ongoing Neurologic Surveillance [CHAMPIONS] and Optic Neuritis Treatment Trial [ONTT])[2], [3] who were seen within a year of their first attack and who had brain magnetic resonance imaging (MRI) within six months of symptom onset were included. At UCSF, three different cohorts were evaluated. The first was the UCSF MS Center database, in which data for all patients (children and adults) seen in the first year of MS onset are entered. Routine clinic visits for UCSF patients typically take place every 6 months. Subjects from two trials (the atorvastatin and riluzole trials) were also included [4], [5]. Atorvastatin participants were seen within 90 days of onset and were evaluated monthly for three months and then quarterly (total duration of follow-up was 18 months). For the riluzole study (which was incomplete when the data were frozen for this analysis), patients were enrolled in the first year of MS onset; they were evaluated monthly for six months and then quarterly for two to three years. In all three cohorts, additional visits were scheduled for suspected relapses. At Aix Marseille Université, records of patients seen within six months of disease onset who participated in a prospective longitudinal study of MS that began in 2000 were retrieved from the EDMUS database [6]. In this cohort, patients had clinic visits and MRIs every 3 months for a year, then every 6 months for 3 years, and annually thereafter. Patients at L'Université de Montreal, Canada were offered participation in an observational study from September, 2007 until March, 2009. Visits occurred at least once a year but were more frequent if the disease was active. Patients evaluated at the University of Navarra, Spain were enrolled at the time of MS onset in a prospective MS biomarker study (using the EDMUS database) that began in 2001 and were clinically evaluated every 3 to 6 months thereafter [7]. At the Universitario Puerta de Hierro, CIS patients evaluated within a year of symptom onset had a neurologic examination at three months and were followed biannually thereafter, with additional visits for suspected relapses. The Optic Neuritis Treatment Trial (ONTT) enrolled participants within 8 days of optic neuritis onset [8]. Because some had had prior symptoms, only those without symptoms consistent with a previous MS/CIS attack were included. The participants were re-evaluated on days 4, 15, and 30, weeks 7, 13, and 19, months 6 and 12, and then annually [3]. The Controlled High Risk Avonex Multiple Sclerosis Prevention Study in Ongoing Neurologic Surveillance (CHAMPIONS) study was an extension study (open-label) of Controlled High Risk Subjects Avonex Multiple Sclerosis Prevention Study (CHAMPS) [2], [9]. CHAMPS enrolled subjects with a first attack if they were evaluated within 27 days of the onset of the neurologic symptoms and had an abnormal brain MRI (at least 2 lesions consistent with MS that had no corresponding clinical symptoms). Participants were randomized to intramuscular interferon beta-1a (weekly) versus placebo. CHAMPIONS enrolled CHAMPS subjects who gave consent into an open-label follow-up study. Subjects were evaluated biannually in both studies and within a week (CHAMPS) or two weeks (CHAMPIONS) of the onset of new symptoms concerning for relapse.

Outcomes

Relapses, by definition, were new or recurring neurological symptoms that were referable to the CNS and lasted for at least 48 hours after a period of at least 30 days since the prior attack in the absence of fever or known infection. Transient recurrence of neurologic symptoms that began in the context of infection or fever was considered a pseudoexacerbation and was excluded. Based on clinical history and examination, each patient's relapses were coded as occurring in the spinal cord, brainstem/cerebellum, optic nerve, or cerebrum.

Predictors

The predictors of interest were 16 non-HLA genetic polymorphisms that had been validated for MS susceptibility as of 2010 (Table S1 in File S1) [10][20]. In the UCSF laboratory of Dr. Jorge Oksenberg, TaqMan SNP genotyping assays were used for genotyping (Applied Biosystems Inc., Foster City, CA, USA). Each PCR reaction contained 10 ng of DNA, 1 x TaqMan Genotyping Master Mix, and 1 x SNP assay (both from Applied Biosystems, Inc.). Amplification was performed with an ABI 97000 GeneAmp PCR system (Applied Biosystems, Inc.). The PCR program used was: 95° Celsius for 10 minutes followed by 50 cycles of 95° Celsius for 15 seconds and then 62° Celsius for one minute. The plates were read with an ABI prism 7900HT Sequence Detection System (with SDS 2. software; Applied Biosystems, Inc.). DRB1 genotyping was performed using a PCR locus-specific amplification, as previously described [19].

Statistics

Summaries are given as percentages for categorical variables and as mean ± standard deviation (SD) or median (interquartile range) for continuous variables. For each of the four attack locations, a dichotomous outcome variable was defined as any involvement of the region versus no involvement. The presence of at least one copy of each risk allele versus no copies of the allele was assessed for its association with attacks involving each CNS location using generalized estimating equations with robust standard errors. This repeated measures approach was used to account for within-individual correlations and to take into account the location data from both the first and second attack in the same model. All of the non-HLA susceptibility genes, as well as HLA-DRB1 status, were included in the models so that the independent effect of each gene could be determined. We also looked for interactions (with strong evidence defined by p value for interaction term <0.1) between HLA-DRB1 and EVI5 and HRA-DRB1 and CD226, because such interactions have been identified in the MS susceptibility studies [15], [17].

Results

Patient and attack characteristics

We identified 503 patients (359 [71%] were women) seen within one year of MS onset (n = 199 from UCSF; n = 43 from Marseille; n = 36 from Montreal; n = 13 from Pamplona; n = 60 from Madrid; n = 68 from CHAMPIONS; n = 84 from ONTT). The average age of onset was 33±9 years; median disease duration was 46 days (interquartile range: 1, 351). Nearly all (459; 91%) patients had an abnormal brain MRI at baseline. On average, there were 7.5 (±5.2) years of follow-up. During the first attack, the following number of subjects had involvement of each location: spinal cord (207 [41%]); brainstem/cerebellum (126 [25%]); optic nerve (198 [39%]); cerebrum (12 [2%]). First attack location could not be definitively ascertained for 3 individuals. Of the 349 patients who had a second attack, a definite location could be ascribed for 302 attacks; of these, 171 (58%) had spinal cord involvement, 74 (25%) had brainstem/cerebellar involvement, 61 (21%) had optic nerve involvement, and 7 (2%) had cerebral involvement. The location of 45 of the 58 second attacks experienced by ONTT subjects could not be ascertained; second attack location could not be definitively ascertained for 2 additional individuals. Because some patients experienced attacks in more than one CNS location (and because attack location was missing for other participants), the percentages do not add up to 100% (and the numbers don’t add up to the total number of attacks). Because so few subjects had cerebral attacks, we were unable to assess it as an outcome. Disease-modifying therapy (DMT) was started in 345 patients (69%) at some point in the follow-up period, but only 134 (27%) had received treatment for the 90 days immediately preceding the second attack or the end of follow-up (if no second attack occurred).

Details of genotype frequencies are provided in Table S1 in File S1. One gene (IRF8) in one patient failed genotyping. There were no subjects without at least one copy of the TYK2 G allele, so for this gene we assessed if two G alleles versus one was associated with attack location. No subject had more than one copy of the TNFRSF1A T allele; as such, we evaluated if having one versus no copies of the T allele was associated with attack location. Approximately half (n = 244; 49%) of the subjects had at least one copy of HLA-DRB1*1501 (the risk allele).

Predictors of spinal cord involvement

The multivariate associations of the susceptibility polymorphisms with spinal cord involvement during the first two attacks are presented in Table? 1. The IL12A polymorphism was associated with greater odds of spinal cord attacks (OR 1.52, 95% CI [1.11, 2.07], p = 0.009). IRF8 was also associated with greater odds of spinal cord involvement (OR = 2.40, 95% CI [1.04, 5.50], p = 0.040). There were trends for associations of TNFRSF1A and TMEM39A with increased odds of spinal cord involvement and of IL2RA and GPC5 with reduced odds thereof. There was no strong evidence for interaction of HLA-DRB1 with EVI5 (p value for interaction term = 0.28) or with CD226a (p value for interaction term = 0.66). The estimates of association were similar in the univariate models (Table S2 in File S1).

thumbnail

Table 1. Predictors of attack location (multivariate models).

doi:10.1371/journal.pone.0075565.t001

Predictors of brainstem/cerebellum involvement

In the multivariate model, the IL7R polymorphism was associated with lower odds of brainstem/cerebellar involvement (OR 0.46, 95% CI [0.22, 0.97], p = 0.041). The IL12A (OR 0.72, 95% CI [0.51, 1.02], p = 0.068) and TNFRSF1A (OR 0.55, 95% CI [0.30, 1.02], p = 0.056) polymorphisms tended to be associated with reduced odds of involvement of the brainstem/cerebellum, while the opposite was true of the MPHOSPH9 polymorphism, although the confidence intervals included 1.0. Again, there was no strong evidence for interaction of HLA-DRB1 with EVI5 (p value for interaction term = 0.89) or with CD226a (p value for interaction term = 0.17). The univariate models had similar results (Table S2 in File S1).

Predictors of optic nerve involvement

The CD6 polymorphism was associated with reduced odds of optic neuritis during the first two attacks (OR = 0.69, 95% CI [0.49, 0.97], p = 0.034). There were trends for associations of the TMEM39A and MPHOSPH9 polymorphisms with reduced odds of optic neuritis as well. There was no strong evidence for interaction of HLA-DRB1 with EVI5 (p value for interaction term = 0.48) or with CD226a (p value for interaction term = 0.95). The results were not meaningfully different in the univariate models (Table S2 in File S1).

Discussion

In this study, we have identified, among genetic polymorphisms known to increase the risk of MS, several that also may play a role in the CNS location of early MS attacks. While the confidence intervals are wide, such that the associations should not be over-interpreted, the results are intriguing from a scientific standpoint. Both the IL12A and the IRF8 polymorphisms were associated with spinal cord involvement. Secreted by dendritic cells and monocytes, IL12 stimulates IFNγ and suppresses TH2 cells [20]. The initiation of treatment with an approved MS disease-modifying therapy, interferon beta-1a, leads to increased expression of IL12 [21]. In light of these data, that the IL12A polymorphism is associated with increased spinal cord involvement is of interest given our previous work showing that patients treated with interferon beta were more likely to experience spinal cord involvement during attacks, an association that was independent of having had prior attacks involving the spinal cord [1], [22]. In contrast, subgroup analyses of the CHAMPS study showed that treatment with interferon seemed to confer the strongest protective effect against conversion to MS after the first attack in those whose first attack localized to the spinal cord [23]. The exact role that IL12 plays in the frequency and location of MS attacks, as well as the functional consequences of the polymorphism known to be associated with MS risk, thus demands further investigation.

The connection between polymorphisms in the other genes identified here and altered MS attack location is less clear. IRF8 has been associated with regulating innate immunity, with B cell development, and with the differentiation of CD8+ T cells [24]. IL7 is important to many immune functions, including T cell differentiation and maintenance, and the IL7R polymorphism confers a functional change; specifically, it leads to an increase in the soluble (versus membrane-bound) form of the IL7R α chain [25]. Finally, the CD6 gene product is involved in continuing the activation of T cells; CD4+ T cell proliferation is reduced in healthy people with two copies of the risk allele [26]. Alternatively, the polymorphisms may influence MS attack location based on local molecular properties in these regions, such as differential expression of adhesion blood-brain barrier molecules, similar to the finding that aquaporin-4 protein is located in the spinal cord, hypothalamus, and optic nerve, driving the involvement of these regions in neuromyelitis optica. The genes could also have roles that vary depending on CNS location.

Limitations of the study include that the results may not be generalizable to non-white patients with MS. That some ONTT patients were missing detailed data regarding second attack locations may have introduced some bias. Also, in some of the models, the 95% CIs included 1.0 even when a convincing direction of association emerged. Because genotyping for this study occurred prior to the publication of the 2011 MS genome-wide association study [27], we were unable to evaluate some of the susceptibility SNPs identified in that study for their association with relapse location. Finally, although the factors all have some a priori plausibility for influencing the phenotype of MS due to their known influence on susceptibility, it is possible that these findings are still due to chance given the number of analyses. Because this was a preliminary study in which the goal was to explore, rather than definitively ascertain, the associations between the SNPs and relapse location, we were more interested in evaluating the magnitudes and directions of the effects seen rather than only focusing on the p-values and performing adjustments for multiple comparisons. Our results thus need replication in a large, independent dataset, ideally including the more recently-identified MS risk alleles at the time such a study is conducted [27].

The location of early MS attacks within the CNS does not appear to be random. Here, we have identified that some of the genes important to MS susceptibility also influence which anatomic structures are affected. While the interplay between these genes and other as yet unknown factors important to attack location is not completely clear, the results presented here add to the growing support for utilizing genetics to predict disease outcomes in MS and to tailor therapies, based on those predictions, to each individual with the disease.

Supporting Information

File S1.

Table S1. Non-HLA genes associated with MS as of 2010. Table S2. Univariate predictors of location.

doi:10.1371/journal.pone.0075565.s001

(DOCX)

Acknowledgments

The authors thank Kelley Weinfurtner and Tammy Skaramagas for their dataset help. Genotyping was conducted by Stacy Caillier in the laboratory of Dr. Jorge Oksenberg at UCSF. DNA samples were provided by the Optic Neuritis Treatment Trial.

Author Contributions

Conceived and designed the experiments: EMM PB EW. Performed the experiments: EMM PB EW. Analyzed the data: EMM PB EW. Contributed reagents/materials/analysis tools: EMM RFC MRB JP PD PV IM ER RPK JM EW. Wrote the paper: EMM PB EW. Critical revisions to manuscript: EMM RFC MRB JP PD PV IM ER RPK JM PB EW.

References

  1. 1. Mowry EM, Deen S, Malikova I, Pelletier J, Bacchetti P, et al. (2009) The onset location of multiple sclerosis predicts the location of subsequent relapses. J? Neurol Neurosurg Psychiatry 80: 400–403. doi: 10.1136/jnnp.2008.157305
  2. 2. Kinkel RP, Kollman C, O'Connor P, Murray TJ, Simon J, et al. (2006) IM interferon beta-1a delays definite multiple sclerosis 5 years after a first demyelinating event. Neurology 66: 678–684. doi: 10.1212/01.wnl.0000200778.65597.ae
  3. 3. Beck RW, Clearly PA, Anderson MM, Keltner JL, Shults WT, et al. (1992) A randomized, controlled trial of corticosteroids in the treatment of acute optic neuritis. N? Engl? J? Med 326: 581–588. doi: 10.1056/nejm199202273260901
  4. 4. Waubant E, Pelletier D, Mass M, Cohen JA, Kita M, et al. (2012) Randomized controlled trial of atorvastatin in CIS: the STAyCIS study. Neurology 78: 1171–1178. doi: 10.1212/wnl.0b013e31824f7fdd
  5. 5. Marcus JF, Spain RI, McDonald J, Julian LJ, Okuda DT, et al. (2010) Baseline data and rationale for study of riluzole for neuroprotection in early relapsing remitting multiple sclerosis. Mult Scler 16: 1004.
  6. 6. Confavreux C (1994) Establishment and use of multiple sclerosis registers—EDMUS. Ann Neurol 36: S136–139. doi: 10.1002/ana.410360733
  7. 7. Toledo J, Sepulcre J, Salinas-Alaman A, Garcia-Layana A, Murie-Fernandez M, et al. (2008) Retinal nerve fiber layer atrophy is associated with physical and cognitive disability in multiple sclerosis. Mult Scler 14: 906–912. doi: 10.1177/1352458508090221
  8. 8. Optic Neuritis Study Group (1991) The clinical profile of optic neuritis: experience of the optic neuritis treatment trial. Arch Ophthalmol 109: ;1673–1678
  9. 9. Jacobs LD, Beck RW, Simon JH, Kinkel RP, Brownscheidle CM, et al. (2000) Intramuscular interferon beta-1a therapy initiated during a first demyelinating event in multiple sclerosis. CHAMPS study group. N? Engl? J? Med 343: 898–904. doi: 10.1016/s0002-9394(00)00915-6
  10. 10. International Multiple Sclerosis Genetics Consortium (2007) Risk alleles for multiple sclerosis identified by a genomewide study. N? Engl? J? Med 357: 851–862. doi: 10.1056/nejmoa073493
  11. 11. International Multiple Sclerosis Genetics Consortium (2010) Comprehensive follow-up of the first genome-wide association study of multiple sclerosis identifies KIF21B and TMEM39A as susceptibility loci. Hum Mol Gen 19: 953–962. doi: 10.1093/hmg/ddp542
  12. 12. DeJager PL, Baecher-Allan C, Maier LM, Arthur AT, Ottoboni L, et al. (2009) The role of the CD58 locus in multiple sclerosis. Proc Natl Acad Sci? U? S? A 106: 5264–5269.
  13. 13. Ban M, Goris A, Lorentzen AR, Baker A, Mihalova T, et al. (2009) Replication analysis identifies TYK2 as a multiple sclerosis susceptibility gene. Eur? J? Hum Genet 17: 1309–1013. doi: 10.1038/ejhg.2009.41
  14. 14. DeJager PL, Jia X, Wang J, deBakker PI, Ottoboni L, et al. (2009) Meta-analysis of genome scans and replication identify CD6, IRF8 and TNFRSF1A as new multiple sclerosis susceptibility loci. Nat Genet 41: 776–782. doi: 10.1038/ng.401
  15. 15. Australia and New Zealand Multiple Sclerosis Genetics Consortium (2009) Genome-wide association study identifies new multiple sclerosis susceptibility loci on chromosomes 12 and 20. Nat Genet 41: 824–828.
  16. 16. Baranzini SE, Wang J, Gibson RA, Galwey N, Naegelin Y, et al. (2009) Genome-wide association analysis of susceptibility and clinical phenotype in multiple sclerosis. Hum Mol Genet 18: 767–778. doi: 10.1093/hmg/ddn388
  17. 17. Johnson BA, Wang J, Taylor EM, Caillier SJ, Herbert J, et al. (2010) Multiple sclerosis susceptibility alleles in African Americans. Genes Immun 11: 343–350. doi: 10.1038/gene.2009.81
  18. 18. Hoppenbrouwers IA, Aulchenko YS, Janssens AC, Ramagopalan SV, Broer L, et al. (2009) Replication of CD58 and CLEC16 as genome-wide significant risk genes for multiple sclerosis. J? Hum Genet 54: 676–680. doi: 10.1038/jhg.2009.96
  19. 19. International Multiple Sclerosis Genetics Consortium (2009) The expanding genetic overlap between multiple sclerosis and type I diabetes. Genes Immun 10: 11–14. doi: 10.1038/gene.2008.83
  20. 20. International Multiple Sclerosis Genetics Consortium (2010) IL12A, PHOSPH9/CDK2AP1 and RGS1 are novel multiple sclerosis susceptibility loci. Genes Immun 11: 397–405. doi: 10.1038/gene.2010.28
  21. 21. Kress-Bennett JM, Ehrlich GD, Bruno A, Post JC, Hu FZ, et al. (2011) Preliminary study: treatment with intramuscular interferon beta-1a results in increased levels of IL12Rβ2+ and decreased levels of IL23R+ CD4+ T-lymphocytes in multiple sclerosis. BMC Neurol 11: 155. doi: 10.1186/1471-2377-11-155
  22. 22. Deen S, Bacchetti P, High A, Bostrom A, Waubant E (2008) Predictors of the location of multiple sclerosis relapse. J? Neurol Neurosurg Psychiatry 79: 1190–1193. doi: 10.1136/jnnp.2007.136440
  23. 23. Beck RW, Chandler DL, Cole SR, Simon JH, Jacobs LD, et al. (2002) Interferon beta-1a for early multiple sclerosis: CHAMPS trial subgroup analysis. Ann Neurol 51: 481–490. doi: 10.1002/ana.10148
  24. 24. Miyagawa F, Zhang H, Terunuma A, Ozato K, Tagaya Y, et al. (2012) Interferon regulatory factor 8 integrates T-cell receptor and cytokine-signaling pathways and drives effector differentiation of CD8+ T cells. Proc Natl Acad Sci? U? S? A 109: 12123–12128. doi: 10.1073/pnas.1201453109
  25. 25. Gregory SG, Schmidt S, Seth P, Oksenberg JR, Hart J, et al. (2007) Interleukin 7 receptor α chain (IL7R) shows allelic and functional association with multiple sclerosis. Nat Genet 39: 1083–1091. doi: 10.1038/ng2103
  26. 26. Kofler DM, Severson CA, Mousissian N, DeJager PL, Hafler DA (2011) The CD6 multiple sclerosis susceptibility allele is associated with alterations in CD4+ T cell proliferation. J? Immunol 187: 3286–2391. doi: 10.4049/jimmunol.1100626
  27. 27. Sawcer S, Hellenthal G, Pirinen M, Spencer CC, Patsopoulos NA, et al. (2011) Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature 476: 214–219.