Multiple Sclerosis Resource Centre

Welcome to the Multiple Sclerosis Resource Centre. This website is intended for international healthcare professionals with an interest in Multiple Sclerosis. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

Genetic variation in PBMC-produced IFN-γ and TNF-α associations with relapse in multiple sclerosis

Journal of the Neurological Sciences, 1-2, 349, pages 40 - 44

Abstract

Background

Alterations in peripheral blood mononuclear cell (PBMC) cytokine production have been found in multiple sclerosis (MS) compared to healthy controls. We have previously found that stimulated PBMC-produced TNF-α and IFN-γ modulated MS relapse risk, such that raised TNF-α was protective, while raised IFN-γ increased relapse risk.

Objective

To assess whether SNPs within genes for relevant cytokines and their receptors modulate the associations of TNF-α and IFN-γ with relapse, thus providing additional information about these cytokine effects and the roles of these genes in MS.

Methods

Prospective cohort of 91 participants with relapsing–remitting MS and cytokine and genotype data. SNPs (N = 361) within a window of 10 kb around each cytokine/cytokine receptor gene (N = 84) were selected for analysis. Predictors of PBMC cytokines were evaluated by multilevel mixed-effects linear regression. Predictors of relapse were evaluated by Cox proportional hazards regression. Bonferroni correction was used to adjust for multiple testing; thusp < 1.39 × 10− 4was defined as significant.

Results

Individuals of GG genotype of rs3218295 (within the geneIL2RB) demonstrated a significant protective effect of TNF-α on relapse while those of GA/AA genotype showed a significant positive association (pinteraction = 5.04 × 10− 5). Carriers of CC genotype of rs522807 (3′ region ofTNFRSF1B) and the AA genotype of rs25879 (5′ region ofIL3) showed a strong association between IFN-γ and increased relapse risk (pinteraction = 8.21 × 10− 5and 1.70 × 10− 5, respectively).

Conclusions

Our results show novel modulation of TNF-α and IFN-γ associations with relapse by SNPs in major cytokines. These findings suggest the potential for these genes and/or their products as potential therapeutic targets in MS.

Highlights

 

  • We studied whether SNPs in cytokine genes modulated cytokine associations with relapse risk in MS.
  • GG genotype of rs32183295 inIL2RBenhanced TNF-α anti-relapse association w/relapse.
  • A allele in rs522807 inTNFRSF1Benhanced TNF-α anti-relapse association w/relapse.
  • G allele of rs25879 ofIL3abrogated IFN-γ association w/relapse.
  • A allele of rs522807 ofTNFRSF1Babrogated IFN-γ association w/relapse.

Keywords: Multiple sclerosis, Epidemiology, Interferon-gamma, Tumour necrosis factor alpha, Peripheral blood mononuclear cell, Relapse.

1. Introduction

Multiple sclerosis (MS) is a progressive demyelinating neurological disease of the central nervous system resulting from the effects of both environmental risk factors and susceptibility loci [1] . Its major clinical subtype is relapsing–remitting MS (RRMS), which is characterised by intermittent and irregular relapses followed by periods of relative clinical inactivity (remission). In our previous work, we found that stimulated PBMC-produced TNF-α was associated with a lower risk of relapse, while stimulated PBMC-produced IFN-γ was significantly associated with an increased risk of relapse. We further found that one MS-related SNP, rs1800693, within the TNF-α-receptor gene significantly modified TNF-α association with relapse, such that the secreted form blocked the protective effect of TNF-α [2] .

In this paper, we provide further genetic context by assessing whether SNPs within a greater number of genes for cytokines and their receptors significantly modulate the associations of TNF-α and IFN-γ with relapse, thus providing information about the effects of these cytokine and the potential role of these loci in MS.

2. Methods

2.1. Study design

The Southern Tasmanian Multiple Sclerosis Longitudinal (MSL) Study followed a cohort of 203 persons with clinically definite MS living in southern Tasmania, Australia over 2002–2005 [3] . This population-based study solicited participants from the sole MS clinic servicing southern Tasmania operating out of the public hospital (19.2%), with additional recruitment from participants in previous studies (80.8%) and via MS Society referral. Accordingly, an estimated 78% (203/259) of eligible cases in the region were included, and data from 198 participants were obtained for analysis [3] . Of these 198, 183 persons completed the summer or winter 2004 reviews, allowing them to be included in the present analysis. Of the 183 persons, 160 (87.4%) had data on any of the four cytokines assessed and 141 (77.1%) had data on all cytokines. As the primary clinical outcome was the occurrence of relapse during the study, analysis was restricted to the 119 persons with relapsing–remitting MS followed beyond one review with cytokine data. Analysis was then restricted to the 91 participants with data on genotype.

At each biannual review participants were asked about relevant environmental and behavioural factors, including medications, smoking, physical activity, time in the sun and diet.

Ethics approval was obtained from the Southern Tasmania Human Research Ethics Committee. All participants provided informed consent.

2.2. Measurement of relapses

A relapse was defined according to the 2001 McDonald Criteria [4] as the acute or subacute appearance or reappearance of a neurological abnormality (lasting at least 24 h), immediately preceded by a stable, improving, or slowly progressive neurological state for 30 days, in the absence of fever, known infection, concurrent steroid withdrawal, or externally derived increases in body temperature. Relapses were reported in real time by phone or at biannual review, and all reports were validated by the study neurologist.

2.3. Biological samples

At each biannual review, blood samples were obtained by standard venipuncture and processed to separate serum and buffy coat. In two of these reviews, Summer 2004 and Winter 2004, buffy coats were processed to separate out PBMCs: collected blood was layered over histopaque (Sigma), spun at 400 gfor 30 min, and the interface containing the PBMC was removed and washed twice in incomplete RPMI-1640 (JRH Biosciences, Melbourne, Australia). The cells were then resuspended in a 10% DMSO/RPMI solution and stored at –80 °C until required. To avoid a storage effect, as soon as a season's collection was completed, the cells were thawed and counted using a haemocytometer. Cell viability was ascertained by Trypan blue exclusion, and any samples with viability below 90% were discarded. Cells were then washed twice in incomplete RPMI and resuspended in complete RPMI containing 2 mMl-glutamine (Sigma, Castle Hill, Australia), 100 U Penicillin, 100 U Gentamicin (Pharmacia, Murarrie, Australia) and 10% FCS (JRH Biosciences) at 1 × 106 cells/ml in 24 well cell culture plates. Five micrograms per milliliter phytohaemagglutinin (PHA) (Sigma) was added to each sample, and the samples were incubated for 24 h at 37 °C, 8% CO2. Supernatant was then stored at − 80 °C for the duration of the study , for analysis by ELISA thereafter.

2.4. ELISA protocol

Samples for IL-4 and IL-10 analyses were assayed undiluted, and those for IFN-γ and TNF-α analyses were diluted 1:4 with Assay Diluent (BD Biosciences, San Diego). ELISAs were performed using BD OptEIA ELISA kits and reagents, as per the manufacturer's instructions. To ensure consistency, each plate (Nunc, Roskilde, Denmark) contained matched MS summer/winter samples, control samples and standards. Preliminary analyses showed that storage of supernatants for approximately 12 months had no effect on cytokine levels.

2.5. SNP identification & genotyping

SNPs (N = 361) within a window of 10 kb around all the cytokine genes and their receptor genes (N = 84) were selected for analysis. A list of these genes and the SNPs are in the Supplemental table. These SNPs were directly genotyped in 189 of our samples as either part of the ANZgene MS GWAS or by an additional GWAS, extensively described in our previous study [5] . Briefly, 164 MS cases were genotyped as a part of ANZgene MS GWAS and additionally 29 MS cases were newly genotyped using the Illumina HumanOmniExpress-12v1_A array. All the samples were previously identified as being from individuals of European descent [6] .

2.6. Statistical analysis

As in the previous work [2] , the associations of cytokines with relapse were evaluated in a multivariable model wherein IFN-γ, TNF-α and IL-10 were mutually adjusted, while a categorical of IL-4 was stratified upon (allowing the baseline hazard to vary by level of the IL-4 categorical) due to its violation of the proportional hazards assumption.

Predictors of PBMC cytokines were evaluated using multilevel mixed-effects linear regression. Transformation was applied as required to satisfy homoscedasticity; however all coefficients are reported on the scale of the original value.

All covariates satisfied the proportional hazards assumption, except IL-4, so all models were stratified by a four-level categorical rendition of IL-4 (11.5–52.9, > 52.9–86.8, > 86.8–127.6, > 127.6–511.9).

Interaction of SNPs and cytokines with relapse was evaluated by Cox proportional hazards for repeated events, as described previously [7] , whereby multiple relapses by the same persons are treated as independent observations but accounted for at the intra-individual level, and the time until a prior event does not influence the composition of the risk set for a subsequent event. Where interaction was assessed, a product term of the SNP and cytokine was generated and included in the multivariable model, the significance of this term defining the significance of statistical interaction.

Adjusting for multiple testing using the Bonferroni correction method,p < 1.39 × 10− 4was defined as significant.

All analyses were performed using Stata/SE for Windows (Version 12.1; StataCorp LP College Station, TX, USA).

3. Results

As discussed previously, our cohort was majority female, of middle age and low disability, and a minority had any relapses, both during the study and during the analysis period here, reflecting their largely being on immunomodulatory therapy ( Table 1 ).

Table 1 Cohort characteristics of sample analysed.

  n (%)
All persons 119
Sex  
 Male 31 (26.1)
 Female 88 (74.0)
Age at study entry (years)  
 21–39 30 (25.2)
 > 39–45 28 (23.5)
 > 45–52 31 (26.1)
 > 52–77 30 (25.2)
Progression to SPMS during study  
 No 105 (88.2)
 Yes 14 (11.8)
Relapse during study?  
 No  
 Yes 59 (49.6)
Relapse during analysis period?  
 No  
 Yes 33 (27.7)
Any immunomodulatory therapy during study?  
 No  
 Yes 98 (82.4)
MS duration from symptom onset (years) 10 (5–17)
EDSS 2.5 (2–4)
Number of relapses during analysis period 0 (0–1)

SPMS = secondary-progressive multiple sclerosis; EDSS = Expanded Disability Status Scale.

3.1. Cytokine SNPs to evaluate associations with TNF-α and IFN-γ production

After adjusting for multiple testing using the Bonferroni correction method (p < 1.39 × 10− 4was defined as significant), we did not observe any cytokine-related SNPs which significantly impacted on TNF-α and IFN-γ production (data not shown).

3.2. Differential associations of TNF-α and IFN-γ with relapse by SNP genotype

As in Table 2 , we found one SNP, rs3218295, located within the geneIL2RB, which showed a significant interaction with TNF-α as a predictor of relapse (pinteraction = 5.04 × 10− 5). Carriers of the GG genotype showed a significant protective effect of TNF-α on relapse (p = 1.46 × 10− 5), while carriers of the GA or AA genotypes showed a significantly increased risk of relapse from increased TNF-α (p = 0.001) ( Table 1 ). Additionally, the SNP rs522807, located within the 3′ region of the geneTNFRSF1B, interacted with TNF-α's reduction in the risk of relapse (pinteraction = 3.83 × 10− 4), such that carriers of the minor A allele showed a much greater protective effect of TNF-α on relapse risk than those homozygous for the CC genotype, who had an association no different from the parent cohort.

Table 2 SNPs showing an interactive association with TNF-α as a predictor of relapse (p < 1.39 × 10− 4was defined as significant).

SNP Type Related genes Genotype # of relapses HR (95% CI) Significance
Aggregate         0.48 (0.18, 1.29) p = 0.150
rs522807 (A/C) 3′ region, intronic TNFRSF1B CC 35 0.47 (0.15, 1.43) p = 0.180
TFBS   CA+AA 3 0.05 (0.01, 0.32) p = 0.001
          pinteraction = 3.83 × 10− 4
rs3218295 (A/G) Within gene, intronic IL2RB GG 30 0.18 (0.08, 0.39) p = 1.46 × 10− 5
    GA+AA 8 7.41 (2.20, 24.95) p = 0.001
          pinteraction = 5.04 × 10− 5

All analyses mutually adjusted for cytokines (IFN-γ, TNF-α, and IL-10), and for age, sex, time in sun in preceding 2/52, ever smoked, IFN-beta medication use, and stratified by IL4 quartile categorical. Results are presented for 100-unit increments of the cytokine. Figures in boldface denote statistically significant associations (p<0.05).

pinteraction < 0.0001 was defined as significant (pinteractionrepresents thepvalue for interaction).

Abbreviations: TFBS: transcriptional factor binding sites.

We also found two SNPs showing interactive associations with IFN-γ as a predictor of relapse ( Table 3 ). The rs522807 SNP, located within the 3′ region of the geneTNFRSF1B, showed a significant interaction with IFN-γ (pinteraction = 8.21 × 10− 5), such that carriers of the CC genotype had an enhanced risk of relapse from increased IFN-γ, while carriers of the CA or AA genotypes showed no association between IFN-γ and relapse. Likewise the rs25879 SNP, located within the 5′ region of the gene ofIL3, interacted significantly with IFN-γ (pinteraction = 1.70 × 10− 5), showing a significantly increased association of IFN-γ with relapse among carriers of the AA genotype, while carriers of the AG or GG genotypes showed no association between IFN-γ and relapse. Both SNPs are within the transcriptional factor binding sites of their respective genes.

Table 3 SNPs showing an interactive association with IFN-γ as a predictor of relapse (p < 1.39 × 10− 4was defined as significant).

SNP Type Related genes Genotype # of relapses HR (95% CI) Significance
Aggregate         1.75 (1.04, 2.94) p = 0.035
rs522807 (A/C) 3′ region, intronic TNFRSF1B CC 35 2.51 (1.46, 4.30) p = 0.001
TFBS   CA+AA 3 0.56 (0.25, 1.27) p = 0.160
          pinteraction = 8.21 × 10− 5
rs25879 (G/A) 5′ region, intronic IL3 AA 29 2.65 (1.48, 4.74) p = 0.001
TFBS   AG+GG 9 0.50 (0.21, 1.22) p = 0.130
          pinteraction = 1.7 × 10− 5

All analyses mutually adjusted for cytokines (IFN-γ, TNF-α, and IL-10), and for age, sex, time in sun in preceding 2/52, ever smoked, IFN-beta medication use, and stratified by IL4 quartile categorical. Results are presented for 100-unit increments of the cytokine. Figures in boldface denote statistically significant associations (p<0.05).

pinteraction < 0.0001 was defined as significant (pinteractionrepresents thepvalue for interaction).

Abbreviations: TFBS: transcriptional factor binding sites.

4. Discussion

In this work, we systematically examined the SNPs within a large number of genes for cytokines and their receptors to assess whether there was an interaction with the previously demonstrated associations [2] of TNF-α and IFN-γ with relapse. Individuals of GG genotype of rs3218295 (within the geneIL2RB) demonstrated a significant protective effect of TNF-α on relapse while those of GA/AA genotype did not (pinteraction = 5.04 × 10− 5). Carriers of CC genotype of rs522807 (3′ region ofTNFRSF1B) and the AA genotype of rs25879 (5′ region ofIL3) showed a strong association between IFN-γ and increased relapse risk (pinteraction = 8.21 × 10− 5and 1.70 × 10− 5respectively). That we have here shown allelic variants in a SNP in the TNF-α receptor gene modulates both the associations of TNF-α and IFN-γ, with carriers of the A allele showing a potentiated protective effect against relapse from increased TNF-alpha, and an abrogation of the increased relapse hazard from increased IFN-gamma, is supportive of a true effect for this SNP. The findings of significant interactions with SNPs in the IL-2 receptor and IL-3 genes with TNF-α and IFN-γ, respectively, are also interesting and reflect the complex interplay between the various cytokines and cytokine receptors of the immune system.

Despite multiple MS risk GWAS[8] and [9], no susceptibility loci in or near the gene ofIL2RBorIL3have previously been demonstrated, though certainly TNF-α receptor is now a well-recognised MS susceptibility locus [10] . That these loci have not been identified in MS GWAS is not unexpected, because GWAS are focused more on cross-sectional data and is limited by a greater multiple testing burden. This reflects the utility in examining SNPs in parameters relevant to the already demonstrated cytokine risk factors, because by casting the net in this fashion and using a systems biology approach, we have identified results that may uncover modulators of these biological risk factors. Similar approaches have been taken with regard to vitamin D/UV[11], [12], [13], and [14], childhood infections [15] and EBV [16] that have either uncovered novel mechanisms of already known risk alleles for MS[12], [13], [14], [15], and [16]or uncovered novel loci that modulate the course of MS [11] .

Our findings of effects for the three SNPs presented here are not wholly novel in terms of relevance to neurological function. The impact ofTNFRSF1BSNPs on MS has been well-demonstrated [10] . Our other SNPs have additional evidence in support of them being biologically plausible candidates. A locus nearIL3(rs31480) is in high linkage disequilibrium with our significant SNP (rs25879, r2 > 0.8). This SNP could influence the binding affinity of the transcription factor SP1 and thus impact upon the expression ofIL3. Others have shown that the level ofIL3promoted the proliferation and survival of neural progenitors [17] . Consistent with this finding, our major allele genotype AA of rs25879 and the major allele genotype CC of rs31480 showed the same effect direction. That is, AA of rs25879 interacted significantly with IFN-γ to increase relapse risk, while CC of rs31480 in their findings was associated with lower brain volume and lowerIL3expression which would decrease neural progenitors' survival and proliferation. These results suggest a possible mechanism by whichIL3might interact with IFN-γ, potentially regulating brain processes and affecting susceptibility to relapse.

IL2RBencodes the beta chain of the IL-2 receptor, expressed on T- and B-cells, monocytes, neutrophils and dendritic cells. In mice, the blocked expression ofIL2RBmanifested in the dysregulation of T- and B-cell activation and behaviour, as well as the loss of thymocytes, leading to death by 12 weeks of age [18] . Thus, IL-2R-beta appears to be required for proper control and function of immune cell activation, reducing the potential for inappropriate immune activation [18] . IL-2R-beta can bind to IL-15, whose three-dimensional structure is similar to that of IL-2 [19] . IL-15 plays an important role in the maintenance and activation of CD8 T lymphocytes, a prominent lymphocyte population found in MS lesions [20] . In MS, IL-15 serum levels were higher in patients with relapse compared with patients in stable periods of the disease (p = 0.001) [21] . Although a role ofIL2RBin susceptibility to other inflammatory diseases like rheumatoid arthritis [22] and asthma [23] has been found in some studies, its role in MS remains unknown, so further studies are required to better understand its role in disease.

This study benefits from its prospective cohort design and real-time reporting of relapses. The current analysis is limited to some extent in its being conducted on only a subset of the MSL study cohort, but nonetheless provides valuable information on the nature of the immune response in MS clinical course and the potential genetic impacts thereon. The biannually obtained PBMCs do allow a reasonably representative ascription of immune state around the time of the relapses, allowing these to be a predictor of subsequent relapse risk. That said, there is variability over time due to changing season, changes in medication and infection state, among others, which can impact on the responsiveness of immune cells. That said, our use of a standardised and appreciable stimulus to induce cytokine production should allow a comparable cytokine production between persons that can overcome this micro-variation over time to give a representation of the general immune state of the individual.

In conclusion, our results provide insight into the complex interactions between cytokines and genetic predictors that may modulate the risk of relapse in MS. While a validation study replicating our results is needed in order to establish them for future use in clinical practice, these findings may provide some clues to better understand the immunopathology of MS and may suggest possible points of intervention in moderating clinical course. They also highlight the potential of more targeted future therapeutics for MS that take into account a person's genotype.

Contribution statement

BT, IvdM, A-LP, and TD were involved in the conception, planning, and acquisition of funding for the study. BT, IvdM, FP, A-LP, and TD were involved in the acquisition of data for the study. Analysis of peripheral blood mononuclear cell cytokine production by NS. Conception of current analysis by SSJ. Statistical analysis of PBMC–SNP interactive relationship predicting relapse undertaken by YZ under supervision by SSJ, with statistical support by LB and genetic analysis support by JC. YZ, SSJ and BT involved in the drafting of the article. All authors were involved in the critical revision of the manuscript and approved it for submission.

Funding statement

The MS Longitudinal Study was funded by a grant from the Australian National Health & Medical Research Council (Project 211308). SSJ is supported by a postdoctoral fellowship from Multiple Sclerosis Research Australia (12051). IvdM is supported by a fellowship from the Australian Research Council (FT100100553).

Conflict of interest

The authors have no conflicts of interest to disclose.

Acknowledgements

We would like to thank all the study participants and the research staff who assisted with this project, particularly our MSL research nurse P Groom.

We also thank Andrew Kemp for the helpful comments on an earlier iteration of this paper.

Appendix A. Supplementary data

 

Download file

Supplemental Table: Tab 1 lists the genes evaluated in the current analysis, including the gene name, the start and stop positions in the genome and the chromosome it is located in (genes in red could not be evaluated, either because they were on the X chromosome or we did not have any SNPs in these genes in our dataset). Tab 2 lists the SNPs evaluated in the current analysis, including the gene in which they were located and information about the gene, the SNP name and position, and the major and minor allele for that SNP.

References

  • [1] A. Compston, H. Lassmann, K. Smith. The neurobiology of multiple sclerosis. A. Compston (Ed.) McAlpine's multiple sclerosis 4th ed. (Churchill Livingston, Philadelphia, 2005) 449-490
  • [2] S. Simpson Jr., N. Stewart, I. van der Mei, P. Otahal, J. Charlesworth, A.L. Ponsonby, et al. Stimulated PBMC-produced IFN-gamma and TNF-alpha are associated with altered relapse risk in multiple sclerosis: results from a prospective cohort study. J Neurol Neurosurg Psychiatry. 2014;10.1136/jnnp-2013-307336
  • [3] F. Pittas, A.L. Ponsonby, I.A. van der Mei, B.V. Taylor, L. Blizzard, P. Groom, et al. Smoking is associated with progressive disease course and increased progression in clinical disability in a prospective cohort of people with multiple sclerosis. J Neurol. 2009;256:577-585 Crossref
  • [4] W.I. McDonald, A. Compston, G. Edan, D. Goodkin, H.P. Hartung, F.D. Lublin, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol. 2001;50:121-127 Crossref
  • [5] R. Lin, B.V. Taylor, S. Simpson Jr., J. Charlesworth, A.L. Ponsonby, F. Pittas, et al. Novel modulating effects of PKC family genes on the relationship between serum vitamin D and relapse in multiple sclerosis. J Neurol Neurosurg Psychiatry. 2014;85:399-404 Crossref
  • [6] Australia, New Zealand Multiple Sclerosis Genetics C. Genome-wide association study identifies new multiple sclerosis susceptibility loci on chromosomes 12 and 20. Nat Genet. 2009;41:824-828
  • [7] S. Simpson Jr., B. Taylor, L. Blizzard, A.L. Ponsonby, F. Pittas, H. Tremlett, et al. Higher 25-hydroxyvitamin D is associated with lower relapse risk in multiple sclerosis. Ann Neurol. 2010;68:193-203
  • [8] M.L. Cavanillas, A. Alcina, C. Nunez, V de las Heras, M Fernandez-Arquero, M Bartolome, et al. Polymorphisms in the IL2, IL2RA and IL2RB genes in multiple sclerosis risk. Eur J Hum Genet. 2010;18:794-799 Crossref
  • [9] International Multiple Sclerosis Genetics, Wellcome Trust Case Control, S. Sawcer, G. Hellenthal, M. Pirinen, C.C. Spencer, et al. Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature. 2011;476:214-219
  • [10] A.P. Gregory, C.A. Dendrou, K.E. Attfield, A. Haghikia, D.K. Xifara, F. Butter, et al. TNF receptor 1 genetic risk mirrors outcome of anti-TNF therapy in multiple sclerosis. Nature. 2012;488:508-511 Crossref
  • [11] R. Lin, B.V. Taylor, S. Simpson Jr., J. Charlesworth, A.L. Ponsonby, F. Pittas, et al. Novel modulating effects of PKC family genes on the relationship between serum vitamin D and relapse in multiple sclerosis. J Neurol Neurosurg Psychiatry. 2013;85:399-404
  • [12] S.V. Ramagopalan, H.E. Hanwell, G. Giovannoni, P.M. Knappskog, H.I. Nyland, K.M. Myhr, et al. Vitamin D-dependent rickets, HLA-DRB1, and the risk of multiple sclerosis. Arch Neurol. 2010;67:1034-1035
  • [13] R. Lin, B.V. Taylor, S. Simpson Jr., J. Charlesworth, A.L. Ponsonby, F. Pittas, et al. Association between multiple sclerosis risk-associated SNPs and relapse and disability — a prospective cohort study. Mult. Scler.. 2013;20:313-321
  • [14] S.V. Ramagopalan, J. Link, J.K. Byrnes, D.A. Dyment, G. Giovannoni, R.Q. Hintzen, et al. HLA-DRB1 and month of birth in multiple sclerosis. Neurology. 2009;73:2107-2111 Crossref
  • [15] S.V. Ramagopalan, D.A. Dyment, G. Giovannoni, A.D. Sadovnick, G.C. Ebers. HLA-DRB1*15, low infant sibling exposure, and multiple sclerosis gene–environment interaction. Ann Neurol. 2010;67:694-695
  • [16] S.V. Ramagopalan, A.D. Sadovnick, G.C. Ebers, G. Giovannoni. Effects of infectious mononucleosis and HLA-DRB1*15 in multiple sclerosis. Mult Scler. 2010;16:127-128 Crossref
  • [17] X.J. Luo, M. Li, L. Huang, K. Nho, M. Deng, Q. Chen, et al. The interleukin 3 gene (IL3) contributes to human brain volume variation by regulating proliferation and survival of neural progenitors. PLoS One. 2012;7:e50375 Crossref
  • [18] H. Suzuki, T.M. Kundig, C. Furlonger, A. Wakeham, E. Timms, T. Matsuyama, et al. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor beta. Science. 1995;268:1472-1476
  • [19] S.C. Wuest, J.H. Edwan, J.F. Martin, S. Han, J.S. Perry, C.M. Cartagena, et al. A role for interleukin-2 trans-presentation in dendritic cell-mediated T cell activation in humans, as revealed by daclizumab therapy. Nat Med. 2011;17:604-609 Crossref
  • [20] P. Saikali, J.P. Antel, C.L. Pittet, J. Newcombe, N. Arbour. Contribution of astrocyte-derived IL-15 to CD8 T cell effector functions in multiple sclerosis. J Immunol. 2010;185:5693-5703 Crossref
  • [21] M. Rentzos, C. Nikolaou, A. Rombos, M.E. Evangelopoulos, A. Dimitrakopoulos, E. Kararizou, et al. Circulating interleukin-15 and RANTES chemokine in MS patients: effect of treatment with methylprednisolone in patients with relapse. Neurol Res. 2010;32:684-689 Crossref
  • [22] D. Diogo, F. Kurreeman, E.A. Stahl, K.P. Liao, N. Gupta, J.D. Greenberg, et al. Rare, low-frequency, and common variants in the protein-coding sequence of biological candidate genes from GWASs contribute to risk of rheumatoid arthritis. Am J Hum Genet. 2013;92:15-27 Crossref
  • [23] L. Akhabir, A.J. Sandford. Genome-wide association studies for discovery of genes involved in asthma. Respirology. 2011;16:396-406 Crossref

Footnotes

a Menzies Institute for Medical Research, University of Tasmania, Australia

b School of Pharmacy, University of Tasmania, Australia

c School of Medicine, University of Tasmania, Australia

d Murdoch Children's Research Institute, University of Melbourne, Australia

lowast Corresponding author at: Menzies Research Institute Tasmania, University of Tasmania, Hobart, TAS, Australia. Tel.: + 61 3 6226 4718; fax: + 61 3 6226 7704.