Human interleukin 10 (hIL-10) is a pleiotropic cytokine with multifaceted functions in the regulation of immune and inflammatory responses 1. It limits cell-mediated immune responses by inhibiting the activation of macrophages, monocytes, and dendritic cells. In B cells, hIL-10 promotes cell proliferation and differentiation, and also plays some anti-apoptotic function. These biological activities of hIL-10 are advantageous for viruses that infect and persist in B cells. Indeed, Epstein-Barr virus (EBV), as well as some other members of the herpes virus family, appears to use mimicry of hIL-10 functions as part of their survival strategies in the host. Specifically, an open reading frame of EBV, BCRF1, shows a strikingly high degree of sequence homology with hIL-10 (84% homology in amino acid sequence) and viral IL-10 shares many of hIL-10's biological activities 1. Based on these observations, we hypothesize that genetic variations in the host's hIL-10 gene may modulate the host-organism interaction with this ubiquitous virus encoding an hIL-10 mimic.

This hypothesis is of particular interest as both the risk of multiple sclerosis (MS) and the cytosine-allele frequency at position -819 in the promoter region of hIL-10 (hIL-10 T-819C) are considerably higher in Western, compared to Chinese and Japanese populations (MS prevalence of < 5 per 100,000 in China and Japan 2 vs. > 50 per 100,000 in the US 3 and the cytosine-allele frequency at hIL-10 T-819C of 28.8–32.7% in Chinese and Japanese 4567 vs. 71.8–83.6% in Caucasians 789101112). Furthermore, there is a strong positive association between MS risk and elevated pre-diagnosis IgG antibody titers against Epstein-Barr viral capsid antigen (anti-EBVCA IgG titers) 131415161718192021 although an inverse association has been observed in one prospective study 22.

To test the hypothesis, anti-EB-VCA IgG titers were measured as an indicator of the host-virus interaction, and compared across the genotypes of the hIL-10 promoter single nucleotide polymorphism (SNP) at hIL-10 T-819C.

Table 1 shows the geometric mean and standard deviation of anti-EB-VCA IgG titers by the hIL-10 T-819C genotype, stratified by sex. The geometric mean titer increased monotonically with the number of cytosine alleles among women, but not among men. In the 123 women, the proportion with elevated anti-EB-VCA IgG titers increased from 53.7% in the T/T group to 66.7% in the T/C group and 83.3% in the C/C groups (P-trend = 0.037) (Table 2). This was not seen among the 118 men (P-trend = 0.84). The difference of the trend between the sexes was marginally statistically significant (P = 0.097).

In this study, we found a monotonic increase in anti-EB-VCA IgG titers with the number of cytosine alleles at the hIL-10 T-819C locus in Japanese women, however not in men. Considering the higher prevalence of both MS 23 and the specific SNP in the promoter of hIL-10T-819C 456789101112 in Western populations compared to Asian populations, and the strong positive association of MS risk with anti-EB-VCA IgG titers 131415161718192021, our finding has an important implication on MS susceptibility. Specifically, it suggests a genetic predisposition to MS in the hIL-10 gene promoter, relevant only in women.

Such female-specific MS-risk-elevating function is consistent with the notion that gonadal steroids regulate the gender dimorphism in MS and other autoimmune diseases 27. Indeed, in the mouse model of MS, experimental autoimmune encephalomyelitis in SJL mice, disease severity is increased by castration in male mice 28 and decreased by testosterone implantation in females 29. The immunomodulatory effects of gonadal steroids, especially, testosterone, may underlie the female-specific association of anti-EB-VCA IgG titers with hIL-10 T-819C observed here.

Studies that have investigated human MS risk in relation to polymorphisms in hIL-10 have found no association 3031323334353637. These studies, however, investigated Western high-risk populations where the cytosine-allele frequencies at hIL-10 T-819C are very high (> 70%). They did not investigate, with a few exceptions, the polymorphism at hIL-10 T-819C whose genotype distributions are markedly different between the high- and low-risk populations of MS, and did not stratify the association analysis by gender. It is also a possibility that the T-819C polymorphism and the associated IL-10 levels may modulate the immune response and MS risk, given some observed correlation between the promoter polymorphism and the IL-10 levels in in vitro/in vivo 38. It would be of interest to evaluate MS risk in relation to the hIL-10 T-819C genotype: (1) in populations where thymine-allele frequencies at hIL-10 T-819C are sufficiently high (e.g., Chinese and Japanese); and perhaps more importantly, (2) in any population stratified by gender.

Our findings are, however, consistent with a recent study of a small patient cohort showing a statistically significant increase in CC genotype of hIL-10 SNPs at -819 and -592 loci in MS patients 39. The hIL-10 promoter polymorphism at the -819 locus is often analyzed along with the hIL-10 promoter polymorphisms at the -1082 and -592, which is justified by their strong linkage disequilibrium, in addition to the three main haplotypes (GCC, ACC, ATA, and very rarely GTA, at the -1082, -819, and -592 locus) found to segregate in most populations. Because we did not study the -1082 locus, we were unable to separate the GCC and ACC haplotypes. However, G alleles at the -1082 locus are very rare (G-allele frequency < 4%) in Japanese and Chinese populations 4567.

Given the high homology of viral IL-10 to hIL-10, the established association of MS with elevated anti-EB-VCA IgG titers, and the public health/personal burden of MS, our finding of female-specific increase in anti-EB-VCA IgG titers with the number of cytosine alleles at hIL-10 T-819C warrants further investigation. Such investigations may lead to an explanation of high MS risk in Western populations, where the MS disease burden is more than 10 times higher than that in Japan and China, and the cytosine-allele frequencies at hIL-10 T-819C are at least twice as high (> 70% vs. 30%) as those in Chinese and Japanese populations.

The author(s) declare that they have no competing interests.

NH, TN, KT designed and conducted the epidemiologic study on which this study was based. The original idea of this study was conceived by NH and YY. NH performed the polymorphism analysis. Data analysis was performed by YY. The paper was drafted by YY, revised by JDP and NSED, finalized and approved by all authors.