Introduction
IL-18 is produced by a wide range of immune cells, such as monocytes, macrophages, and dendritic cells. Initially thought to be an IFN-γ-inducing factor of T cells and natural killer cells [
1,
2], IL-18 has been found to have multiple biological functions. Interestingly, IL-18 is a unique cytokine that stimulates both T helper (Th)1- and Th2-type immune responses, depending on its cytokine milieu [
3]. In combination with IL-12, IL-18 induces IFN-γ production in Th1 cells, B cells, and natural killer cells, promoting Th1-type immune responses [
4]. When cultured with IL-2, however, IL-18 induces Th2 lineage in CD4+ T cells [
5]. In basophils and mast cells, IL-18 together with IL-3 also induces production of Th2 cytokines [
6].
In vivo, IL-18 regulates innate and acquired immune responses, controlling either Th1 or Th2 cytokine balance. To date, inappropriate IL-18 production has been reported both in chronic inflammatory diseases such as Crohn's disease [
7] and rheumatoid arthritis (RA) [
8,
9], and in allergic diseases such as bronchial asthma and atopic dermatitis [
10].
We previously reported that serum IL-18 levels were significantly elevated and well correlated with disease activity and severity [
11] in adult-onset Still's diseases (AOSD), which is characterized by high spiking fever, polyarthralgia, evanescent salmon-colored rash, liver dysfunction, splenomegaly and hyper ferritinemia [
12]. Although serum levels of inflammatory cytokines are generally high in AOSD patients [
13], the levels of IL-18 were enormously high, reaching more than 1,000 times the levels found in normal controls and other chronic inflammatory diseases such as RA [
11,
14]. Therefore, we have suggested that IL-18 is closely involved in the pathogenesis of AOSD.
Juvenile idiopathic arthritis (JIA) is a chronic arthritis of childhood and belongs to a group of clinically heterogeneous disorders including oligoarthritis, polyarthritis, systemic arthritis, secondary arthritis, and other arthritis types. According to the International League of Associations for Rheumatology (ILAR) classification criteria, oligoarthritis is further subdivided into persistent and extended oligoarthritis, and polyarthritis is further subdivided into rheumatoid factor-positive and negative polyarthritis. Thus, JIA is categorized into seven clinically distinct presentations [
15], and the pathogenesis differs among the subgroups. Systemic JIA is characterized by systemic involvement, such as high spiking fever, skin rash, serositis and hepatosplenomegaly, and overproduction of inflammatory cytokines [
16,
17]. Maeno and colleagues [
18] reported the serum levels of IL-18 were strikingly high in systemic JIA compared with other JIA subgroups and other childhood inflammatory disorders. Among systemic JIA patients, individuals with hepatosplenomegaly or serositis showed higher serum IL-18 levels than those without such manifestations [
18]. In addition to similar clinical findings, aberrant production of IL-18 is a characteristic feature in both AOSD and systemic JIA, which is the reason why many investigators may consider these two diseases to be the same entity.
The human
IL18 gene is located on chromosome 11q22.2 – q22.3 [
19] and is composed of six exons; the translation-starting site is present in exon 2 [
20]. It lacks a TATA box, and its expression is regulated by at least two distinct promoter regions, one of which is located upstream of untranslated exon 1 (promoter 1) [
20‐
24], and the other upstream of exon 2 (promoter 2) [
23]. Promoter 1 is up-regulated by various stimulants in macrophages [
20], intestinal epithelial cells [
21], and epithelial-like cell lines [
24]. A 108 base-pair (bp) 5' flanking region upstream of exon 1 contains a
PU.1 (purine-rich sequence) consensus-binding site and a GC-rich region; this region is critical for the sodium butyrate-stimulated promoter 1 activity [
21,
25]. Promoter 2 is considered to act constitutively [
23], but regulation of human
IL18 gene expression has not been fully examined.
We speculated that genetic polymorphisms within the promoter region of the
IL18 gene might contribute to the high IL-18 production in AOSD. Thus, in the previous study, we performed a systematic search for polymorphisms in a 6.7 kb sequence, including a putative promoter region of the
IL18 gene, and then identified 10 single nucleotide polymorphisms (SNPs) and a single 9 bp insertion [
26]. The region had been reported to be upstream of exon 2 (intron 1) [
24]. Later, Kalina and colleagues [
20] determined that the 6.7 kb sequence was upstream of exon 1 instead of exon 2 by using 5' rapid amplification of cDNA ends; thus, this 6.7 kb region includes promoter 1 of the
IL18 gene. We identified some of these polymorphisms as components of haplotypes and found there were three major haplotypes in the Japanese population (S01, S02, and S03). Furthermore, the frequency of haplotype S01, especially the diplotype configuration of S01/S01, was significantly higher in AOSD than in RA as well as in normal controls [
26].
In the present study, we conducted a case-control study to evaluate possible associations of haplotype S01 of the IL18 gene with susceptibility to JIA, especially to systemic JIA.
Discussion
The present study has demonstrated a strong association between the haplotype S01 of the
IL18 gene and JIA as well as AOSD and RA in the Japanese population. Initially, we speculated that haplotype S01 is specific to systemic JIA because of previous studies that showed enormously high IL-18 production both in AOSD and systemic JIA [
11,
14,
18] and genetic skewing of the haplotype S01 in AOSD [
26]. However, haplotype S01 was associated with all the subgroups of JIA in the Japanese population, and we did not find the expected statistically significant correlation of haplotype S01 with susceptibility to systemic JIA. Furthermore, haplotype S01 was found to be associated with the whole group of arthritis diseases, including JIA, ASOD and RA.
Genetic polymorphisms of the human
IL18 gene have been associated with a wide variety of diseases, including allergic diseases and inflammatory diseases [
32,
34‐
36]. These observations might reflect the redundancy of the IL-18 protein, which has multiple biological functions promoting both Th1- and Th2-type immune responses. Previously reported polymorphic sites concentrated on two promoter regions and the untranslated exon 1 of the
IL18 gene. Most of the investigators have compared the frequency of each allele in the disease population with that of a healthy population. In a previous study [
26], we demonstrated that all of these genetic polymorphisms comprised a haplotype. We therefore checked the genotypic data reported by other investigators and estimated the carrying rate of each haplotype in the disease populations. Interestingly, we found differently genetically skewed haplotypes in each disorder. For example, in the white population of Germany, patients with atopic phenotypes were predisposed to have -137 C in promoter 1, +113 G in exon 1, +127 T in exon 1, and -133 G in promoter 2 [
32], and each allele corresponded to the minor allele of SNPs 11, 12, 13, and 15, respectively. All of these were the components of haplotype S03. Other investigators reported that SNPs specific for haplotype S03 are linked to susceptibility to atopic eczema in Germany [
34], and asthma in Japanese [
35]. It was likely that haplotype S03 of the
IL18 gene was associated with atopic disorders, typical of Th2-dominant disease. On the other hand, Japanese patients with sarcoidosis were reported to be associated with polymorphisms specific for haplotype S02 [
36]. The present study shows that haplotype S01 is associated with JIA as well as AOSD and RA in the Japanese population, with each having arthritis as a phenotype. In view of imbalanced immune responses, these diseases may be considered as Th1-dominant disorders [
37‐
39]. By analyzing haplotypes, it became clear that haplotypes S01 and S03 of the
IL18 gene are involved in the susceptibility of quite different types of disorders.
It is important to note there is an essential racial difference in the distribution of haplotypes. According to our data, the frequency of haplotype S01 of the
IL18 gene is 37.9% and that of the S01/S01 diplotype configuration is 13.9% in the Japanese population. We also estimated the frequency of haplotypes and/or the diplotype configurations in the different ethnic groups of previous reports. The frequency of the S01/S01 diplotype configuration was considered to be only 1% in the healthy Swedish population [
22] and only 0.05% in the Chinese population [
38], much lower than that of the Japanese population. Also, in German whites, the frequency of haplotype S01 was 13% [
40]. Interestingly, the existence of the fourth major haplotype, which we tentatively call haplotype S04, was reported in the Polish population [
41]. The haplotype S04 is composed of combinations of nucleotides of SNP 10 C and SNP 11 C [
41]; we have never observed this haplotype in the Japanese population. As described, the carrying rate of each haplotype is quite different among ethnic groups, which would be one reason why a disease-associated haplotype is not necessarily common among them. Indeed, no increase in the frequency of the S01/S01 diplotype configuration was observed in RA patients in Chinese patients [
42] and in JIA patients in Germany [
40]. One reason why haplotype S01 of the
IL18 gene is skewed in Japanese arthritis patients is due to a basal high frequency of the haplotype S01 in the Japanese population.
The findings in the present study raise the question of how these genetic polymorphisms within the human
IL18 gene influence the phenotype. One consideration is that nucleotide substitution directly influences the transcription of the gene. SNP11 is located at the GATA3 binding site, which is involved strongly in Th2 differentiation [
43]. The C allele of SNP11 is specific to the haplotype S03, which might be a good explanation for the association of haplotype S03 and atopic disorders. As for haplotype S01, there have been few good explanations as to why it is linked to extraordinarily high IL-18 production. Nucleotide substitutions specific for haplotype S01 (SNP2, 4, 5 and 14) do not create known nucleotide factor-binding sequences; however, it is still possible that undefined genetic polymorphisms in linkage disequilibrium with haplotype S01 exist in other regions of the
IL18 gene and influence the expression of IL-18 protein.
The aim of the present study was to examine whether haplotype S01 of the
IL18 gene was skewed in systemic JIA. The frequency of the S01/S01 diplotype configuration seemed to be higher in patients with systemic JIA than in healthy controls (23.5% versus 13.9%), but the difference was not statistically significant, probably because of the relatively small study populations. Although we could not prove haplotype S01 was associated with disease susceptibility, the haplotype was linked to some clinical features. The patients carrying the S01/S01 diplotype configuration showed significantly higher IL-18 protein levels than those without it (
P = 0.017). Thus, it is possible that haplotype S01 of the
IL18 gene is linked to the phenotype of high IL-18 production. Because serum IL-18 levels indicate disease severity [
18], the haplotype S01 might be a marker of severe disease condition in systemic JIA. In addition, although the differences were not statistically significant, the patients with the S01/S01 diplotype configuration tended to be younger at disease onset, need immunosuppressive drugs in addition to high doses of corticosteroids to control severe disease activity, and have the complication macrophage-activating syndrome (data not shown). Similarly, in AOSD, the S01/S01 diplotype configuration correlated with disease severity (unpublished data).
The mechanism of induction of enormously high IL-18 protein production in AOSD and systemic JIA is unclear, but it is likely that infectious agents such as viruses trigger the activation of macrophages and induce IL-18 production, and that some immunological breakdowns fail to control sustained activation of macrophages and consequently cause continuous IL-18 production. Several candidate genes would be involved in these phenomena besides the
IL18 gene itself. They might include factors regulating IL-18 mRNA transcription or translation and caspase-1, which cleaves pro-IL-18 into a mature form [
44]. We suggest that haplotype S01 of the
IL18 gene might influence IL-18 protein production via unknown mechanisms, which explains in part the enormously high IL-18 production in systemic JIA. Furthermore, homozygosity for S01 would be more important for susceptibility to JIA as well as AOSD and RA than only carrying the S01 haplotype, since the
P value for the former was smaller than that of the latter.
The present study also shows that haplotype S01 of the
IL18 gene is widely linked to susceptibility to arthritis in the Japanese population. As described above, IL-18 is a key cytokine involved in the pathogenesis of both AOSD and systemic JIA [
11,
14,
18]. High IL-18 protein could induce liver injury [
45] or arthritis [
8]. In RA, several lines of evidences suggest that IL-18 plays a role in the pathogenesis because IL-18 is up-regulated and induces production of inflammatory cytokines such as tumor necrosis factor-alpha in synovium [
8,
9]. In collagen-induced arthritis, IL-18 promotes arthritis via tumor necrosis factor-alpha induction [
46]. In another study, IL-18 knockout mice showed a reduced degree of inflammation, and the administration of recombinant IL-18 reversed collagen-induced arthritis [
47]. In these disorders, including JIA, AOSD and RA, overproduction of IL-18, probably together with IL-12, would shift the immune response to the Th1 lineage. Indeed, peripheral blood obtained from AOSD patients showed more IFN-γ producing Th cells and a higher Th1/Th2 ratio than in healthy controls [
37]. Th1 cytokine expression has been demonstrated in the synovium of RA and JIA [
38,
39]. As described, IL-18 would be involved in the pathogenesis of arthritis directly or via production of other cytokines, and probably would promote a Th1-type immune response. Among three haplotypes of the
IL18 gene, haplotype S01 might contribute genetically to the development of arthritis in the Japanese population. This is supported by the observation that the frequency of the S01/S01 diplotype configuration was significantly higher in arthritis patients as a whole, including JIA, AOSD and RA, than normal controls (
P = 0.0013, OR = 2.36, 95% CI = 1.36–4.12) or patients with collagen diseases (
P = 0.0069, OR = 2.39, 95% CI = 1.22–4.75, data not shown).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
TS conceived the study and drafted the manuscript. NM was responsible for the recruitment and classification of patients, performed ELISA, and helped to draft the manuscript. Y Kawaguchi participated in the design and coordination of the study, and recruited a subset of the patients. ST conceived the study together with Y Kawano and participated in the design of the study. HI and HTI recruited a subset of patients. MH recruited a subset of patients and participated in coordination of the study. NK participated in the design and coordination of the study. All authors read and approved the final manuscript.