Discussion
EBV-HLH is a typical cytokine storm syndrome, and cytokines play a crucial role in the pathogenesis of EBV-HLH. Wada et al [8]. observed that EBV-infected patients had different serum cytokine levels and different severities of clinical manifestations. The synthesis, secretion, and biological effects of cytokines are regulated by their genes, and many studies have demonstrated important roles of cytokine gene polymorphisms in regulating cytokine levels and influencing the development of diseases[9-11]. SNPs are the most common form of gene polymorphisms, and we therefore hypothesized that cytokine SNPs might play an important role in the pathogenesis of and susceptibility to EBV-HLH.
Soluble IL-2 receptor (sIL-2R, also known as sCD25) is a marker of T-lymphocyte activation and is elevated in conditions such as infection, trauma, malignant hematological disorders, and autoimmune diseases. sIL-2R is currently used as an indicator of the severity of several diseases, and its level was shown to be directly proportional to the degree of inflammation[12]. Excessive activation of CD8+ T cells and significantly increased serum levels of sIL-2R have made sIL-2R an important diagnostic and disease marker for HLH in patients with EBV. The HLH-2004 diagnostic guidelines even added serum sIL-2R level ≥2400 U/ml as a new diagnostic criterion for HLH[13]. Furthermore, SNPs of the IL2RAgene encoding sIL-2R are associated with the development of autoimmune diseases, such as diabetes, multiple sclerosis, and baldness[14-16].
In the current study, we investigated the three most common IL2RAloci, rs2104286, rs12722489, and rs11594656, and found that the rs2104286 AA genotype and A allele frequencies were significantly higher in children with EBV-HLH than in controls, suggesting that these might be risk factors for the development of EBV-HLH in children. rs2104286 is located in the first intron of the IL2RA gene, which is not a coding or regulatory region; however, this SNP has been associated with several autoimmune diseases, including multiple sclerosis and rheumatoid arthritis[17]. Carriers of the rs2104286 A allele have elevated sIL-2R levels, diminished IL-2R signaling, increased granulocyte-macrophage colony-stimulating factor production by memory CD4+ T cells, increased frequency of CD25+ naïve T cells, and decreased CD25 expression on the surface of memory CD127+CD25+ and Treg cells[15, 18]. Moreover, the rs2104286 polymorphism can affect the activity of enhancer elements in the first intron of the IL2RA gene and the binding affinity of the transcription factor TRAF4, as well as mRNA processing and half-life, to influence IL2RA expression[17]. Dendrouet al. [19] found that the rs2104286 AA genotype upregulated CD25 (IL-2Rα) expression levels on CD4+ naïve T cells and CD14+CD16+ monocytes, thereby enhancing T-cell activation, and this genotypic effect persisted after T cell activation, leading to a high proportion of CD69+CD4+ naïve T cells that upregulated CD25 in donors carrying the susceptible rs2104286 allele. We thus deduced that the rs2104286 polymorphism might affect the pathogenesis of EBV-HLH by upregulating CD25 expression on CD4+ naïve T cells, thus increasing sIL-2R levels.
IL-10 is produced by Th2 cells and is an important anti-inflammatory factor. Cellular secretion of IL-10 is significantly increased during microbial infections and autoimmune diseases. It is involved in the development of HLH by modulating immunity and suppressing the inflammatory response. Significantly elevated serum IL-10 levels are a characteristic cytokine pattern of HLH[20]. Sánchez et al. [21] reported that certain SNPs in the promoter region of the IL-10 gene might influence IL-10 production and response intensity, and were associated with disease development. IL-10 SNPs were shown to be associated with various diseases, such as sepsis, asthma, multiple sclerosis, ankylosing spondylitis, gastric cancer, and coronary heart disease[4, 21, 22]. In the current study, we selected rs1800896, rs1800871, and rs1800872 for testing and showed that the frequencies of the AA genotype and A allele of rs1800896 and the CC genotype and C allele of rs1800872 were significant higher in children with EBV-HLH than in controls, suggesting that these might be predisposing risk factors for the development of EBV-HLH in children. Wang et al .[9] studied the relationship between the rs1800872 and children with EBV-HLH, and similarly showed that the rs1800872 CC genotype and C allele frequencies were higher in children with EBV-HLH than in patients with infectious mononucleosis (IM) and in healthy controls. rs1800896 is located at a putative ETS-like transcription factor-binding site and rs1800872 at a putative signal transducer and activator of transcription 3-binding site and regions of negative regulatory function[23, 24]. Polymorphisms at these sites may thus influence transcription factor binding and consequent IL-10 transcription and IL-10 serum levels, thus affecting disease susceptibility[25, 26]. IL-10 levels are significantly elevated in children with EBV-HLH. Numerous studies have shown that the C allele of rs1800872 is associated with high levels of [27, 28], while the relationships between the G and A alleles of rs1800896 and IL-10 levels remain controversial. Rees et al. [24, 29]showed that the rs1800896 A allele enhanced transcription of theIL-10 gene and resulted in high levels of IL-10. In contrast, Engelhardt et al. [30-32] showed that the rs1800896 G allele was associated with higher IL-10 production, while the rs1800896 A allele resulted in lower IL-10 production. The frequencies of IL-10 polymorphisms differ among populations[33], and IL-10 plays different roles in different diseases, although its regulatory mechanisms are not fully understood. Numerous genetic correlation studies and a few functional studies have shown that the regulation of IL-10 is complex and multifactorial. Further studies are therefore needed to determine how rs1800896 and rs1800872 regulate the production of IL-10 and affect the susceptibility to and clinical course of EBV-HLH in children. In addition, the rs1800896 polymorphism can also influence EBV infection status. Consistent with the results of the present study, Helminenet al. [34] also showed that individuals carrying the rs1800896 A allele were more susceptible to severe EBV infection. The rs1800896 and rs1800872 polymorphisms may thus influence EBV-HLH susceptibility by synergistically regulating IL-10 expression and EBV infection status.
Haplotype analysis provides more accurate genetic information than analysis of individual SNPs. Comparisons between children with EBV-HLH and controls based on estimated haplotype frequencies thus provide a better understanding of the role of cytokine SNPs in EBV-HLH susceptibility. The present results showed that the frequency of theIL2RA AGT (rs2104286-rs12722489-rs11594656) haplotype was significantly increased while that of the GGT haplotype was significantly decreased in children with EBV-HLH. The frequency of theIL-10 ACC (rs1800896-rs1800871-rs1800872) haplotype was also significantly increased, whereas the frequencies of the ACA and GTA haplotypes were significantly decreased. In addition, some of these haplotypes showed strong linkage disequilibrium, suggesting thatIL2RA and IL-10 haplotype association and linkage disequilibrium may play important roles in the susceptibility to EBV-HLH. The results of this study suggest that the IL2RA AGT haplotype may increase the risk of EBV-HLH; however, no other studies have simultaneously tested the IL2RA rs2104286, rs12722489, and rs11594656 loci, and studies of the AGT haplotype in other diseases are lacking, and further studies are therefore needed to validate the roles of the AGT and GGT haplotypes. The common IL-10rs1800896-rs1800871-rs1800872 haplotypes are GCC, ACC, and ATA. The ACC haplotype is associated with moderate IL-10 production[35], but its role in different diseases is controversial. Marangon et al. [36]showed that this haplotype was a protective factor in diffuse large B-cell lymphoma, while Gao et al. [28]showed that it was associated with an increased risk of IgA nephropathy.
We also analyzed IFN-γ rs2430561, IRF5 rs2004640, andCCR2 rs2004640, but found no significant associations between these SNPs and susceptibility to EBV-HLH. IFN-γ is an important cytokine produced by a variety of immune cells in response to inflammatory stimuli. It regulates the body’s immune response, and is significantly elevated in children with EBV-HLH. IFN-γ rs2430561 has been shown to affect susceptibility to diseases such as systemic lupus erythematosus, sepsis, ankylosing spondylitis, and breast cancer[37-40], but no previous association between this SNP and HLH has been reported. IRF5 is a member of the IRF transcription factor family that plays important roles in both the interferon pathway and the Toll-like receptor signaling pathway, and is closely associated with many autoimmune diseases[41, 42]. IRF5 rs2004640 has been associated with susceptibility to diseases such as systemic juvenile idiopathic arthritis(sJIA), systemic lupus erythematosus(SLE), and systemic sclerosis(SS)[43, 44]. Yanagimachi et al. [42] investigated the relationship between rs2004640 and susceptibility to secondary HLH, such as macrophage activation syndrome and EBV-HLH, and found that the GT/TT genotype at this locus increased overall susceptibility to secondary HLH, but not to EBV-HLH, consistent with the results of the present study. The chemokine receptor CCR2 plays a key role in several diseases, and the CCR2rs1799864 gene polymorphism has also been associated with susceptibility to diseases such as ischemic stroke and multiple sclerosis[45, 46]. Ou et al .[47] found no significant difference in the rs1799864 polymorphism between children with and without HLH, suggesting that this locus may be independent of HLH susceptibility, consistent with the current study. However, the paucity of relevant studies and the limited numbers of cases mean that further studies are needed to confirm the relationships between the above genetic polymorphisms and EBV-HLH.
This study had some limitations. First, the relatively small sample size may have limited the statistical power, and more multicenter studies with larger sample sizes are therefore needed to validate our findings. Second, because of technical limitations in the quantitative analysis of cytokines before 2014, cytokine profiles were not available for more than half of the patients. Unfortunately, we were therefore unable to analyze the relationships among SNPs, cytokine secretion, and HLH.