Introduction

Kawasaki disease (KD) is one of the common childhood vasculitides and is being increasingly recognized as an important cause of acquired heart disease in young children.1 The highest incidence of KD has been reported from Japan, Korea, and Taiwan.2,3,4 Reports from several centers in India over the past two decades suggest that the incidence of KD may be increasing.2,5,6 Two hospital-based studies from Chandigarh, India showed that KD is the most common cause of vasculitis in young children.7,8,9 One of the main concerns in KD is its propensity to affect the heart and cardiovascular system and this can result in the development of coronary artery lesions (CALs).10,11,12 Treatment with IVIG not only reduces the occurrence of CAL but also decreases myocardial inflammation.13,14

Although therapy of KD has evolved over the past three decades,1,15,16 the etiology of this disease remains a mystery. Many etiological associations have been proposed for KD, but the cause and effect relationship has not yet been established.17 Genetic association of KD has been suspected because of the high incidence of the disease in children in families of Japanese ancestry and siblings or offsprings of patients with KD.18,19 Inositol 1,4,5 triphosphate 3-kinase C (ITPKC) is a negative regulator of Ca2+/nuclear factor of activated T cell (NFAT) pathway in T cells.20 Onouchi et al. reported the association of polymorphism of rs28493229 of ITPKC with susceptibility to KD and CAL in Japanese and American KD cohorts.21 This finding was, however, not replicated in studies from other countries like China and Taiwan.22,23 Similarly, while Peng et al. reported an association of polymorphism of rs2290692 of ITPKC with susceptibility to KD in China,22 the finding was not replicated in studies from Taiwan and Korea.24,25 There is no study on SNPs of ITPKC in association with KD from India to date. As the phenotype of KD in India is different, it is important to study the genotype of KD in Indian children.26,27,28 In this study, we have studied the SNPs rs28493229 at intron 1 and rs2290692 at 3′-UTR (untranslated region) of the ITPKC gene in North Indian KD patients (with or without CAL) and compared the results with healthy controls.

Methods

Study population

This study was carried out in Pediatric Allergy Immunology Unit, Advanced Pediatrics Centre, Postgraduate Institute of Medical Education and Research, Chandigarh, India. Our institute serves as a not-for-profit tertiary care referral center for North India. Patients and controls were of North Indian ethnicity. Fifty cases (25 KD patients without CAL and 25 KD patients with CALs) and 50 age- and sex-matched controls were selected for sampling over a period of 14 months (January 2018–April 2019). The mean (SD) age of children with KD was 3.9 (2.53) years, whereas the mean (SD) age of controls was 3.92 (2.46) years. There were 30 (60%) boys and 20 (40%) girls (M:F = 1.5:1) in both the KD group and control group.

Diagnosis of KD was based on American Heart Association (AHA) guidelines.1,29 A CAL was defined as any coronary artery abnormality and included ectasia, dilatation (with the coronary z-score value of ≥2 but <2.5), and aneurysms (with the coronary z-score value of ≥2.5).1 Written informed consent was taken from all the subjects for the purposes of this study. The study protocol was approved by Institute Ethics Committee and Institute Thesis Committee. The manuscript has been approved by Departmental Review Board. Two-dimensional echocardiographic coronary assessment of patients was carried out in the febrile phase and ~4–6 weeks later on follow-up.

Genetic study

Peripheral venous blood sample (2–3 ml) was drawn from cases and control subjects in ethylenediaminetetraacetic acid (EDTA) vacutainer.

Genomic DNA was extracted from peripheral blood using QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany). DNA quality was determined using 1% agarose gel electrophoresis, followed by staining with ethidium bromide. The purity of DNA was determined by taking optical density (OD) of samples at 260 and 280 nm using TECAN infinite M200 Pro with Nanoquant plate (Tecan group (Life-sciences and Diagnostics) AG, Switzerland) and was stored at −80 °C till further analysis.

Genotyping for SNPs rs28493229 and rs2290692 was carried out using PCR and restriction fragment length polymorphism (RFLP). 3′-UTR and intron 1 of ITPKC gene were amplified using PCR at controlled conditions using specific oligonucleotide primers as previously used by Peng et al.22 PCR primers (Integrated DNA Technologies, Lowa), restriction enzymes (BanI and AvaI) (New England Biolabs Inc., Massachusetts), length of the PCR products, and the digested fragments are shown in Table 1A. Genotyping was performed by the PCR-RFLP method. Primer sequence, restriction enzymes, and band size details have been mentioned in Table 1A. PCR cycle conditions for amplification are available on request.

Table 1 (A) Primers and restriction enzymes for concerned SNPs. (B) Review of published literature on SNP rs28493229 in association with susceptibility to KD and CAL. (C) Review of published literature on SNP rs2290692 in association with susceptibility to KD and CAL.
Full size table

Representative gel pictures showing the genotypes of both the SNPs have been given in Fig. 1a, b. PCR products of samples with the presence of C allele and G allele were confirmed with bidirectional Sanger sequencing (Applied Biosystems, Foster, CA) (Fig. 1c, d). Results were reproducible with no discrepancy in genotyping.

Fig. 1: Representative pictures of gel electrophoresis and sanger sequencing.
figure 1

a Gel electrophoresis for confirmation of CG genotypes with restriction fragments highlighted as bands of 36, 60, 141 and 177 bp in digested column and band of 237 bp in an undigested column of rs28493229 in intron 1 of ITPKC. b Gel electrophoresis picture showing bands of restriction fragments (digested and undigested columns) of rs2290692 in 3′-UTR of ITPKC. c Sanger sequencing of control and representative cases of different genotypes of rs28493229 at intron 1 of ITPKC. d Sanger sequencing results of control and representative cases of rs2290692 at 3′-UTR of ITPKC.

Full size image

Meta-analysis

In addition, all previous studies on SNPs rs28493229 and rs2290692 of the ITPKC gene in association with susceptibility and coronary complication of KD were gathered. A meta-analysis was performed for genotypes of both SNPs and their association with susceptibility to KD.

Statistical analysis

Allele and genotype frequencies of SNPs were compared between patients with KD and control subjects. Similar comparisons were also carried out between patients of KD with and without CALs. The association of categorical variables with KD patients was analyzed using χ2 test/Fisher’s exact tests, whereas comparisons of quantitative variables between two study groups were carried out using an independent-sample t test (parametric) or Mann–Whitney U test (nonparametric). A p value of ≤0.05 was taken as significant. Data analysis was done using R software version 5.3.0 with RStudio IDE (The R Foundation) in Windows 10 platform. Deviation of SNPs from Hardy–Weinberg law was evaluated with Stata IC version 14 (StataCorp LLC, TX) using “genass” package. Meta-analyses were performed with R and Stata software.

Results

Allele and genotype frequencies

Genotype distribution of SNPs rs28493229 (Pearson’s χ2 0.378, p = 0.54) and rs2290692 (χ2 = 0.263, p = 0.76) were in Hardy–Weinberg equilibrium. We analyzed the association of these two SNPs with KD and control groups. G and C allele frequency of rs28493229 observed among cases was 89% and 11%, respectively. Among controls, 92% G alleles and 8% C alleles were found (Table 2A). GG and CG genotype of this SNP was found to be 39 (78%) and 11 (22%) among cases. We could not get any case or control with the CC genotype of rs28493229 in our study. Similarly, the allele frequency of rs2290692 at 3′-UTR among cases was 63% and 37% for the G and C allele, respectively. Nineteen GG (38%), 25 CG (50%), and 6 CC (12%) genotypes of SNP rs2290692 were found in cases, whereas among controls, the frequencies were 20 (40%), 16 (32%), and 14 (28%), respectively. No statistical differences were found in allele, carrier, and genotype frequencies of both SNPs between cases and controls (Table 2A).

Table 2 (A) Allele, genotype, and carrier frequencies of single-nucleotide polymorphisms rs28493229 and rs2290692 in test subjects and controls. (B) Frequencies of allele and genotype of rs28493229 and rs2290692 SNPs and results of logistic regression analysis in KD patients with CAL and those without CAL. (C) Results of univariate as well as multivariate logistic regression analysis of combined genotypes with the dependent outcome as the occurrence of KD.
Full size table

Association between ITPKC polymorphism and occurrence of KD

On multivariate logistic regression, adjusted odds of KD occurrence were increased by nearly one and half times (odds ratio (OR) 1.53; 95% confidence interval (CI): 0.54–4.65; p = 0.43) in subjects with CG genotype on rs28493229. Similarly, adjusted odds of KD occurrence were increased by 1.8 times (OR 1.8; 95% CI: 0.72–4.56; p = 0.21) in subjects with CG genotype on rs2290692. However, these associations failed to reach statistical significance (Table 2A). We observed an over-representation of CG genotype in rs28493229 as well as rs2290692 at 3′-UTR in patients with KD. Adjusted odds of occurrence of KD using SNP predictors are also demonstrated in a forest plot (Supplemental Fig. 1A).

Effects of combined genotypes for the risk of KD were also analyzed. On multivariate logistic regression, adjusted odds of occurrence of KD were increased by more than 4 times in subjects with any genotype with G allele (i.e., CG + GG) of rs2290692 (OR 4.14; 95% CI: 1.38–13.83; p = 0.015) after adjusting the confounding effects of sex and another SNP (Table 2C).

Association between ITPKC polymorphism and occurrence of coronary abnormality in KD

Allele, genotype, and carrier frequencies were analyzed between the KD patients with and without CALs. Effect size in genetic characteristics of both SNPs failed to reach statistical significance (Table 2B). Adjusted odds of occurrence of coronary abnormality in KD were lowered by 50% in subjects with CC genotype on rs2290692 at 3′-UTR (OR 0.50; 95% CI: 0.06–3.35; p = 0.49) after adjusting the confounding effects of sex and other SNPs (Supplemental Fig. 1B).

Association of ITPKC polymorphism and size of aneurysms

The z-scores for the size of aneurysms were divided into tertiles. The highest tertile scores represented higher z-scores. No significant association of SNPs was observed with the size of aneurysms. These effect sizes were adjusted for the age at which the illness occurred. Predictive margins of two SNPs with a probability of largest aneurysms (z-score) are illustrated in Supplemental Fig. 1C, D.

Meta-analyses

We further performed a meta-analysis for a possible association of both SNPs and susceptibility and complications of KD by combining the results of our study with all previous studies on rs28493229 and rs2290692 of the ITPKC gene. In the meta-analysis for the association of CC genotype of rs28493229, combined data confers a 1.5 times higher risk for this genotype to develop KD (OR = 1.46, 95% CI: 0.96–2.23) (Table 3A). Latter just crossed the line of null hypothesis showing only the trend for significance (Fig. 2a).

Table 3 The meta-analysis of the association of different genotypes of SNPs of the ITPKC gene with KD.
Full size table
Fig. 2: Forest plot for the association of different genotypes of rs28493229 and rs2290692 of ITPKC gene.
figure 2

a Forest plot for the association between CC genotype of rs28493229 and risk of having KD (estimates of odds ratio and its 95% CI are plotted with a box and horizontal line). b Forest plot for the association between GG genotype of rs2290692 and risk of having KD (estimates of odds AQ8ratio and its 95% CI are plotted with a box and horizontal line). c Forest plot for the association between CC genotype of rs2290692 and risk of having KD (estimates of odds ratio and its 95% CI are plotted with a box and horizontal line).

Full size image

Similarly, a meta-analysis was performed individually for the association of CC and GG genotypes of rs2290692 with the risk of having KD. In the meta-analysis for the association of GG genotype of rs2290692, combined effect showed that odds of occurrence of GG in KD was 21% lower than in controls in this meta-analysis with an OR of 0.7881 and its 95% CI of 0.6085–1.0205 (Table 3B). The combined average only showed a trend for significance (Fig. 2b).

In the meta-analysis of CC genotype of our study and all previous studies of SNP rs2290692, combined OR did not confer any significant association with the susceptibility of KD (OR = 1.07, 95% CI: 0.663–1.731) (Table 3C and Fig. 2c).

Discussion

KD is a common medium vessel vasculitis of young children. It may lead to complications like myocarditis, KD shock syndrome, and the development of CALs.10,14 Etiology of KD still remains an enigma and its pathogenesis is poorly understood.3,17,30 A high incidence of KD in some Asian populations, particularly Japanese and in native Japanese population residing in Hawaii, points towards some genetic link of KD.3,19 A study by Onouchi et al. in 2008 revealed a significant association of polymorphisms of rs28493229 of ITPKC gene with susceptibility and severity of KD in Japanese and American patients.21 Functional SNPs of ITPKC have been shown to result in altered messenger RNA (mRNA) splicing and gene transcription in patients with KD. Polymorphisms of ITPKC are reported as an important predisposing factor for KD.21,31 We studied the association of two SNPs of the ITPKC gene in patients with KD and healthy controls of North Indian ethnicity.

Although certain polymorphisms of the ITPKC gene are shown to be associated with susceptibility to KD from some ethnicities, the results have not been uniformly replicated across ethnicities (Table 1B, C).22,23,24,25,26,27,28,29,30,31,32 Such associations between SNP and KD need to be studied extensively in different ethnicities before coming to a conclusion. Moreover, the phenotype of KD in India is different.26,27,28 Contrary to the pattern of KD in Eastern Asian and North American children, proportionately greater number of male patients, higher incidence of older children (almost half of them of age around 5 or more than that), early appearance of periungual peeling (prior to day 10 of fever), and early appearance of thrombocytosis were observed in Indian cohorts of KD. This suggests the possibility of different genotypic associations of these SNPs in patients with KD from India.

In the present study, we performed PCR-RFLP and representative bidirectional Sanger sequencing for two SNPs (rs28493229 at intron 1 and rs2290692 at 3′-UTR) of the ITPKC gene in North Indian children with KD. Diagnosis of KD was made on the basis of AHA guidelines.1,29 According to the 1000 Genome Project, the prevalence of C and G allele of rs28493229 in the South Asian population is 13% and 87%, respectively.33 In our study, C and G Allele frequency of rs28493229 in the KD group was 11% and 89%, respectively. Ancestral GG genotype of rs28493229 was found in 78% of patients with KD (Table 1B).

We failed to show an association of any allele or genotype of rs28493229 with susceptibility to disease. This result is in accordance with the findings of studies on Chinese, Taiwanese, and Filipino patients (Table 1B).22,23,32,34,35 Although two previous meta-analyses had shown significant association of this SNP with KD,36,37 our meta-analysis combining our study with all previous studies on rs2849229 did not show any association with KD (Table 3A and Fig. 2a). This suggests that lowering of splicing efficiency of mRNA level due to the presence of the C allele of rs2849229 may not be a universal phenomenon.21 This may also highlight the remarkable heterogeneity of functional SNP on KD across ethnicities.

Although the association of the C allele of rs2290692 of 3′-UTR and KD was reported from China,22 our study did not show any association of any single allele or genotype with KD. Similar findings were reported in children with KD from Taiwanese and Korean ethnicities (Table 1C).24,25 Meta-analyses combining our study with previous studies on rs2290692 also did not show any significant association with susceptibility to KD (Table 3B, C and Fig. 2b, c). This may simply reflect ethnic variation or the limited role of a single allele or genotype of the SNP on the pathogenesis of KD.

However, analyzing any genotype containing G allele (i.e., CG + GG) in rs2290692, we observed a statistically significant association of genotypes with G allele with KD occurrence in both univariate and multivariate analysis (p = 0.015; OR = 4.16, 95% CI: 1.38–13.83) (Table 2C). Such an association of rs2290692 is not described in previous studies. It is postulated that polymorphism in any allele may result in altered regulation efficiency of miRNAs. These miRNAs are involved in the regulation of gene expression. Our result showing an overrepresentation of the G allele in the KD group suggests that associated miRNAs may bind to the G allele altering the gene expression.38 In addition, there may be evolutionary and epigenetic effects on the genetic structure of individuals in different ethnicities.

It appears that no specific SNP of the ITPKC gene may have a uniform association with susceptibility to KD and CAL formation across ethnicities. KD is undoubtedly a multifactorial disease and these SNPs may only confer some predisposition to develop KD.39 Pathogenesis of KD may reflect the effect of several genes involved in calcium-NFAT signaling, which ultimately activates gene transcription for T cell activation and cytokine secretion.35 Larger multicentric studies incorporating different ethnicities are required to understand the genetic basis of KD.

The relevance, strength, and limitations of the study

This is the first study of ITPKC SNP association with KD and CAL from India. Results from RFLP were confirmed by Sanger sequencing. Age and sex matching eliminate possible differences due to exposure prevalence and the effect of variable exposure in individuals. More importantly, the likelihood of false negatives due to the population stratification effect is also nullified by matching age and sex. However, this study has some limitations. This was a single-center hospital-based study and the numbers are understandably small. We were unable to conduct the functional level of studied gene polymorphism. Further, the effect of other genes (e.g., Ca2+/NFAT or tumor growth factor-β pathway) in the studied cohort also cannot be excluded.

Conclusions

In conclusion, this is the first study in the Indian population that has explored the association of KD with two SNPs (rs28493229 and rs2290692) of the ITPKC gene. The frequency of the C allele of rs28493229 of ITPKC was found to be 11%, which is lesser than that in the Japanese population (22%), but comparable with KD cohorts from Taiwan and China. We documented a higher frequency of G allele (63%) of rs2290692 in patients with KD in our study. This study, and meta-analysis, did not provide any evidence to support the association of rs28493229 with KD susceptibility or CAL in children with KD of North Indian ethnicity. Combined genotypes containing the G allele of rs2290692 (CG + GG) were found to be significantly associated with the risk of KD but not with CAL. Further studies with a larger sample size are required to confirm this finding. In addition, studies on multiple SNPs of the ITPKC gene together with haplotype analysis are needed to explore the relevance of ITPKC gene polymorphisms in KD in the Indian population.