Association Study Reveals One Susceptibility Locus with Vitiligo in the Chinese Han Population

Abstract

Objective:

The purpose of this study was to investigate associations between the 14 reported loci (from a meta-analysis of genome-wide association studies [GWAS] in the Caucasian population) and vitiligo in the Chinese Han population.

Materials and Methods:

In this study 14 single nucleotide polymorphisms (SNPs) at 14 different genetic loci were evaluated for their association with viteligo in a Chinese Han cohort, including 1472 cases and 1472 controls of by using the Sequenom MassArray iPLEX1 system. A Bonferroni adjustment was used for multiple comparisons and p_Bonferroni <0.0056 was considered statistically significant.

Results:

The T allele of the locus within the FBXO45-NRROS gene (3q29) was significantly associated with vitiligo (odds ratio = 1.22, 95% confidence interval: 1.10-1.36, p = 0.0001). Association at the genotype level was strong (p = 0.0007). The other SNPs were not associated with vitiligo (p_Bonferroni >0.0056).

Conclusion:

A SNP at the rs6583331 locus 3q29 is associated with the susceptibility of vitiligo in the Chinese Han population, which suggests that there is a common genetic factor predisposing to the development of vitiligo in the Chinese and Caucasian populations.

Introduction

Vitiligo is a common autoimmune skin disease in which the loss function of melanocyte results in the pigmentation of the skin or mucosa. Its main features include that the loss of melanocytes can lead to localized or generalized depigmentation or loss of plaque in the epidermis and mucous membranes (Nunes and Esser, 2011). Global epidemiological surveys have shown that the incidence of vitiligo is ∼0.5-1% (Ezzedine et al., 2012). The incidence of vitiligo in Chinese population is ∼0.56% in China (Wang et al., 2013). Its underlying pathobiology has remained largely unknown, which may be related to the interaction of immune, genetic, and acquired environments. In the past, multiple potential susceptibility genes were reported based on candidate gene association analysis; however, most of them were also associated with autoimmune diseases (Jin et al., 2010a). In the multiple populations, >60 susceptibility genes/loci were reported by genome-wide association studies (GWAS) of vitiligo (Jin et al., 2010a, 2010b, 2011, 2012; Quan et al., 2010; Tang et al., 2013).

Recently, Jin et al. (2016) identified many susceptibility genes/loci for vitiligo by using GWAS-meta in the Caucasian population, such as, 1q24.3 (rs78037977), 1q31.3-q32.1 (rs16843742), 2q13 (rs4308124), 2q33.2 (rs231725), 2q37.3 (rs41342147), 3p24.3 (rs35161626), 3q29 (rs6583331), 4q24 (rs1031034), 11q13.1 (rs12421615), 17q21.2 (rs11079035), 18q21.33 (rs8083511), 19p13.3 (rs4807000), 20q11.22 (rs6059655), and 20q13.13 (rs6012953). However, these new risk loci have never been validated in Chinese Han population with vitiligo. This study first investigates the associations between 14 reported single nucleotide polymorphisms (SNPs) and vitiligo in Chinese Han population.

Materials and Methods

Subjects

The sample size of this study includes 1472 cases and 1472 controls (Table 1). All samples were from the Resource Library of the Institute of Dermatology, Anhui Medical University. All study individuals were given informed consent, and signed and finished a complete epidemiological questionnaire. Based on the Declaration of Helsinki principles, the study was passed by the Institutional Ethical Committee of Anhui Medical University. All cases meet the diagnostic criteria of the Vitiligo European Task Force (Taieb and Picardo, 2007). All controls were healthy individuals without vitiligo, any other autoimmune diseases or systemic disorders, and without any family history of vitiligo (including first-, second-, and third-degree relatives). Patients and healthy controls were matched each other (Zhao et al., 2017). Based on the QIAamp DNA Blood kit (Qiagen, Valencia, CA) and their instructions, DNA was extracted from the peripheral venous blood, and was normalized to the concentrations for platform with 20-25 ng/mL.

Table 1.

Summary Information of Vitiligo Patients and Controls

	Case	Control
Total number	1472	1472
Age (years)	25.33 ± 15.39	30.78 ± 14.90
Male	782	769
Female	690	703
Nonsegmental	1359	NA
Segmental	73	NA
With autoimmune disease^a	121	NA
Without autoimmune disease	1351	NA

Including thyroid disease (66), alopecia areata (19), type I diabetes mellitus (3), systemic lupus erythematosus (1), rheumatoid arthritis (23), myasthenia gravis (6), and scleroderma (3).

SNP selection and genotyping

A recent study reported many new risk SNPs in Caucasian population by GWAS of vitiligo (Jin et al., 2016). Of them, the minor allele frequency of three SNPs (rs12203592, rs117744081, and rs73456411) was 0.00 in East Asian population from 1000 Genomes data and the design of PCR primers for six SNPs (rs10200159, rs78521699, rs10986311, rs71508903, rs35860234, and rs2304206) was failed. Finally, 14 SNPs at 1q24.3 (rs78037977), 1q31.3-q32.1 (rs16843742), 2q13 (rs4308124), 2q33.2 (rs231725), 2q37.3 (rs41342147), 3p24.3 (rs35161626), 3q29 (rs6583331), 4q24 (rs1031034), 11q13.1 (rs12421615), 17q21.2 (rs11079035), 18q21.33 (rs8083511), 19p13.3 (rs4807000), 20q11.22 (rs6059655), and 20q13.13 (rs6012953) were selected and then genotyped in an independent cohort (Chinese Han), including 1472 patients with vitiligo and 1472 controls. By using the Sequenom MassArray system, all SNPs were genotyped. The primers for specific detection and PCR were designed by using the Assay Design 3.0 software (Sequenom). The DNA samples were amplified by multiplex PCRs, and then products were used for locus-specific single-base extension reactions. The PCR products were moved to a 384-element plate after amplified by multiplex PCRs. Alleles were analyzed by using MALDI-TOF MS, and then the mass spectrograms were scanned by using MassARRAY Typer software (Quan et al., 2010).

Statistical analysis

The associations for SNPs were analyzed by comparing the minor allele frequency between cases and controls (PLINK 1.07 software) (Purcell et al., 2007); meanwhile, the Hardy-Weinberg equilibrium (HWE) was tested in cases and controls to evaluate the significant deviation by using online DeFinetti software. As a part of quality control, SNPs were excluded if they had a call rate <95%, or a significant deviation from HWE in the controls (p < 0.05). Of the 14 genotyped SNPs, 5 SNPs (rs41342147, rs1031034, rs11079035, rs8083511, and rs4807000) deviated from HWE in controls (p < 0.05, date not shown) due to the possible genotyping errors that resulted from the bad genotyping clusters for these SNPs. Finally, nine SNPs were included for the association analysis. Based on the Japanese in Tokyo, Japan (JPT) and Han Chinese in Beijing, China (CHB) data (hg18/HapMap Phase II), the regional plot was produced by using LocusZoom. The dominant and recessive models for the associated SNPs were also performed. The analyses of genotype-phenotype for associated SNPs were also performed, the clinical phenotype includes age of onset (Hu et al., 2011), family history, gender, clinical classification, and autoimmune disease involvement. The p value of Bonferroni corrections was regarded as statistically significant (p_Bonferroni <0.05/9 = 0.0056).

Results

Association between SNPs and vitiligo

The analysis results of all SNPs are shown in Table 2. The T allele of rs6583331 at 3q29 was significantly associated with vitiligo (odds ratio [OR] = 1.22, confidence interval [95% CI]: 1.10-1.36, p = 0.0001), even after correcting for multiple testing (p < 0.0056). The dominant and recessive models for rs6583331 are shown in Table 3. Association analysis at the full genotype was significant (p = 0.0007). Individuals with carrying genotypes TA (p = 0.0155, OR = 1.25, 95% CI: 1.04-1.49) and TT (p = 0.0001, OR = 1.50, 95% CI: 1.21-1.84) had a higher risk of developing the disease. They also showed that the dominant model (p = 0.0011, OR = 1.32, 95% CI 1.12-1.57) might be better fit than recessive model (p = 0.0032, OR = 1.30, 95% CI 1.09-1.54) for SNP rs6583331. However, no significant association evidence was observed for the rest SNPs (p_Bonferroni >0.0056, Table 2).

Table 2.

Summary of Association Results of 9 Single Nucleotide Polymorphisms in 14 Loci Between Cases and Controls

Chr	Gene	SNP	Allele ^a	MAF		p	OR	95% CI	p_HWE		Call rate ^b
Chr	Gene	SNP	Allele ^a	Case	Control	p	OR	95% CI	p_case	p_control	Call rate ^b
1q24.3	FASLG	rs78037977	G/A	0.0014	0.0014	0.9898	0.99	0.25-3.97	1.0000	1.0000	0.9739
1q31.3-32.1	PTPRC	rs16843742	C/T	0.2101	0.2028	0.4900	1.05	0.92-1.19	0.4783	1.0000	0.9912
2q13	BCL2L11-MIR4435-2HG	rs4308124	C/T	0.2137	0.2120	0.8730	1.01	0.89-1.15	1.0000	0.4813	0.9918
2q33.2	CTLA4	rs231725	G/A	0.3577	0.3719	0.2651	0.94	0.84-1.05	0.2060	0.8208	0.9711
3p24.3	UBE2E2	rs35161626	—/A	0.2490	0.2356	0.2324	1.08	0.95-1.21	0.5296	0.3824	0.9912
3q29	FBXO45-NRROS	rs6583331	T/A	0.5164	0.4659	0.0001	1.22	1.10-1.36	0.9578	0.8312	0.9694
11q13.1	PPP1R14B-PLCB3-BAD-GPR137-KCNK4-TEX40-ESRRA-TRMT112-PRDX5	rs12421615	A/G	0.0726	0.0784	0.4061	0.92	0.76-1.12	1.0000	0.5881	0.9901
20q11.22	RALY-EIF252-ASIP-AHCY-ITCH	rs6059655	A/G	0.0028	0.0014	0.2507	2.00	0.60-6.63	1.0000	1.0000	0.9681
20q13.13	PTPN1	rs6012953	G/A	0.2794	0.2908	0.3328	0.95	0.84-1.06	0.1353	1.0000	0.9925

Minor allele/major allele.

The call rates of all SNPs were >95%.

CI, confidence interval; HWE, Hardy-Weinberg equilibrium; MAF, minor allele frequency; OR, odds ratio; SNPs, single nucleotide polymorphisms.

Table 3.

Distribution of Genotypes and Genetic Model Analysis for rs6583331 in Cases and Controls

	Cases (%) (n = 1432)^a	Controls (%) (n = 1422)^a	OR (95% CI)	p
Full genotype
AA	334 (23.32)	408 (28.69)	Reference	0.0007^b
TT	381 (26.61)	311 (21.87)	1.50 (1.21-1.84)	0.0001
TA	717 (50.07)	703 (49.44)	1.25 (1.04-1.49)	0.0155
Recessive model
TT/(TA+AA)	381 (26.61)/1051 (73.39)	311 (21.87)/1111 (78.13)	1.30 (1.09-1.54)	0.0032
Dominant model
TT+TA/AA	1098 (76.68)/334 (23.32)	1014 (71.31)/408 (28.69)	1.32 (1.12-1.57)	0.0011

Genotyping of 40 cases and 50 controls failed.

p Value for full genotypes (AA vs. TT vs. TA).

CI, confidence interval; T, minor allele (effect/risk allele); A, major allele (reference allele); dominant model, (DD, Dd) versus dd; recessive model, DD versus (Dd, dd), where D is the minor allele and d the major allele; OR, odds ratio.

Genotype and subphenotype analysis

The analyses of genotype-phenotype (age of onset, gender, family history, clinical classification, and concomitant autoimmune diseases) of the rs6583331 was also performed. The genotype and allele distributions for SNP rs6583331 in different subgroups are summarized in Table 4. However, no statistical difference in different subgroups was observed for the SNP rs6583331 by comparing the genotype and allele frequencies (p > 0.05, Table 4).

Table 4.

Distribution of Genotypes and Alleles for Single Nucleotide Polymorphism rs6583331 in Subgroups

Subphenotypes	Genotype/allele frequency (%)					Genotype	Allele
Subphenotypes	TT	TA	AA	T	A	p^a	OR (95% CI)	p^b
Age at onset
≤20 years (n = 736)	190 (25.82)	373 (50.68)	173 (23.50)	753 (51.15)	719 (48.85)	0.7828	0.96 (0.83-1.11)	0.5924
>20 years (n = 696)	191 (27.44)	344 (49.43)	161 (23.13)	726 (52.16)	666 (47.84)	0.7828	0.96 (0.83-1.11)	0.5924
Gender
Male (n = 760)	208 (27.37)	378 (49.74)	174 (22.89)	794 (52.24)	726 (47.76)	0.7720	1.05 (0.91-1.22)	0.4974
Female (n = 672)	173 (25.74)	339 (50.45)	160 (23.81)	685 (50.97)	659 (49.03)	0.7720	1.05 (0.91-1.22)	0.4974
Family history
Positive (n = 151)	41 (27.15)	72 (47.68)	38 (25.17)	154 (50.99)	148 (49.01)	0.7959	0.97 (0.77-1.23)	0.8112
Negative (1281)	340 (26.54)	645 (50.35)	296 (23.11)	1325 (51.72)	1237 (48.28)	0.7959	0.97 (0.77-1.23)	0.8112
Vitiligo type
Nonsegmental (n = 1359)	359 (26.42)	687 (50.55)	313 (23.03)	1405 (51.69)	1313 (48.31)	0.2760	1.04 (0.75-1.45)	0.8124
Segmental (n = 73)	22 (30.14)	30 (41.09)	21 (28.77)	74 (50.68)	72 (49.32)	0.2760	1.04 (0.75-1.45)	0.8124
Autoimmune disease involvement
Positive (n = 115)	30 (26.09)	59 (51.30)	26 (22.61)	119 (51.74)	111 (48.26)	0.9617	1.00 (0.77-1.32)	0.9752
Negative (n = 1317)	351 (26.65)	658 (49.96)	308 (23.39)	1360 (51.63)	1274 (48.37)	0.9617	1.00 (0.77-1.32)	0.9752

For genotype using a 2 × 3 contingency table.

For allele using a 2 × 2 contingency table.

Discussion

It is well known that vitiligo is a complex disease; many disease susceptibility loci have been identified with the application of GWAS, most of which encoded proteins play important biological pathway in immune cells and regulate the immune system, even causing the destruction of its melanocytes (Spritz, 2012), which highlight that the genetic factor and immune regulations play important roles in pathogenetic mechanisms for vitiligo. This study further confirms that rs6583331 at 3q29 is significantly associated with vitiligo susceptibility in an independent cohort from Chinese Han population; it also reveals a common susceptibility to vitiligo in the Caucasian and Chinese populations.

The susceptible SNP rs6583331 at 3q29 is located in a linkage disequilibrium (LD) block that includes two important genes, FBXO45 and NRROS (Fig. 1). FBXO45 contains three exons and encodes an FBXO protein (belongs to the estrogen-induced protein) currently known to contain the SPRY domain (that interacts with Par-4) (Yoshida, 2005; Chen et al., 2014). Its expression level is associated with several cellular processes, including those that are associated with tumorigenic pathways, which include cell-cycle progression, apoptosis, and transcription (Diaz and de Herreros, 2016). The mechanisms of transcription and post-transcription can regulate the biological pathway of NOS2 by the activation of toll-like receptor (TLR) and nuclear factor kappa B (NF-κB) signaling pathways (Pautz et al., 2010). Interestingly, FBXO45, as a direct NOS2 interactor and interacting with Par-4, might involve in the NF-kB biological pathways (Foster et al., 2013). In addition to inflammation, the inhibition of NF-kB might be participated in the immunobiology pathway of vitiligo (Arase et al., 2016).

FIG. 1.

Regional association plot for the associated SNP rs6583331 in 3q29 region. SNP is plotted by chromosomal position (GRCh36/hg18; x axis) and association with vitiligo from this study (−log10Pvalue; y axis). Estimated recombination rates (based on the combined CHB and JPT samples from the HapMap project) were plotted in gray. Genes are indicated in the lower panel of the plot. SNP, single nucleotide polymorphism.

For this locus, another gene NRROS contains three exons and encodes a leucine-rich repeat containing transmembrane protein in the endoplasmic reticulum, which may involve in the regulation of immunity and inflammation by activating reactive oxygen species (ROS) system and controlling NOX2 protein biological utility (Bonini and Malik, 2014; Noubade et al., 2014). Study showed that NRROS can negatively regulate the activation of TLR signaling pathway, which might affect the development for allergic and autoimmune-related diseases (Liew et al., 2005). In addition, it also involves TLR-mediated immune responses by regulating the activation of NF-κB and JNK signaling pathways (Liu et al., 2013; Su et al., 2014). Therefore, it is speculated that the encoded proteins by FBXO45 and NRROS in 3q29 might involve a variety of immunobiological pathways, which further indicates that genetic and immune factors play an important role in the pathogenesis of vitiligo.

Genetic risk for autoimmunity in HLA genes is most often attributed to structural specificity resulting in presentation of self-antigens. Autoimmune vitiligo is strongly associated with the major histocompatibility complex (MHC) class I and class II regions by GWAS. In Caucasians, the most associated genes included HLA-A in class I gene region, HLA-DRB1 and HLA-DQA1 in the class II region (Jin et al., 2010a). In the Chinese, the major MHC association signals were in the HLA class III gene region and the HLA-C-HLA-B region (Quan et al., 2010). Additional studies show that several genetic variants in the MHC class II region are also associated with the age of onset for vitiligo, such as rs7758128 located near c6orf10-BTNL2 (Jin et al., 2011) and rs145954018 located near HLA-DRA (Jin et al., 2019), which indicates that the risk loci in the MHC class II region might affect clinical phenotype of disease.

Some studies show that several haplotypes of MHC loci regulatory variants might quantitatively affect HLA protein expression. The haplotype rs145954018del-rs9271597A (a major signal within MHC class II locus from previous GWAS) is specifically associated with increased expression of HLA-DQB1 mRNA and HLA-DQ protein by monocytes and dendritic cells (Jin et al., 2019). An SNP haplotype is a transcriptional regulator of the HLA-A that induces elevated expression of HLA-A RNA, and strong linkage disequilibrium with an HLA-A *02:01 (Hayashi et al., 2016). The high-risk haplotype for several SNPs located within a predicted super enhancer between HLA-DRB1 and HLA-DQA1 elevated surface expression of HLA-DR and HLA-DQ, which increase in production of IFN-γ and IL-1β in the peripheral blood mononuclear cells (Cavalli et al., 2016). These studies demonstrate that the MHC regulatory variations may represent a significant component of genetic risk for vitiligo. Outside MHC region, the high-risk haplotypes for several nonsynonymous variants in NLRP1 not only were associated with vitiligo, but also increases the amounts of the IL-1β precursor to mature bioactive IL-1β under basal (resting) conditions by activation of the NLRP1 inflammasome (Levandowski et al., 2013).

This study reveals a locus (3q29) that contributes to vitiligo susceptibility. The encoded proteins for plausible candidate genes FBXO45 and NRROS in this locus may play important roles in TLR, NF-κB, and JNK signaling pathways, which further emphasize that the immunoregulatory genes in non-MHC loci play important roles in the pathogenetic mechanisms for disease. The current findings provide an indirection for function studies of candidate gene for this locus in future. The SNP is only an associated signal, and there may be exit causal variants in the LD block region. Therefore, it is still necessary to identify the causal gene(s) by the targeted sequencing and function studies.

Footnotes

Acknowledgments

The authors thank all individuals who took part in this research.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was funded by the Youth Program of the National Natural Science Foundation of China (No. 81402591).

References

Arase

, Yang

, Tanemura

, et al. (2016) The effect of rhododendrol inhibition of NF-kappaB on melanocytes in the presence of tyrosinase. J Dermatol Sci, 83:157-159.

Bonini

, Malik

. (2014) Regulating the regulator of ROS production. Cell Res, 24:908-909.

Cavalli

, Hayashi

, Jin

, et al. (2016) MHC class II super-enhancer increases surface expression of HLA-DR and HLA-DQ and affects cytokine production in autoimmune vitiligo. Proc Natl Acad Sci U S A, 113:1363-1368.

Chen

, Sahasrabuddhe

, Szankasi

, et al. (2014) Fbxo45-mediated degradation of the tumor-suppressor Par-4 regulates cancer cell survival. Cell Death Differ, 21:1535-1545.

Diaz

, de Herreros

(2016) F-box proteins: Keeping the epithelial-to-mesenchymal transition (EMT) in check. Semin Cancer Biol, 36:71-79.

Ezzedine

, Lim

, Suzuki

, et al. (2012) Revised classification/nomenclature of vitiligo and related issues: The Vitiligo Global Issues Consensus Conference. Pigment Cell Melanoma Res, 25:E1-E13.

Foster

, Thompson

, Forrester

, et al. (2013) Proteomic analysis of the NOS2 interactome in human airway epithelial cells. Nitric Oxide, 34:37-46.

Hayashi

, Jin

, Yorgov

, et al. (2016) Autoimmune vitiligo is associated with gain-of-function by a transcriptional regulator that elevates expression of HLA-A*02:01 in vivo. Proc Natl Acad Sci U S A, 113:1357-1362.

, Ren

, Zhu

, et al. (2011) Comparisons of clinical features of HLA-DRB1*07 positive and negative vitiligo patients in Chinese Han population. J Eur Acad Dermatol Venereol, 25:1299-1303.

10.

Jin

, Andersen

, Yorgov

, et al. (2016) Genome-wide association studies of autoimmune vitiligo identify 23 new risk loci and highlight key pathways and regulatory variants. Nat Genet, 48:1418-1424.

11.

Jin

, Birlea

, Fain

, et al. (2010a) Variant of TYR and autoimmunity susceptibility loci in generalized vitiligo. N Engl J Med, 362:1686-1697.

12.

Jin

, Birlea

, Fain

, et al. (2010b) Common variants in FOXP1 are associated with generalized vitiligo. Nat Genet, 42:576-578.

13.

Jin

, Birlea

, Fain

, et al. (2011) Genome-wide analysis identifies a quantitative trait locus in the MHC class II region associated with generalized vitiligo age of onset. J Invest Dermatol, 131:1308-1312.

14.

Jin

, Birlea

, Fain

, et al. (2012) Genome-wide association analyses identify 13 new susceptibility loci for generalized vitiligo. Nat Genet, 44:676-680.

15.

Jin

, Roberts

GHL

, Ferrara

, et al. (2019) Early-onset autoimmune vitiligo associated with an enhancer variant haplotype that upregulates class II HLA expression. Nat Commun, 10:391.

16.

Levandowski

, Mailloux

, Ferrara

, et al. (2013) NLRP1 haplotypes associated with vitiligo and autoimmunity increase interleukin-1beta processing via the NLRP1 inflammasome. Proc Natl Acad Sci U S A, 110:2952-2956.

17.

Liew

, Xu

, Brint

, O'Neill

. (2005) Negative regulation of toll-like receptor-mediated immune responses. Nat Rev Immunol, 5:446-458.

18.

Liu

, Zhang

, Chai

, et al. (2013) Identification and characterization of a unique leucine-rich repeat protein (LRRC33) that inhibits Toll-like receptor-mediated NF-kappaB activation. Biochem Biophys Res Commun, 434:28-34.

19.

Noubade

, Wong

, Ota

, et al. (2014) NRROS negatively regulates reactive oxygen species during host defence and autoimmunity. Nature, 509:235-239.

20.

Nunes

, Esser

. (2011) Vitiligo epidemiological profile and the association with thyroid disease. An Bras Dermatol, 86:241-248.

21.

Pautz

, Art

, Hahn

, et al. (2010) Regulation of the expression of inducible nitric oxide synthase. Nitric Oxide, 23:75-93.

22.

Purcell

, Neale

, Todd-Brown

, et al. (2007) PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet, 81:559-575.

23.

Quan

, Ren

, Xiang

, et al. (2010) Genome-wide association study for vitiligo identifies susceptibility loci at 6q27 and the MHC. Nat Genet, 42:614-618.

24.

Spritz

. (2012) Six decades of vitiligo genetics: Genome-wide studies provide insights into autoimmune pathogenesis. J Invest Dermatol, 132:268-273.

25.

, Mei

, Liang

, et al. (2014) Epigenetically modulated LRRC33 acts as a negative physiological regulator for multiple Toll-like receptors. J Leukoc Biol, 96:17-26.

26.

Taieb

, Picardo

. (2007) The definition and assessment of vitiligo: A consensus report of the Vitiligo European Task Force. Pigment Cell Res, 20:27-35.

27.

Tang

, Zhang

, Hu

, et al. (2013) Association analyses identify three susceptibility Loci for vitiligo in the Chinese Han population. J Invest Dermatol, 133:403-410.

28.

Wang

, Du

, Wang

, et al. (2013) Prevalence and clinical profile of vitiligo in China: A community-based study in six cities. Acta Derm Venereol, 93:62-65.

29.

Yoshida

. (2005) Characterization of estrogen-induced F-box protein FBXO45. Oncol Rep, 14:531-535.

30.

Zhao

, Fang

, Tang

, et al. (2017) An in-depth analysis identifies two new independent signals in 11q23.3 associated with vitiligo in the Chinese Han population. J Dermatol Sci, 88:103-109.