AHR promoter variant modulates its transcription and downstream effectors by allele-specific AHR-SP1 interaction functioning as a genetic marker for vitiligo

Vitiligo is an acquired depigmentation disorder largely caused by defective melanocyte- or autoimmunity-induced melanocyte destruction. The aryl hydrocarbon receptor (AHR) is essential for melanocyte homeostasis and immune process, and abnormal AHR was observed in vitiligo. We previously identified the T allele of AHR −129C > T variant as a protective factor against vitiligo. However, biological characterization underlying such effects is not fully certain, further validation by mechanistic research is warranted and was conducted in the present study. We showed that −129T allele promoted AHR transcriptional activity through facilitating its interaction with SP1 transcription factor (SP1) compared with −129C allele. We subsequently found reduced peripheral AHR and SP1 transcript expressions in vitiligo and a negative correlation of AHR level with disease duration. We also investigated AHR-related cytokines and observed increased serum TNF-α concentration and diminished serum levels of IL-10 and TGF-β1 in vitiligo. Further genetic analysis showed that -129T carriers possessed higher levels of AHR and IL-10 than −129C carriers. Therefore, our study indicates that the modulation of AHR transcription by a promoter variant has a profound influence on vitiligo, not only advancing our understanding on AHR function but also providing novel insight into the pathogenesis of degenerative or autoimmune diseases including vitiligo.

The aryl hydrocarbon receptor (AHR) is a ligand-activated transcription factor and belongs to the basic-helix-loop-helix family 10 . Upon binding ligand, AHR translocates into the nucleus to govern target genes 11 . AHR is well characterized for immune regulation through mediating T-cell differentiation and cytokine milieu 12,13 , and more recently, scientific evidence strongly supports that AHR is vital to melanocyte homeostasis. Activation of AHR pathway stimulated melanogenesis by improving expressions of TYR and TYRP in human melanocytes 14 . Ahr-deficient mice showed decreased melanocyte number and hypopigmentation in the tail skin caused by reduced levels of SCF and C-Kit 15 . Additionally, significantly declined expressions of AHR and its target genes were observed in the epidermis of vitiligo patients 16 . These data imply that mutation or dysfunction of AHR might be involved in vitiligo.
In consideration of the critical role of AHR in both melanocyte and cellular immunity and aberrant AHR pathway in vitiligo, we previously evaluated the potential association between AHR polymorphisms and vitiligo susceptibility. Our data demonstrated that the T allele of − 129C > T variant (rs10249788) in the AHR promoter region is associated with a protective effect on vitiligo in Han Chinese populations 17 , which might be a functional variation through altering AHR transcription process. The promoter of human AHR gene lacks TATA and CCAAT boxes but possesses several putative SP1 transcription factor (SP1) binding sites within a highly GC-rich region 18 . SP1 is a Cys 2 /His 2 -type zinc-finger transcription factor that binds to GC box elements (5′ -GGGCGG-3′ ) within promoter region 19 . SP1 is particularly important to the TATA-less genes, which regulates transcription of multiple target genes involved in cell growth, differentiation, apoptosis and immune response 20 . Early researches have revealed that SP1 dominates the maximal constitutive activity and basal expression of AHR gene via binding to these GC-rich motifs 21 . The abnormal interaction between SP1 and AHR promoter is responsible for AHR down-regulation in human diseases 22,23 . As the − 129C > T polymorphism is in the core promoter region of AHR gene and is juxtaposed 5′ to the SP1 recognition sequence, we therefore hypothesized that the AHR − 129C > T polymorphism could influence its transcription and downstream effectors in melanocyte biology or immune system, which could further affect the development of vitiligo. In the present study, we performed functional research to explore the molecular mechanisms underlying such genetic marker for vitiligo.

Results
Effects of AHR −129C > T polymorphism on AHR transcriptional activity. To assess the AHR promoter activity related to − 129C > T polymorphism, C or T promoter constructs were transiently transfected in human normal melanocyte PIG1 cells, human malignant melanoma LiBr cells and human embryonic kidney 293T cells. As shown in Fig. 1, the vectors with − 129T allele had enhanced relative luciferase activity compared with that of those with − 129C allele (P = 0.018 for PIG1 cells; P = 0.010 for LiBr cells; P = 0.023 for 293T cells). These results suggest that the AHR − 129T allele possesses an increased AHR transcriptional activity.
Impacts of AHR −129C > T polymorphism on SP1 binding to AHR promoter. Both nucleotide sequences and DNA shape influence specific binding of proteins to DNA 24,25 . The hydroxyl radical cleavage pattern embodies information on sequence-dependent variation in DNA structure, including solvent accessibility, minor groove width and electrostatic potential 26,27 . Where the minor groove is wide, and deoxyribose backbone hydrogens are exposed, cleavage intensity is high; where the groove is narrow, and backbone hydrogens are diminished in exposure, cleavage intensity is low 27,28 . Minor groove width and electrostatic potential are important for protein binding 25 . Specifically, narrowing of the groove is associated with more negative electrostatic potential, which is beneficial for proteins to insert side chains with positive charge into the groove 28,29 . To test the effect of AHR − 129C > T variant on DNA local structure, bioinformatic analysis was performed using the OH Radical Cleavage Intensity Database (ORChID) and showed different hydroxyl radical cleavage patterns among variants of the AHR − 129C > T polymorphism. As shown in Fig. 2c, the T allele with relatively lower score reflects narrowing of the DNA backbone, which might function as an electrostatic groove to further improve the DNA-protein interaction.
To detect the role of − 129C > T polymorphism in the interplay between AHR promoter and transcriptional regulators, we carried out electrophoretic mobility shift assays (EMSA) and found the DNA-protein complex with stronger intensity for the − 129T allele than − 129C allele (Fig. 2a, lanes 2 and 6; P = 0.008, Fig. 2b), which was abolished by excessive unlabeled probes (Fig. 2a, lanes 3 and 7). Since the AHR − 129C > T polymorphism is close to the binding site of SP1, we then conducted supershift assays and detected that anti-SP1 antibody supershifted the C and T allele-specific bands (Fig. 2a,  lanes 4 and 8), implying that this promoter variant could modify allele-specific binding affinity of SP1. We next performed chromatin immunoprecipitation assays (ChIP) to verify these observations within PIG1 cells (− 129CC) and LiBr cells (− 129TT). Both cell lines expressed comparable levels of SP1 protein (Fig. 2d). Similarly, − 129TT genotype had higher abundance in the interaction between SP1 and AHR promoter than that of − 129CC genotype (Fig. 2e, lane 3). To affirm the role of SP1 activation in AHR expression, PIG1 and LiBr cells were treated with 100 nM mithramycin A, a specific SP1 inhibitor that competitively binds to the SP1-binding sites 30 . The basal expression of AHR protein was relatively higher in LiBr cells (− 129TT) than PIG1 cells (− 129CC), and mithramycin A treatment suppressed AHR protein expression in PIG1 and LiBr cells (Fig. 2f). Taken together, these data reveal that SP1 prefers to bind to the AHR promoter containing − 129T allele rather than − 129C allele.
AHR and SP1 transcript levels in vitiligo patients and controls with different genotypes for the AHR −129C > T polymorphism. To investigate the impact of − 129C > T polymorphism on AHR transcript level and the correlation between SP1 and AHR expressions, we measured the AHR and SP1 mRNA levels in the peripheral blood mononuclear cells (PBMCs) from 23 vitiligo patients and 23 age-and sex-matched controls harboring different AHR − 129 genotypes (n = 16 for CC, n = 6 for CT and n = 1 for TT in vitiligo patients; n = 17 for CC, n = 5 for CT and n = 1 for TT in controls; Hardy-Weinberg equilibrium test: χ 2 = 0.571, P = 0.450). We observed positive correlation between AHR and SP1 expressions (R = 0.805, P < 0.001, Fig. 3d), both of which were obviously diminished in vitiligo patients compared with controls (P = 0.007 for AHR, Fig. 3a; P < 0.001 for SP1, Fig. 3b). Moreover, the AHR transcript expression was negatively correlated with total disease duration (R = − 0.556, P = 0.020, Fig. 3e), but no significant relevance with disease onset age and body surface area (BSA) involvement ( Fig. 3f,g). As expected, individuals with protective − 129TT/CT genotypes had higher AHR level than those with − 129CC genotype (6.059 ± 1.467 vs. 2.476 ± 0.356, P = 0.017, Fig. 3c). These findings implicate that the AHR − 129C > T polymorphism reflects on the level of AHR transcript expression.

Association between the AHR −129C > T polymorphism and serum cytokine levels in vitiligo.
Because T cell-mediated autoimmunity and altered cytokines are involved in vitiligo as well as AHR plays a pivotal role in T cell development and cytokine production, mutation or abnormal expression of AHR may influence cytokine levels associated with vitiligo. We therefore examined the major cytokines of T helper and regulatory T cells in 16 vitiligo patients and 16 age-and sex-matched controls carrying various AHR − 129 genotypes (n = 10 for CC, n = 4 for CT and n = 2 for TT in vitiligo patients; n = 11 for CC, n = 4 for CT and n = 1 for TT in controls; Hardy-Weinberg equilibrium test: χ 2 = 0.515, P = 0.473). We detected significantly increased TNF-α concentration and decreased levels of IL-10 and TGF-β 1 in the serum of vitiligo patients compared with controls (29.451 ± 4.748 ng/ml vs. 66.500 ± 17.805 ng/ml, P = 0.022 for TNF-α ; 20.649 ± 9.514 vs. 6.375 ± 1.885 ng/ml, P = 0.019 for IL-10; 7258.105 ± 687.912 ng/ ml vs. 5895.036 ± 365.716 ng/ml, P = 0.004 for TGF-β 1; Fig. 4a,e,g). No difference was observed between the cases and controls in the levels of IFN-γ , IL-17A and IL-22 (Fig. 4c,i,k). Furthermore, individuals with − 129TT/CT genotypes had greatly enhanced IL-10 expression compared with the group with − 129CC genotype (7.118 ± 1.869 ng/ml vs. 25.719 ± 13.565 ng/ml, P = 0.023; Fig. 4f). However, the AHR − 129C > T polymorphism seems not to be related to other cytokines (Fig. 4b,d,h,j,l). The evidence supports that the AHR − 129C > T polymorphism might be associated with IL-10 production.

Discussion
In the current study, we investigated the molecular mechanisms of AHR functional variation (− 129C > T polymorphism) underlying its association with vitiligo. Our data demonstrated that the protective T allele of this polymorphism dramatically enhanced AHR promoter activity through promoting the binding activity of SP1. We also found decreased AHR transcript expression in the PBMCs of vitiligo patients, which was negatively correlated with disease duration. Further genetic analysis showed that − 129T carriers had higher AHR expression than − 129C carriers. We further detected AHR related cytokines in vitiligo patients and observed declined serum IL-10 level and its association with the AHR − 129C > T Scientific RepoRts | 5:13542 | DOi: 10.1038/srep13542 variant. Thus, our scientific evidence strengthens our previous finding 17 and further clarifies that the AHR − 129C > T polymorphism is indeed a functional genetic marker for vitiligo.
In this report, we first proved the contribution of − 129C > T variation to AHR promoter function, which might influence AHR transcription by altering its interaction with transcriptional regulators. Human AHR gene contains features of a housekeeping gene as possessing GC-rich regions 31 . The core promoter region (− 1 ∼ − 250) has a GC content of 70%, containing several putative binding sites for SP1 32 . Deletion of the SP1 recognition site reduced basal promoter activity of human AHR gene to 20% 21 . In addition, impaired interaction between SP1 and AHR promoter caused by SP1 inhibitor or AHR hypermethylation led to AHR downregulation 22,23,33 . Consistent with previous study, we also found that SP1 inhibitor down-regulated AHR expression in PIG1 and LiBr cells. These results suggest that the SP1-binding motif is sufficient for maximal constitutive expression of AHR genes. Given the pivotal role of SP1 in AHR expression and AHR − 129C > T variation located nearby the core binding site for SP1, we then performed EMSA and ChIP assays and observed the − 129T allele with higher SP1 binding affinity than the − 129C allele. Li D. et al. previously reported that the − 129C > T variant interfered the binding activity of nuclear factor 1-C, which might down-regulate AHR transcription via antagonizing SP1 occupancy 34 . Although the mutual effect of SP1 and other regulators on AHR transcription remains unclear, it is conceivable that the AHR − 129C > T variant may affect the binding of SP1 to the AHR promoter in a direct or indirect manner.
The association between AHR expression and vitiligo remains to be elucidated. Gene expression analysis of PBMCs from independent cases and controls showed apparently decreased AHR mRNA level in vitiligo, which was clearly negatively correlated with disease duration of vitiligo patients, suggesting that down-regulation of AHR transcription is relative to vitiligo. We also found that declined SP1 mRNA level in vitiligo patients and a strongly positive correlation between SP1 and AHR expressions. It has been reported that the expression levels of SP1 and AHR vary greatly among different tissues 21 , but highest mRNA levels of SP1 and AHR were observed in the same tissues, such as thymus, lung, and spleen 35,36 . Differential SP1 expression and abnormal SP1 binding activity were demonstrated to mediate reduced expressions of AHR and its target genes in human breast cancer cells 22,37 . In order to investigate whether the regulatory role of − 129C > T mutation in SP1 binding affinity and AHR promoter activity could eventually influence AHR transcript expression, we performed genetic analysis showing that the protective − 129T allele is associated with elevated expression of AHR mRNA. Both in-vitro and in-vivo transcriptional analyses provided consistent results, indicating the potential effect of − 129C > T mutation on AHR expression, which might in turn influence its function in regulating melanogenic factors including TYR, TYRP, SCF and C-Kit 14,15 . Preliminary studies have implicated that abnormalities of TYR and TYRP, such as genetic mutation, decreased expression, altered folding or maturation and aberrant retention, may lead to increased toxic metabolites and subsequent melanocyte apoptosis in vitiligo 38,39 . TYRP could prevent premature melanocyte death in animals 39 . Moreover, Mutation of SCF or C-Kit caused hair hypopigmentation in mice 40 , and aberrant SCF/C-Kit pathway contributed to vitiligo and piebaldism in human [41][42][43][44] . SCF could protect primary human melanocytes against apoptosis and promote repigmentation in vitiligo mouse model 45,46 . Accordingly, it could be speculate that the AHR − 129C > T polymorphism might be associated with vitiligo through influencing AHR expression and its downstream melanogenic factors.
To further understand another significance of AHR − 129C > T polymorphism, we assessed the impact of this functional variant on the immune cytokines in vitiligo. Consistent with previous findings 47, 48 , we found increased TNF-α concentration and declined levels of IL-10 and TGF-β 1 in the serum of vitiligo patients. Further genetic analysis showed that IL-10 was strikingly higher in subjects carrying AHR − 129T allele than − 129C allele. Recent studies have demonstrated that AHR could affect IL-10 expression via mediating the differentiation of Treg cells 49,50 . AHR activation promotes Treg cells development by diverse mechanisms, including transcription regulation 13 , epigenetic modification 51 , cooperation with transcription factors Smad1 (SMAD family member 1) and Aiolos (IKAROS family zinc finger 3) 50 , inhibition of IL-2 52 and limitation of STAT1 (signal transducer and activator of transcription 1) activation 53 . These Treg cells induced by AHR pathway were shown to ameliorate autoimmune diseases such as type I diabetes 54 , multiple sclerosis 13 , colitis 51 and graft versus host disease 55 . Further, naive T cells from Ahr null mice inefficiently generated Tregs with decreased IL-10 levels 53,56 , indicating a specific role of AHR in Treg generation and IL-10 production. Recently, abnormal Tregs were identified in vitiligo, with evidence of reduced number, impaired skin homing, defective function and declined IL-10 expression 48,57-59 . Consequently, our data of genetic analysis imply that the AHR − 129C > T mutation might be related to vitiligo partially by affecting IL-10 production, which further supports previous findings about AHR function in immune regulation. Nevertheless, immune system contains many biological structures and complex processes, which could be modulated by mutual effort of multiple factors including AHR. Thus, functional studies may be needed to explore the regulatory mechanisms of AHR pathway in the immune pathology of vitiligo further.
There are several limitations within our study. First, owing to absence of melanocytes in vitligo epidermis and Koebner phenomenon exiting in vitiligo, it is hard to acquire primary melanocytes or tissue samples from vitiligo patients and we therefore used PBMCs instead of melanocytes to confirm these findings. Second, multiple genes and various environmental factors make collaborative contribution to vitiligo. These highlight the need to explore the gene-gene or gene-environment interaction in vitiligo in our future genetic studies.
In summary, we provide evidence that AHR -129C > T variant could lead to allele-specific binding of SP1 to AHR promoter and further modify its transcription and downstream effectors in vitiligo. Therefore, our findings have not only advanced our understanding on AHR physiological function but also provided significant insight into the pathogenesis of degenerative disorders or autoimmune diseases including vitiligo.

Methods
In silico AHR promoter analysis. ORChID (dna.bu.edu/orchid) was used to predict the hydroxyl radical cleavage pattern of local variation in solvent-accessible surface area of duplex DNA, and provided information on the local structural DNA profiles from nucleotides sequence containing the AHR − 129C > T polymorphism. Blood samples. For assays of gene expression, 23 patients with vitiligo and 23 sex-and age-matched control subjects were enrolled in the present study. We obtained 10 ml of peripheral blood from each participant for PBMCs isolation as described 58 . For cytokines measurement, we used 4 ml of peripheral blood from 16 vitiligo patients and 16 sex-and age-matched health subjects for serum collection. The detailed methods of recruiting subjects and genotyping of AHR − 129C > T variant have been described previously 17 . In particular, information on demographics and other characteristics, including disease duration, onset age and BSA, was obtained using questionnaires (details shown in Supplementary Table S1 online). All the subjects gave written informed consent. The study was approved by the ethics review board of the Fourth Military Medical University and was conducted according to the Declaration of Helsinki Principles.

Construction of reporter plasmids.
A 534-bp DNA fragment (− 534 to − 1) of human AHR promoter region was directly synthesized with added KpnI and MluI sites at 5′ ends in forward and reverse sequences, respectively (Cnservice Invitrogen, Shanghai, China). The DNA sequence was inserted between KpnI and MluI sites in the firefly luciferase reporter vector pGL3-Basic (Promega, Madison, WI, USA). The − 129C > T polymorphism was introduced into the recombinant pGL3 reporter vector by performing site-directed mutagenesis. The sequence of constructs was confirmed by direct DNA sequencing. EMSA assay. The DNA-protein interactions were detected using LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The biotin labeled and unlabeled double-stranded oligonucleotides corresponding to a 30-bp sequence 5′ -BIO-AAGAC/ TGGAATGGAATCCAGATGGGCGGGGG-3′ (SP1 binding site underlined) were synthesized by Viagene Biotech Inc (Viagene Biotech Inc, Tampa, Florida, USA). Nuclear extracts (10 μ g) were isolated from PIG1 cells using Viagene Subcellular Protein Fraction Kit (Viagene Biotech Inc, Tampa, Florida, USA) and protein concentration was determined by BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, Massachusetts, USA). For each gel shift reaction (15 μ l), a total of 300 fmol labeled probes were combined with 10 μ g of nuclear extract, 1.5 μ g of poly (deoxyinosinic-deoxycytidylic acid) and 10 × binding buffer. For competition assays, a 100-fold molar excess of unlabeled either − 129C or − 129T probes was preincubated for 20 min at room temperature with nuclear extracts before addition of the labeled probes. For each supershift reaction (15 μ l), 3.5 μ l of anti-SP1 antibody (Abcam Inc., Cambridge, MA, USA) was incubated with nuclear extracts at 4 °C for 15 min, followed by an additional incubation for 20 min at room temperature with the labeled probes. After incubation, samples were separated on a native non-denaturing 6.5% polyacrylamide gel and then transferred to a nylon membrane. The positions of biotin-labeled probes in the membrane were detected by a chemiluminescent reaction with the stabilized streptavidin-horseradish peroxidase conjugate according to the manufacturer's instructions and visualized by autoradiography.
ChIP assay. Before ChIP assay, PIG1 cells and LiBr cells were genotyped and sequenced to be CC and TT homozygous, respectively, which were performed similarly to our previous procedures (see Supplementary Fig. S1 online) 17 . ChIP assays were performed with ChIP-IT Express Kit (Active Motif, Carlsbad, USA) according to the manufacturer's protocols. Briefly, 1 × 10 6 of PIG1 cells and LiBr cells were treated with 2.7% formaldehyde for 10 min, followed by 2.5 M glycine stop solution. Cells were then collected and centrifuged at 4 °C and resuspended in lysis buffer for 30 min. The extracts were sonicated, and the DNA was sheared using enzymatic shearing cocktail until the DNA fragments were 200 ∼ 500-bp in size. Sheared chromatin (10 μ l) was incubated with protein G magnetic beads and 3 μ g of rabbit polyclonal anti-SP1 antibody (Abcam Inc., Cambridge, MA, USA), rabbit control IgG (Santa Cruz Biotechnology) or rabbit polyclonal anti-RNA polymerase II antibody (Abcam Inc., Cambridge, MA, USA) overnight at 4 °C. After extensive washing of the beads, the samples were reverse cross-linked and treated with ribonuclease A and proteinase K. After DNA purification, the immunoprecipitated chromatin fraction was amplified by PCR with primers specific for the AHR promoter (see Supplementary  Table S2 online) and control primers for the GAPDH promoter. The PCR products were analyzed on a 2.0% agarose gel.
Real-time quantitative RT-PCR. The total RNA was extracted from the PBMCs of vitiligo patients and controls with Trizol Reagent (Invitrogen, Inc.) and then converted to cDNA using PrimeScript TM RT Scientific RepoRts | 5:13542 | DOi: 10.1038/srep13542 Master Mix (TaKaRa, Inc.). The mRNA levels of AHR, SP1 and ACTB were measured by real-time quantitative PCR using specific primers (see Supplementary Table S2 online). Fold changes were normalized by the level of ACTB expression, and each assay was done in triplicate.

Western blotting. SP1 and AHR protein expressions in PIG1 and LiBr cells were examined by
Western blots analysis. Western blots (50 μ g protein/lane) were performed as described previously 30 , and probed with monoclonal antibodies specific for AHR or SP1 proteins (Abcam Inc., Cambridge, MA, USA). Hybridization signals for AHR and SP1 were normalized to hybridization signal for β -actin (Abcam Inc., Cambridge, MA, USA).
Measurement of cytokine concentrations. The amounts of TNF-α , IFN-γ , IL-10, TGF-β 1, IL-17A and IL-22 in the serum from each participant were determined using a corresponding ELISA kit (R&D systems, Bender MedSystems, Minneapolis, MN, USA) according to the manufacturer's instructions.
Statistical analysis. Student's t test was performed to evaluate the difference in luciferase activities between different constructs. Unpaired t test with Welch's correction was used to assess the different DNA-protein binding affinities between − 129T and − 129C probes. Non-parametric Mann-Whitney U-test was applied to analyze the results of peripheral AHR and SP1 mRNA expressions and serum cytokine concentrations between vitiligo patients and controls with different AHR − 129 genotypes. Correlation studies were performed using Pearson's correlation test. All tests were two-sided by using SPSS software (Version 13.0, SPSS Inc., Chicago, USA). Statistical significance was defined as P < 0.05.