Polymorphism of keratin 1 associates with systemic lupus erythematosus and systemic sclerosis in a south Chinese population

Both systemic lupus erythematosus (SLE) and systemic sclerosis (SSc) diseases are related to the genetic and environmental factors, causing damage to the skin. The mutations of keratin 1 gene (KRT1) were reported to associate with skin diseases. The single-nucleotide polymorphism (SNP, rs14024) and the indel polymorphism (cds-indel, rs267607656), consisting mostly of the common haplotypes and could be used for genotyping of KRT1. We used the PCR with sequence specific primers (PCR-SSP) to determine the genotype of KRT1 in 164 SLE, 99 SSc patients, and 418 healthy controls. The results showed that the mutant with G at SNP rs14024 was associated with the high risk to SLE (p = 6.48×10−5) and SSc (p = 8.75×10−5), while the deletion allele at rs267607656 was associated with the low risk to SSc (p = 4.89×10−4) comparing to the normal controls. Haplogenotype, Del-/MU+ was associated with high susceptibility to SLE (OR = 1.87, p = 0.001) and SSc (OR = 2.29, p = 2.34×10−4). In contrast, the Haplogenotype Del+/MU- was associated with resistance to SLE (OR = 0.35, p = 6.24×10−5) and SSc (OR = 0.34, p = 0.001). This study demonstrates that the variations in KRT1 and the specific polymorphism of KRT1 in this Chinese Han population are associated with autoimmune diseases SLE and SSc. Typing KRT1 might be helpful to identify SLE and SSc patients.


Introduction
Systemic lupus erythematosus (SLE) is an autoimmune inflammatory disease with symptoms that can affect almost any organ [1]. The disease is caused by the deposition of antibody and immune complexes in blood vessels, leading to inflammation in the skin, joints, serosa, central nervous system, and/or kidney. Both genetic and environmental factors contribute to development of the pathology [2]. Systemic sclerosis (SSc) is a systemic connective tissue disease, also of autoimmune origin [3]. Patients suffer from excessive accumulation of extracellular matrix that results in progressive fibrotic replacement of normal tissue architecture [4], which affects clinical checkup by physician specialists and was tabulated in the database. Clinical experimental data were achieved from the clinic laboratory reports. Additional demographic information was gathered from cases and controls through a structured questionnaire survey. All individuals in the control group were clinically assessed to be without SLE, SSc, other autoimmune diseases, other skin diseases, or any family members (including first, second, and third degree relatives) who have the history of autoimmune and skin disorders. The study was approved by the Institutional Ethical Committee of Xiangya Hospital (201403157) and was conducted according to the Declaration of Helsinki principles. All participants provided written informed consent.
Each PCR reaction with 20 ul of total volume contained 1 x PCR buffer, 0.2 mM each of four deoxy-nucleotides, 1 mM each of forward and corresponding reverse primers, 0.3 units of high fidelity Taq polymerase (Roche Diagnostics, Indianapolis, IN, USA), and 50 ng genomic DNA. The PCR reaction was performed using the following thermal cycling conditions: 1 cycle at 95˚C for 2 min, 30 cycles of 95˚C for 30 s, 65˚C for 60 s, and 72˚C for 40 s respectively, and finally incubated for 10 min at 72˚C. The fragments of the PCR products were visualized in 2.0% agarose gel under UV light and recorded by photograph. The SNP polymorphism at rs14042 was detected with presence or absence of PCR product, and the indel polymorphism (rs267607656, -/GGCTCCGGAGGTAGCAGCTAC)with or without 21-nucleotide deletion in exon 9 of the KRT1 gene was demonstrated with the different product size(s) in the electrophoresed gels.

Haplotype determination
The haplotypes formed by two polymorphic sites of rs14024 and rs267607656 at KRT1 gene were estimated by Haploview software in version 4.2 (Broad Institute of MIT and Harvard, Cambridge, Barrett et al., 2005). The name of haplotypes with two letters (G-L, A-L or A-Del) represent the combination of two polymorphic sites of KRT1 gene.

Statistical analysis
Allele and genotype frequencies of SNP and indel polymorphism were calculated. Hardy-Weinberg Equilibrium (HWE) testing was performed using Arlequin version 3.11 [28]. Statistical significance between patients and controls was determined using a standard chi-square test (SPSS 18.0 software, IBM Corporation, NY, United States), or 2-df chi-square test (3 X 2). SNP data were edited into Haploview format [29] and the linkage disequilibrium (LD) and haplotype map were obtained using Haploview version 4.2. The statistical power (1-β) was calculated using PS Software (Power and Sample Size Calculation) [30]. A p value less than 0.05 was considered significant.

Validation of KRT1 genotyping assay
The genotypes of KRT1 were determined based on two polymorphic sites within exon 9. The genotyping of 6 reference samples with known-genotypes, which were typed using Sanger Sequencing of PCR products, were given in Fig 1. SNP rs14024 with A or G or both at position 388 of KRT1 exon 9 and indel genotypes at rs267607656 were determined with sequencing profiles (Fig 1a). The reference samples were used to validate this novel assay. A 698 bp of GAPDH gene fragment amplified in each PCR test was used as the internal positive control. Specific PCR product with or without a 410 bp size in gel represented either A or G, or both in the SNP, rs14024 (Fig 1b). 21-nuecletide deletion at rs267607656 of KRT1 gene was identified based on the size difference of PCR product running in 2% gel as Fig 1c. These typing results of 6 reference samples completely matched the results of genotypes with Sanger sequencingbased typing (SBT). The approach of PCR-SSP for KRT1 genotyping was validated.

The frequencies of KRT1 genotypes
The genotyping was assigned by the pattern of PCR reactions as showed in Fig 2. Allele and genotype frequencies of each SNP and the indel polymorphism in healthy Chinese Southern Han (CSH), SLE and SSc patients' population were given in Tables 1 and 2. All were found to be in Hardy-Weinberg Equilibrium (p = 0.08~0.95).
KRT1 haplotype analysis. The haplotype combining two polymorphic sites tested was estimated with conventional estimation and maximization (EM) algorithm (significance level setup to 0.05) among three populations. Three of four possible haplotypes were found. As shown in Table 3, wild type (WT, A-L), combining allele-A at rs14024 and no deletion at rs267607656, deletion mutant (Del, A-S) with allele-A at rs14024 linking to 21-neucletide deletion at rs267607656, and mutant (MU, G-L) with G substitution at rs14024 without deletion  Table 3. None of the haplotype MD (G-S) was observed among three groups, while this haplotype MD combining a mutant "G" and 21-nueculetide deletion was supposed to be found in populations. The linkage disequilibrium (LD) of two sites with two alleles was evaluated in each group. Since three of four possible haplotypes were observed, the results showed that two polymorphic sites were in complete linkage disequilibrium in all three groups (D' = 1 and r 2 = 1), mainly due to the absence of haplotype MD (G-S).

Haplotype of KRT1 and disease association
Considering two polymorphic sites as a haplotype of KRT1, the haplotypic frequencies of wild type (WT, A-L) were not different in SLE and SSc groups against control (p = 0.07 and 0.34, respectively, Table 3). The frequencies of the mutant haplotype (KRT1-MU, G-L) were significantly higher in SLE (57.62% vs. 44 Table 3) than that in the control group. Since the frequency of haplotype Del was significantly lower in the SSc group, the deletion of 21 nucleotides at rs267607656 might associate with resistance to, or protection from SSc.

Haplogenotype of KRT1 and disease association
Hypothesizing that the haplotypes of KRT1-Del (low risk) and KRT1-MU (high risk) haplotypes have opposite effects to associate with SLE and SSc as above, the patients were further classified into three sub-groups according to the individuals having Del and no MU genotype (Del+/MU-, low risk alone), having no Del but with MU genotype (Del-/MU+, high risk either A, G or both at rs 14024 is read based on the presence of a 410-bp band in PCR products in different primer pairs. (c): The indel polymorphism in rs267607656 of KRT1 is detected by the sizes of PCR gels with larger bands (L, no deletion) and/or small bands (S, with deletion). The relatively small size (S) detected is stand for the existing of a 21-bp deletion in KRT1 gene.

Discussion
After reviewing the SNPs of KRT1 and their allelic frequencies in populations, two polymorphic sites in exon 9 of KRT1 are highlighted for the analysis in our investigation. SNP rs14024 encodes a nonsynonymous amino acid change from a lysine to arginine in exon 9 because of the codon replacing from AAG to AGG). The coded amino acid is positioned in the carboxy tail region of KRT1 molecule and this location is away from the highly conserved α-helical rod domains in the center of the KRT1 protein [19]. Indel polymorphism at rs267607656 with two size variants of human keratin 1 protein chain described by Korge [31] is characterized by a deletion of 7 amino acids in a repeatable sequence of the transmembrane subdomain of KRT1. This deletion corresponds to loss of an entire glycine loop of seven amino acids. The PCRbased KRT1 allele typing showed the different patterns with three PCR products (Fig 1).
Our data also indicate that both polymorphic sites frequently varied in our local population (Table 1). Our data demonstrated that the allele of rs14024-G was more frequent in patients with SLE and SSc than that in the local normal population (p = 6.48×10 −5 and p = 8.75×10 −5 , Table 1). And the analysis of genotypes of the SNP rs14024 showed that the G/G was associated with SSc (OR: 2.04, 95%CI: 1.29-3.24, p = 0.002). The frequencies of allele with 21bp deletion (S) in KRT1 were lower in SSc patients than that in the local normal population. Moreover, the ratio of deletion-homozygous (S/S) at rs267607656 was also decreased in SSc patient groups comparing to the control (Table 2). Since it was difficult to identify the association of two alleles with the opposite effects, further analysis was performed on tran-cis for haplotypes of KRT1 gene.
Possible four haplotypes were supposed to be observed if two sites with two alleles in each. Our results showed that the haplotype MD (G-S) ( Table 3) was not found in any of the samples tested. The same results have been reported by Dr. Stastny's group [25]. It might be due to the short distance between two sites at KRT1 gene, as the crossover might not happen. In other possibility this kind of haplotype might be lethal. Our data indicated that mutation allele rs14024-G was high risk to associate with SLE (p = 6.48×10 −5 ) and SSc (p = 8.75×10 −5 ) ( Table 1), but the frequency of deletion (S) at rs267607656 showed the resistant trend to SSc (OR: 0.49, 95%CI: 0.33-0.74, p = 4.89×10 −4 , Table 1). These results generated our hypothesis: haplotype MU (G-L) is susceptible to SLE and SSc and haplotype Del (A-S) is resistant to them. To analyze the data with this idea, we clearly showed that the frequency of haplogenotype of Del-/MU+ was significantly higher in SLE (p = 0.001) and SSc (p = 2.32×10 −4 ) than that in control (Table 4). In contrast, the haplogenotype with a 21-base deletion but not with G mutation (Del+/Mu-) was significantly lower in SLE and SSc than that in control (p = 6.24×10 −5 , p = 0.001, respectively, Table 4). As expected, no difference was observed between the control and the patients with Del+/MU+ haplogenotype, as having an opposite effect with two haplotypes. This detail analysis avoided the neutralization or counteraction with opposite effect on tran-cis polymorphic sites. Based on the GWAS study, Protein C-ets-1, DNA-binding protein Ikaros, Ras guanylreleasing protein 3, Solute carrier family 15 member 4, TNF-α-induced protein 3-interacting protein 1 (TNIP1), 7q11.23, 10q11.22, 11q23.3 and 16p11.2 were identified associating with SLE susceptibility in a Chinese Han population [5]. A slice of genetic loci were also confirmed associating with SSc susceptibility by GWAS, such as CD247 [32], TNIP1, Psoriasis susceptibility 1 candidate gene 1 protein, and Rho-related GTP-binding protein RhoB [33]. Among them, several variants were reported involving in the proliferation of keratinocyte [34,35]. As a major protein produced by keratinocyte, keratins were reported participated in the progress of inflammation [20]. Given the skin inflammation was a common symptom of SLE and SSc, we genotyped the KRT1 and explored their association.
Up to now, GWAS have identified more than 80 SLE susceptibility loci [36] and 40 SSc risk genetic regions [37]. Nevertheless, the causal mutations responsible for the diseases remain elusive, it is necessary to conduct a more detailed investigation of the individual gene. Among the founded susceptibility loci, variants on Interferon-induced helicase C domain-containing protein 1 (IFIH1) and Complement C1r subcomponent (C1R) attracted our attention. IFIH1 is a double-strand RNA receptor and play a vital role in innate immune response during virus infection by inducing the expression of interferon. [38] Polymorphisms in IFIH1 were also shown association with Aicardi-Goutières syndrome [39], type 1 diabetes mellitus [40] and psoriasis [41]. It was reported that IFIH1 mutations resulted in the aberrant induction of type I interferon, which was an etiology of Aicardi-Goutières syndrome [42]. Previous studies had shown that the severity of SLE was associated with the level of type I interferon [43,44],consistent with other researcher's finding that the variants of IFIH1 were a genetic risk for SLE [45,46]. The levels of serum complement are reported to link to the pathogenesis of SLE [47] and Demirkaya E et al. shown that the variant in C1R resulting in a dysfunction complete C1r lead to the occurrence of the SLE [48]. The study of a single gene is conductive to explore more detail information about pathogenesis.
Our data showed that the association between two KRT1 polymorphic sites and effects of autoimmune diseases SLE and SSc. More work need to be done to elucidate the genetic diversity of KRT1 and the association with immune diseases. Some of such research work had been done and were reported by others. Pant and colleagues [49] found the KRT1 has extreme allele-specific expression differences in human white blood cells. This differential allelic expression of KRT1 is predominantly controlled by cis-regulatory polymorphism(s) in strong linkage disequilibrium with the gene.
Recently study by Roth et al. revealed a KRT1-mediated gene expression signature similar to atopic eczema and psoriasis, suggesting a functional link between KRT1 and human inflammatory skin diseases [20], and Wallace [50] reported that deletion of K1/K10 affects desmosomal structure and nuclear integrity. It is interesting to understand whether these specific alleles of KRT1 affect the structure of the encoded protein and hence its functions. This information is very useful for us to explore that KRT1 gene variety involves in the pathogenesis of SLE, SSc and other autoimmune diseases. However, expanding our study in a large cohort is needed to confirm our finding.
The data generated from our finding and others' investigation might lead us to measure the amount of the expression of different haplotype of KRT1 in the further study. In our previously study [51], we had detected the antibodies against KRT1 among patients with autoimmune diseases using the antigens encoded by three common KRT1 alleles mentioned in this paper. The distinct epitopes encoded by the specific KRT1 alleles are identified to serve as targets for the antibodies against KRT1 molecules in patients. Therefore, typing for KRT1 genotypes could improve the diagnosis for SLE and SSc, and helpful to identify high risk individuals with potential genetic background. Moreover, it might be benefit to development of genetic therapy for such autoimmune diseases.