Article Text
PDF
Abstract
Objectives X chromosome has been considered as a risk factor for SLE, which is a prototype of autoimmune diseases with a significant sex difference (female:male ratio is around 9:1). Our study aimed at exploring the association of genetic variants in X chromosome and investigating the influence of trisomy X in the development of SLE.
Methods X chromosome-wide association studies were conducted using data from both Thai (835 patients with SLE and 2995 controls) and Chinese populations (1604 patients with SLE and 3324 controls). Association analyses were performed separately in females and males, followed by a meta-analysis of the sex-specific results. In addition, the dosage of X chromosome in females with SLE were also examined.
Results Our analyses replicated the association of TMEM187-IRAK1-MECP2, TLR7, PRPS2 and GPR173 loci with SLE. We also identified two loci suggestively associated with SLE. In addition, making use of the difference in linkage disequilibrium between Thai and Chinese populations, a synonymous variant in TMEM187 was prioritised as a likely causal variant. This variant located in an active enhancer of immune-related cells, with the risk allele associated with decreased expression level of TMEM187. More importantly, we identified trisomy X (47,XXX) in 5 of 2231 (0.22%) females with SLE. The frequency is significantly higher than that found in the female controls (0.08%; two-sided exact binomial test P=0.002).
Conclusion Our study confirmed previous SLE associations in X chromosome, and identified two loci suggestively associated with SLE. More importantly, our study indicated a higher risk of SLE for females with trisomy X.
- lupus erythematosus, systemic
- polymorphism, genetic
- autoimmune diseases
Data availability statement
Data are available upon reasonable request. Our summary statistic analysis is available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
SLE exhibits a bias towards females, and the X chromosome has been identified as a risk factor for the sex bias observed in SLE.
WHAT THIS STUDY ADDS
This study confirmed the association of TMEM187-IRAK1-MECP2, PRPS2, TLR7 and GPR173 loci in X chromosome with Asian SLE and identified two X chromosome loci (LOC389895-SOX3 and CT83-KLHL13) suggestively associated with SLE. In addition, this study showed a higher risk of SLE for females with trisomy X in Asian population.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study may provide insights into the sex bias observed in SLE.
Background
Sex bias towards females is commonly found in autoimmune diseases.1 SLE is a multi-organ affected autoimmune disease with a significant sex difference (female:male ratio is around 9:1).2 Although oestrogen is considered a sex disparity factor, the molecular mechanism that delineates genetic bias remains unclear.
Variations in the dose of X chromosome have been proposed as a genetic factor for the sex biased observed in SLE and other autoimmune diseases.3 4 Notably, male patients with Klinefelter syndrome (47,XXY) have an increased risk of developing SLE by four-fold to eight-fold compared with healthy males, suggesting that disease pathogenesis might be related to the copy number of X chromosome.5 In addition, significant increased prevalence of trisomy X (47,XXX; 7 of 2826 patients with SLE) were found in American-European and African patients with SLE.6 The severity of SLE was also found linked to doses of X chromosome.7 However, only a few of related studies have been reported in Asian population.8 Thus, further studies are needed to confirm the significance of X chromosome dose in Asian population.
In addition, genetic variants in X chromosome were also found to be associated with SLE. Variants in TMEM187-IRAK1-MECP2,9 10 PRPS2,11 TLR7,12 Cxorf21 (TASL)13 and GPR173 loci14 have been found significantly associated with SLE. Among them, the associated variants in TLR7, PRPS2 and GPR173 loci tended to have a stronger effect on SLE in males as compared with females.11 12 14 Thus, genetic variants in X chromosome may also contribute to the sex bias observed in SLE.
In this study, we aimed to examine the association of genetic variations in X chromosome with SLE and the prevalence of trisomy X in SLE by using the data collected from both Thai15 and Chinese populations.16 We successfully replicated previous X chromosome-wide association studies (XWAS) findings and identified potentially novel SLE-associated loci in X chromosome. Furthermore, we confirmed an increased frequency of 47,XXX in Asian patients with SLE.
Methods
Subjects
Infinium Asian Screening Array-24 V.1.0 (ASA) BeadChip was used for genotyping of the Thai samples. Genotypes were called by the Genome Studio data analysis software V.2011.1 (Illumina, San Diego, California, USA). The dataset included 835 SLE cases (778 females and 57 males).16 All patients were recruited from Chulalongkorn Memorial Hospital and Faculty of Medicine, Ramathibodi Hospital, during 2018–2019 (EC no. 590223, Protocol-ID 12-58-12). In addition, 2995 geographically matched controls (1554 females and 1441 males) were obtained from the randomly selected samples from the fourth Thai National Health Examination Survey,17 which were previously genotyped jointly by Faculty of Medicine Ramathibodi Hospital, Mahidol University and Department of Medical Sciences, Ministry of Public Health of Thailand.
The data from Chinese population were retrieved from our previous studies,18 including a total of 1604 patients with SLE (1478 females and 126 males) and 3324 geographically matched controls (1760 females and 1564 males). These samples were genotyped by Infinium OmniZhongHua, the Infinium Global Screening Array-24 V.2.0 and the ASA. To overcome potential influence of varied version of genotyping arrays, variants were removed if their allele frequencies were significantly different across different arrays among the control samples. Details of the correction processes were described in our previous publication.18
Quality controls and X chromosome-wide association studies
Procedures of quality control for samples were similar to our previous studies,16 18 samples with identity-by-descent (IBD) ≥0.125, missing genotypes ≥5% and discordant sex identity were adjusted. In addition, low-quality variants in X chromosome, indicated by genotype calling rate <95%, departuring from Hardy-Weinberg equilibrium in females (p<10−4) and presenting differential missingness between males and females (p<10−7), were excluded. Single-nucleotide polymorphisms (SNPs) with minor allele frequencies <1% in both males and females were also discarded. After the quality control procedure, the data were imputed into the density of the TOPMed (Trans-Omics for Precision Medicine) reference panel using the online server.19 SNPs with imputation quality >0.7 were kept for further analysis.
We performed association studies in the Thai and Chinese datasets separately. In each dataset, association analyses for females and males were also conducted separately with three principle components as covariates using RVtests.20 The summary statistics for these data groups were further combined by METAL programme using inverse variance-weighted method (online supplemental figure S1).21 The inflation factors for the association results were calculated by using R packages ‘qq-man’ after pruning. Locus zoom was used to visualise the association signals in each locus (https://my.locuszoom.org).22
Supplemental material
Functional annotation
SNPs with association p<1E-05 in the meta-analysis were annotated using RegulomeDB V.2.23 Epigenomic data across the immune-related cell types and organs were retrieved from Roadmap Epigenomics project24 and visualised by the WashU Epigenome Browser.25
SLE subphenotype association analysis
Known SLE-associated variants on the X chromosome were initially evaluated in the Thai dataset. We further evaluated their associations with varied subphenotypes of SLE, including arthritis (n=268), haematological disorders (n=358), neurological disorders (n=90), rash (n=352), renal involvement (n=409), serological disorders (n=271) and ulcers (n=159), compared with patients with SLE without these symptoms. Similarly, association analyses were first conducted for females and males separately with three principle components as covariates using RVtests, then the summary statistics were combined by METAL using inverse variance-weighted method.
Identification of trisomy X (47,XXX)
Trisomy X (47,XXX) for each individual was detected through examining the fluorescence intensity of the B allele across all SNPs in X chromosome. Genome Studio data analysis software V.2011.1 (Illumina) was used to quantify the intensity of B allele for SNPs. Scatter plots of B allele intensity were then generated for each individual, with x-axis indicating SNPs across X chromosome and y-axis indicating the intensity of B allele for the SNP. For a SNP with BB homozygous genotype, the expected intensity is 100% in the scatter plot, while the expected intensity is 0% for a SNP with AA homozygosity. In addition, the expectation is 50% for a SNP with AB heterozygosity.
Thus, for females who carried two copies of X chromosome (46,XX), we expected to observe ‘three-band’ pattern in the scatter plot, indicating three groups of the SNPs with different genotypes (AA, AB and BB). However, for females with trisomy X (47,XXX), they have three alleles for each SNP, resulting in four possible genotypes (AAA, AAB, ABB, BBB) in the X chromosome. In this case, we expected to observe four bands in the scatter plot, indicating 0%, 33%, 66% and 100% of intensisty of B allele (online supplemental figure S2).
Results
Meta-analysis of XWAS from the Thai and Chinese datasets
After the procedures of quality control and imputation, we performed XWAS in Thai and Chinese datasets, which included a total of 2439 cases and 6319 controls. Considering the sex difference in X chromosome, we performed the association tests in females and males separately in each dataset. We then combined these summary statistic by using the inverse variance-weighted method. The inflation factor (λgc) for the meta-analysis results was estimated to be 1.09 (online supplemental figure S3).
Our analyses successfully replicated previous findings in TMEM187-IRAK1-MECP2, TLR7, PRPS2 and GPR173 loci (pcombined<0.01 and consistent direction of association in both datasets) (figure 1 and table 1). Leveraging differences in linkage disequilibrium (LD) patterns, analysis of XWAS from Thai and Chinese datasets indicated that SNP rs13397 (OR=0.70, pcombined=4.669E-18) in TMEM187-IRAK1-MECP2 locus was a likely causal variant, with a shared signal between both populations (figure 2A). rs13397 is a synonymous variant of TMEM187 and epigenomic data suggested that it is located in an active enhancer across multiple immune-related cell types (monocytes, B and T cells) and immune-related organs (thymus and spleen) (figure 2B). RegulomeDB (V.2)23 also annotated this one as a potential regulatory variant (with a score of 1f). Moreover, eQTL (expression Quantitative Trait Loci) results from the GTEx project26 showed that the disease risk allele rs13397-A was significantly associated with decreased expression level of TMEM187 in whole blood (p=2.23E-34; figure 2C).
Manhattan plot of the meta-analysis of X chromosome association study. The green dots (
) highlight candidate novel loci associated with SLE; the red dots (
) indicate the known SLE-associated loci.
List of known and potential SLE-associated variants in X chromosome
Fine-mapping analyses for SLE association in TMEM187-IRAK1-MECP2 locus. (A) SLE association signals in the locus from the Thai (y-axis) and Chinese dataset (x-axis). (B) Epigenomic profile in the locus represented by the 15-state model (ChromHMM). (c) eQTL violin plot for TMEM187, retrieved from the GTEx project. eQTL, expression Quantitative Trait Loci.
In addition, a variant in 3’ untranslated region (3’ UTR) of TLR7 was also found significantly associated with SLE in our data (rs3853839; OR=1.36, pcombined=1.35E-11). Although PRPS2 and TLR7 loci are distantly close with each other, our data demonstrated that rs7059565 in PRPS2 locus was independently associated with SLE (OR=1.22, pcombined=6.05E-07), as the LD (R2) with rs3853839 was <0.001 in either Thai or Chinese datasets. Besides, we also repeated the association of GPR173 locus (rs12011862; OR=1.13, pcombined=0.004), known SLE X-linked loci, in our results.
Our data also found some loci suggestively associated with SLE. For example, SNP rs424153 in LOC389895-SOX3 locus (OR=0.83, pcombined=2.17E-06) and SNP rs34418148 in CT83-KLHL13 locus (OR=0.84, pcombined=3.22E-06) passing the suggestive significance thresholds (5E-05; table 1). Reduced expression of Sox3 were found associated with a lupus-like mouse model.27 The KLHL13 gene encodes a substrate specific to B-cell receptors, playing an essential role in mitotic progression and cytokine production.28 However, these findings need to be further confirmed in another cohort.
Examination of SLE subphenotype association in X chromosome
Considering the heterogenous manifestation of SLE, we examined the subphenotype association in those known associated loci. As a comprehensive clinical record is available for Thai samples, we performed the subphenotype association in Thai patients with SLE. Unfortunately, we did not find any variants significantly associated with specific organ damage (table 2), due to limited power.
Association of known SLE-assciated loci with varied subphenotypes in the Thai dataset
Detection of trisomy X (47,XXX)
To evaluate the impact of trisomy X in Asian patients with SLE, we futher detected the samples with trisomy X in our datasets (see ‘Methods’ section). For the analysis, we used raw intensity data, comprising 2231 females with SLE (1453 with Chinese ancestry and 778 with Thai ancestry) and 2410 female controls (856 with Chinese ancestry and 1554 with Thai ancestry). We identified five cases of SLE with trisomy X, and all were detected within the Chinese dataset (figure 3A). In contrast, two samples with 47,XXX were identified in the control group (figure 3B).
Identification of trisomy X (47,XXX) in either SLE (A) or control (B) groups, representing B allele frequency.
The frequency of 47,XXX was approximately 0.22% (5/2231) in females with SLE in our datasets, with a higher frequency in the Chinese dataset (0.34%, 5/1453 females with SLE). In contrast, the frequency of trisomy X among female controls was around 0.08% (2/2410), which was very close to previous reports (1 in 1000 female births) in European population.29 Our data indicated that the frequency of trisomy X was significantly higher in females with SLE than controls (two-sided exact binomial test p=0.002).
Previous findings in individuals with European and African-American ancestries identified trisomy X in 7 of 2826 (0.25%) females with SLE but in 2 of 7074 female controls (0.03%).6 Integrating these findings with our results, a significant enrichment of trisomy X in SLE can be observed (12/5057 females with SLE; OR=5.64, two-sided Fisher’s exact test p=0.001). However, the ancestral differences require additional investigation in future studies.
Discussion
In this study, we conducted XWAS and examined the influence of trisomy X in SLE using data from both Thai and Chinese populations. According to our XWAS results, we successfully replicated several loci in X chromosome, including the association of TMEM187-IRAK1-MECP2 complex region,9 10 PRPS2,11 TLR712 and GPR173 loci14 with SLE. Among these identified variants, the TMEM187-IRAK1-MECP2 region has been consistently replicated across various nationalities.14 30 In contrast, PRPS2 and TLR7 appear to be more prevalent in Asian SLE populations.31 Another variant at CXorf21 (TASL) locus, found in European populations,13 was excluded during the quality control process. Consequently, we were unable to validate its association. Moreover, our data indicate suggestively associations of LOC389895-SOX3 and CT83-KLHL13 loci with SLE. However, these results need to be further confirmed.
Making use of the LD difference between Thai and Chinese populations, we prioritised a synonymous variant in TMEM187 (rs13397) as a likely causal variant. We initially pinpointed the location of this variant and discovered it within an active enhancer region of immune-related cells. We suggested that this allele may link to a reduced expression level of TMEM187 according to GTEx eQTL data. This variant was also reported to be associated with rheumatoid arthritis, another type of autoimmune disease,32 suggesting a shared immunopathogenesis across multiple autoimmune diseases.
As several genes have been shown to skew X chromosome inactivation,3 our results identified variant on TLR7 reported to influence this skewing. The TLR7 gene is known to escape from X inactivation in immune cells.33 Our analysis identified the rs3853839 in the 3’ UTR of TLR7. This variant is known to augment TLR7 mRNA transcription and more pronounced interferon signature.12 Interestingly, studies of effect of rs3853839 specifically in B cells found that this polymorphisms was associated with more active disease and CD19+CD24+CD38+ newly formed transitional B cells enhanced inflammation.34 Contrary to a previous study that linked rs3853839 with lupus nephritis in the Egyptian population,35 our subphenotype analysis found no association between rs3853839 and clinical phenotypes in the Thai population. Thus, the lupus nephritis-specific association needs to be further confirmed in the future study.
In addition, we found trisomy X in approximately 0.22% (5/2231) of Asian females with SLE (Chinese and Thai), with a higher frequency of 0.34% (5 out of 1453 females) in the Chinese subgroup. However, the small sample size limits our ability to fully explore heterogeneity across different ancestries. The observed frequency of trisomy X in our controls aligns with previous findings, which reported trisomy X in 1 of every 1000 females.6 Extra X chromosomes in childhood-onset SLE, a more severe and chronic form, were also noted.36 Detecting 47,XXX may be a useful predictor for SLE in newborn babies.
Conclusion
Our findings suggest that the X chromosome plays a role in SLE development. We successfully replicated previously identified SLE-associated loci on the X chromosome and discovered two loci that are suggestively associated with SLE. Notably, our study confirms that an extra X chromosome can predispose individuals to SLE in Asian populations, similar to what is observed in European and African-American populations.
Data availability statement
Data are available upon reasonable request. Our summary statistic analysis is available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The protocols for the SLE study of the Thai population were approved by the ethics committee of the Faculty of Medicine, Chulalongkorn University and Faculty of Medicine, Ramathibodi Hospital, Mahidol University (EC no. 590223, Protocol-ID 12-58-12). Informed consent was obtained from patients according to the Declaration of Helsinki. The SLE study of Chinese protocols was approved by the ethics committee of the Faculty of Medicine, Hong Kong University (EC no. UW07-119).
Acknowledgments
We thank all the clinicians who helped us to collect the clinical data and anonymous reviewers for their insightful comments and suggestions, which are essential to improve the quality of our study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
PT and YL contributed equally.
Correction notice This article has been corrected since it was published. Missing details in the funding information has been added.
Contributors PT and WYF conceived of the study. YL and KJ conducted the quality control and association analyses in Chinese and Thai, respectively. WYF analysed eQTL. YL, PK and MB performed the imputation. YW, JE, NJ, WA, NN, NS, SM, PS, YY and NH collected samples. Y-FW is responsible for the overall content as the guarantor. All authors read and contributed to the manuscript.
Funding This research is funded by Thailand Science Research and Innovation Fund Chulalongkorn University (6646/2566), Chulalongkorn University (RinUni_65_01_37_02) and partially supported by the Asahi Glass Foundation (RES_65_012_37_001). PT and KJ have received research assistance funding from Chulalongkorn University (GCUGE17). NH has received funding from the Ratchadapiseksompotch Fund, Faculty of Medicine, Chulalongkorn University, grant number (RA-MF-22/66). Y-FW recognises the research start-up from the Chinese University of Hong Kong, Shenzhen (CUHK-Shenzhen; K10120220256 and UDF01002831/UF02002831), Shenzhen-Hong Kong Cooperation Zone for Technology and Innovation, CUHK-Shenzhen Futian Biomedical Innovation R&D Center (HZQB-KCZYB-2020056) and Ganghong Young Scholar Development Fund (E10120210019 and E10120220220).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.