Skip to main content
Download PDF

Phenomic landscape and pharmacogenomic implications for HLA region in a Taiwan Han Chinese population

Biomarker Research volume 12, Article number: 46 (2024) Cite this article

Abstract

Background

The human leukocyte antigen (HLA) genes, exhibiting significant genetic diversity, are associated with susceptibility to various clinical diseases and diverse in drug responses. High costs of HLA sequencing and the population-specific architecture of this genetic region necessitate the establishment of a population-specific HLA imputation reference panel. Moreover, there is a lack of understanding about the genetic and phenotypic landscape of HLA variations within the Taiwanese population.

Methods

We created models for a Taiwanese-specific HLA imputation reference panel. These models were trained with the array genotype data and HLA sequencing data from 845 Taiwanese subjects. HLA imputation was applied for 59,448 Taiwanese subjects to characterize the HLA allele and haplotype frequencies. Additionally, a phenome-wide association study (PheWAS) was conducted to identify the phenotypes associated with HLA variations. The association of the biallelic HLA variants with the binary and quantitative traits were evaluated with additive logistic and linear regression models, respectively. Furthermore, an omnibus test with likelihood-ratio test was applied for each HLA amino acid position in the multiallelic HLA amino acid polymorphisms to compare the difference between a fitted model and a null model following a χ2 distribution of n-1 degree of freedom at a position with n residues. Finally, we estimated the prevalence of adverse drug reactions (ADR)-related HLA alleles in the Taiwanese population.

Results

In this study, the reference panel models displayed remarkable accuracy, with averages of 99.3%, 98.9%, and 99.1% for 2-, 4-, 6-digit alleles of the eight classical HLA genes, respectively. For PheWAS, a total of 18,136 significant associations with HLA variants across 26 phenotypes are identified (p < 5×10-8), highlighting the pleiotropy feature of the HLA region. Among the independent signals, 15 are novel, including the association of HLA-B pos 138 variation with ankylosing spondylitis (AS), and rs9266290 and rs9266292 with allergy. Through an analysis spanning the entire HLA region, we identified clusters of phenotype correlations. Finally, the carriers of pharmacogenomic related HLA alleles, including HLA-C*01:02 (35.86%), HLA-B*58:01 (20.9%), and HLA-B*15:02 (8.38%), were characterized in the Taiwanese general population.

Conclusions

We successfully delivered the HLA imputation for 59,448 Taiwanese subjects and characterized the genetic and phenotypic landscapes of the HLA variations. In addition, we quantified the estimated prevalence of the ADR-related HLA alleles in the Taiwanese population. The developed HLA imputation reference panel could be used for estimation of population HLA allele frequencies, which can facilitate further studies in the role of HLA variants in a wider range of phenotypes in the population.

Background

The human leukocyte antigen (HLA) corresponds to a cluster of genes located on the chromosome 6p21 that encode the major histocompatibility complex (MHC) in humans [1]. These genes play a pivotal role in differentiating self and non-self by presenting a vast amount of antigen peptides to T cells [2]. The HLA genes exhibit an incredible diversity, with more than 35,000 alleles [3]. This intensive polymorphism enables the immune system to fight against a variety of pathogens and diseases. Indeed, the HLA region has been considered as a critical area associated with a wide spectrum of diseases, including type I diabetes, rheumatoid arthritis (RA), and ankylosing spondylitis (AS) [4, 5]. Moreover, the HLA variations have been repeatedly alerted to adverse drug reactions, including abacavir hypersensitivity and carbamazepine-induced Stevens–Johnson syndrome [6,7,8,9]. These studies indicate the pleiotropy of HLA and highlight the importance to investigate the HLA variations in different phenotypes.

Phenome-wide association studies (PheWAS) is research tool to investigate the association of genetic variances across extensive phenotypes. PheWAS is a transposition of GWAS. Unlike GWAS that test the associations of the genetic variants across the genome with a specific phenotype, PheWAS aims to identify the phenotypes associated with certain genetic variants [10, 11]. GWAS has been widely applied to identify disease predisposition genes and pathways involved in complex human traits [10]. On the other hand, PheWAS can be applied to study the shared mechanisms across different traits and to identify the target diseases for drug repurposing [11]. In addition, PheWAS has become feasible and increasingly popular with the emergence of biobanks of densely phenotyped cohorts or link to the participants’ electronic health records. Taiwan biobank TWB has enrolled more than 150 thousand individuals since its establishment in 2012. The comprehensive phenotypes collected with questionnaires and physical and biochemical examinations make it an ideal resource for phenome-wide investigation [12].

Genetic and phenotypic characteristics within the HLA region have been investigated in European, Japanese and Korean populations [13,14,15,16]. However, the structure of linkage disequilibrium (LD) is extremely complex in the MHC region and the complexity leads to diverse haplotypes across different populations [4]. In addition, previous genomic study of TWB revealed a distinct cluster with unique HLA pattern, indicating high degree of LD between HLA haplotype A*33:03-B*58:01 [17]. Furthermore, previous studies showed that certain HLA-phenotype associations can be replicated, while others are population specific [13,14,15,16]. These give prominence to the need of population specific investigation in HLA region.

Direct sequencing-based typing (SBT) is the gold standard for HLA typing in clinical settings. Next-generation sequencing (NGS) -based typing has shown high concordance with SBT [18]. However, sequencing-based HLA typing is labor-intensive, expensive, and not widely applied in large cohorts. On the other hand, extensive genotype data in huge cohorts has been collected with the establishment of biobanks around the world. Furthermore, a variety of strategies to impute HLA alleles from genotype data have been developed to enable the investigation of HLA variations in large cohorts [19]. To extend the use of HLA imputation and to account for the population diversity in HLA region, it is essential to develop population-specific HLA imputation reference panels in different populations.

Recently, Y.H. Huang et al. constructed a Taiwanese-specific HLA imputation reference. This reference panel was applied to the genotype data from 23,972 Taiwanese subjects. Importantly, the results successfully validated the association of HLA variations with RA in the Taiwanese population [20]. However, the investigation of comprehensive phenotypic landscape of the HLA variants was not conducted in Taiwan Han Chinese population yet. In addition, the reference panel constructed in that study was based upon the genotype data of TWBv1 array, which was the first custom single-nucleotide polymorphism (SNP) array designed for genotyping in Taiwan Biobank (TWB) [12, 20]. The TWB has designed another SNP array, namely the TWBv2 array, to include rare coding risk alleles and more variants across the HLA region, and has released an expanded cohort of TWBv2 genotype data [12, 21]. Given these factors, we aimed to develop a Taiwanese-specific HLA imputation reference panel based upon the TWBv2 array for HLA imputation in an expanded cohort of Taiwan Han Chinese population. Moreover, with the imputed HLA alleles, we aimed to investigate the genetic features of the HLA region and the phenotypic characteristics associated with HLA variations in a Taiwan Han Chinese population.

Methods

Study design

A schematic overview of the study is shown in Fig. 1 (created with BioRender.com). The Taiwan Biobank (TWB) data was harnessed in this study and the predominant majority of the TWB subjects are Han Chinese descents [12, 21]. To construct the Taiwanese-specific HLA imputation references using the TWB subjects with both the TWBv2 array genotype data and the next generation sequencing (NGS)-based HLA typing data. Furthermore, the reference panels were applied for HLA imputation on the TWB subjects with TWBv2 array genotype data. With this imputed cohort, we characterize the frequencies of HLA alleles and haplotypes in Taiwanese population. Moreover, a PheWAS was conducted to investigate the associations between HLA variations and 109 traits in 59,448 subjects. We also estimated the HLA region-wide heritability and genetic correlations of the phenotypes. Finally, we assessed the prevalence of adverse drug reactions (ADR)-related HLA alleles in the Taiwanese population. This study was conducted in accordance with The Code of Ethics of the World Medical Association. Deidentified data of human subjects was applied from the Taiwan Biobank, Academia Sinica. This study was approved by the Institutional Review Board of Taipei Medical University (TMU-JIRB N201906005). The access to and use of the Taiwan Biobank resource in this research was approved by the Ethics and Governance Council (EGC) of Taiwan Biobank (approval number: TWBR10906-03).

Fig. 1
figure 1

Schematic overview of this study

Full size image

Data resources and processing

In this study, individual genotype, sequencing-based HLA typing, and phenotype data of subjects enrolled to the TWB between 2012 and 2019 was used for analysis. Subjects without TWBv2 array genotype data were not included. Quality control, phasing and genome-wide imputation of the genotype data were conducted by TWB, as described in the previous study [12]. Overall, 16,222,535 variants in 68,948 subjects with the imputation INFO score > 0.3 were initially included in this study. Furthermore, post-genome-wide imputation quality control (QC) filtering was applied using PLINK 1.9. First of all, low quality variants meeting any of the following criteria were excluded: (1) variant call rate < 0.98; (2) minor allele frequency (MAF) < 0.05; (3) deviation from Hardy-Weinberg equilibrium (HWE) p-value < 1×10-6. Furthermore, subjects meeting any of the criteria below were excluded: (1) incorrect sex status; (2) sample call rate < 0.98; (3) outlying heterozygosity rate (outside 3 times of the standard deviation from the mean); (4) Identity by descent (IBD) > 0.1875. After these quality control, a total of 3,982,815 variants and 59,448 subjects remained. Finally, variants located on the MHC region, which is on chromosome 6, ranging from 28,510,120 to 33,480,577 bp based on GRCh38 (accessed in May 2021), were extracted for further analysis. Overall, 59,448 subjects and 29,279 variants, including 27,055 biallelic single nucleotide variations (SNVs) and 2224 indels, were included for analysis.

Amongst these subjects, 845 had NGS-based HLA typing data of eight classical HLA genes (HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DPA1, HLA-DPB1, HLA-DQA1, and HLA-DQB1) available from TWB. The sequencing library was constructed with the NXTypeTM NGS HLA typing kit (One Lambda, Inc., Canoga Park, CA) and the Ion ChefTM system (Thermo Fisher Scientific, Waltham, MA). Furthermore, sequencing data was performed with the Ion S5TM XL sequencer (Thermo Fisher Scientific, Waltham, MA). Finally, the HLA alleles corresponding to the IMGT/HLA database (version 3.29.0.1) were called using the TypeStreamTM software v.1.1.0 (One Lambda, Inc., Canoga Park, CA)[20].

Phenotypes were collected by TWB through physical examinations and a structured questionnaire for health-related information. Furthermore, the blood and urine specimens provided by the TWB participants were subjected to laboratory tests. Detailed procedure of phenotype collection was described in a previous study [12]. In this study, self-reported diseases based on the questionnaire were extracted as binary traits. Individuals with each self-reported disease were defined as cases, and the other subjects were the corresponding controls. For gynecological diseases, female controls were used. Furthermore, quantitative traits were extracted from the results of the physical and laboratory examinations. A total of 109 traits, including 55 binary traits (Supplementary table 1) and 54 quantitative (Supplementary table 2), were included in the association analysis.

Construction of Taiwanese-specific HLA imputation reference panel

In this study, all 845 TWB subjects with both TWBv2 genotype data and NGS-based HLA typing data were included to construct the Taiwanese-specific HLA imputation reference panel. Taiwanese-specific HLA parameter estimates of two-, four-, and six-digits resolution for the eight classical HLA genes were generated from the 845 TWB subjects with both the TWBv2 array genotype and the HLA typing data using the HIBAG R package (version 1.26.1) [22, 23]. SNVs within the 500 kb flanking region of each HLA gene were included and 100 classifiers were adopted for the generation of each parameter estimate. Performance of each HLA parameter estimate was assessed via internal validation by randomly dividing the study subjects into a training set and a validation set at a 7:3 proportion. The parameter estimates generated from the training set were applied for HLA imputation in the validation set. The imputation accuracies were calculated by comparing the direct HLA typing results and the imputed HLA alleles with call threshold ≥ 0.5 for posterior probability in the validation set.

HLA imputation and haplotype frequencies calculation

After construction of the Taiwanese-specific HLA reference panel, it was applied for HLA imputation on the 59,448 TWB subjects with TWBv2 array genotype data using the HIBAG R package (version 1.26.1) [22, 23]. External validation was carried out by comparing the distribution of HLA allele frequencies in our imputed cohort with those in other Taiwanese cohorts obtained from the allele frequency net database (AFND) [24]. The datasets included for comparisons were: (1) MJ Lai et al., consisted of 46682 subjects from the Tzu Chi Taiwan Marrow Donor Registry with HLA-A, B, and DRB1 genotypes determined by the PCR sequence specific oligonucleotide probe (PCR-SSO), PCR sequence specific primer (PCR-SSP), and the sequencing methods [25]; (2) SH Wen et al., consisted of 710 subjects from the Tzu Chi Taiwan Cord Blood Bank with HLA-A, B, and DRB1 genotypes determined by the PCR-SSP method [26]; and (3) PL Chen et al., consisted of 504 subjects with HLA-A, B, C, DRB1, and DQB1 genotypes determined by the PCR-SSO and PCR-SSP methods [27]. Furthermore, haplotypes of the eight classical HLA genes were phased and the frequencies were calculated with haplo.stats (version 1.8.6) R package. In addition, haplotypes of five HLA genes (HLA-A, HLA-C, HLA-B, HLA-DRB1, and HLA-DQB1) were phased to compare the haplotype frequencies with the frequencies in a Han Chinese population [28].

Phenome-Wide Association Study (PheWAS)

In addition to the biallelic SNVs and indels in the HLA region, the following types of HLA variants were included in the PheWAS analysis: (1) imputed HLA alleles in 2-, 4-, and 6-digit resolution; (2) HLA amino acid polymorphisms (e.g. HLA-B position 138); (3) biallelic HLA amino acid polymorphisms of the respective residue (e.g. HLA-B Q94). The amino acid sequences were translated from the imputed HLA alleles using the HIBAG R package (version 1.26.1) and the amino acid polymorphisms were identified from the amino acid sequences [22, 23]. Additional QC after the HLA imputation and amino acid translation were applied as follows: (1) for the HLA alleles and the biallelic HLA amino acid polymorphisms, variants with missing rate >0.05 or MAF < 0.01 were excluded; (2) for multiallelic HLA amino acid polymorphisms, variants with missing rate >0.05 were excluded. In this study, a total of 30,188 variants, including 27,055 biallelic SNVs, 2,224 indels, 269 HLA alleles, 437 biallelic HLA amino acid polymorphisms, and 203 multiallelic HLA amino acid polymorphisms, were included in the association analysis.

The association between the HLA variants and the phenotypes were evaluated using regressions under additive genetic models using REGENIE (version 2.2.4) [29]. The sex, age, age2, and the top 10 principal components (PCs) obtained from the genome-wide genotype data were included as covariates. The association between the biallelic HLA variants and the binary traits were evaluated with an additive logistic regression model. The quantitative traits were first regressed with the covariates. Furthermore, the association of the biallelic HLA variants with the inverse- normal transformed residuals were evaluated using an additive linear regression model. For the multiallelic HLA amino acid polymorphisms, an omnibus test with likelihood-ratio test was applied for each HLA amino acid position to compare the difference between a fitted model and a null model following a χ2 distribution of n-1 degree of freedom at a position with n residues. The genome-wide significant (GWS) threshold of p < 5×10-8 was used in this study. The biallelic SNVs and indels were annotated using ANNOVAR [30]. Furthermore, these variants were assigned to the eight classical HLA genes if they were in high LD of r2 ≥ 0.7 with any HLA allele or amino acid polymorphism on the corresponding HLA genes.

The forward-type conditional regression analysis was applied for the phenotypes with at least one GWS signal to identify the independent association signals using REGENIE (version 2.2.4) [29]. For each phenotype, the top signal and the corresponding HLA alleles and HLA amino acid polymorphisms in high LD (r2 ≥ 0.7) with it were included into the regression model as covariates. The new top signal identified and its high-LD variants were included as covariates in the next iterative step of the conditional regression. The iterative step was repeated until there was no additional independent signal detected. This modified forward type step-wise conditional analysis additionally include HLA polymorphisms, namely the HLA alleles and the HLA amino acid polymorphisms as covariates and could rule out the effect of the same HLA gene and condition the associations attributable to the HLA gene robustly [15, 16]. GWAS Catalog (Accession date: December 21, 2021) was queried to determine if the associations with the biallelic SNVs and indels were novel. The associations were considered novel if no significant signal was identified within 500 kb of the variant with the same phenotype in the GWAS Catalog. Literature review was conducted to determine if the associations of the phenotypes with the HLA alleles or the amino acid polymorphisms were novel.

The region-wide heritability estimates of each phenotype and the pairwise genetic correlations between the phenotypes accounting for the HLA variants (SNVs and indels in the HLA regions, the HLA alleles and the HLA amino acid polymorphisms) were calculated with the univariate and the bivariate Haseman-Elston (HE) regression using GCTA (version 1.93.2beta), respectively [31]. The genetic correlation network was visualized with the igraph R package (version 1.2.7) [32].

Pharmacogenomics of HLA variations

The burden of adverse drug reactions (ADR)-related HLA alleles in the Taiwanese population was assessed by summarizing the proportion of carriers of selected ADR-related HLA alleles in the imputed cohort. The selected ADR-related HLA alleles were collected through literature search [6, 7, 33, 34].

Results

Internal validation of the Taiwanese HLA imputation reference panel

We constructed a Taiwanese-specific HLA imputation reference panel using array-based (TWBv2) genome-wide genotype data and NGS-based HLA typing data from 845 subjects in TWB. While seven subjects had missing data of self-reported parent originality, all the remaining 838 subjects in the reference panel are self-reported Han Chinese descendants with parents that are either Taiwanese Minnan, Taiwanese Hakka, or Chinese immigrants (Supplementary table 3). This panel was composed of 24 models to predict HLA types of the eight classical HLA genes in 2-, 4-, and 6- digit resolutions from genotype data. A total of 581 2-, 4-, 6-digit HLA alleles derived from the NGS-based HLA typing data were subjected to the model training. As shown in Fig. 2A, the reference panel exhibited high imputation accuracies for eight HLA genes (averages of 99.3%, 98.9%, and 99.1% for 2-, 4-, 6-digit HLA alleles, respectively). Call rates of the panel were also assessed for these HLA genes, with averages of 99.5%, 96.7%, and 97.9% in 2-, 4-, 6-digit models, respectively (Fig. 2B).

Fig. 2
figure 2

Internal validation imputation accuracy (A) and call rate (B) of the Taiwanese-specific HLA reference panels. The internal validation imputation accuracy and call rate were calculated after applying a call threshold ≥ 0.5 (n=845)

Full size image

HLA frequencies in the Taiwanese population

The mean age of the TWB HLA imputed cohort (n = 59448) was 50.5 years old (50.51±10.58), and nearly one third of the subjects were male (31.7%). The distribution of HLA allele frequencies in TWB cohorts by NGS-based HLA typing data (n=845) and by HLA imputation of genome-wide genotyping data (n=59,448) were similar across different resolutions (Supplementary figure 1, Supplementary table 4). Furthermore, we compared the HLA allele distributions of the TWB cohorts with additional three Taiwanese cohorts from AFND. The allele frequencies of HLA-A, HLA-B, HLA-C, HLA-DPB1, HLA-DQB1, and HLA-DRB1 genes exhibited similar distributions in our TWB cohorts and in other Taiwanese cohorts (Fig. 3, Supplementary table 5, Supplementary dataset 1).

Fig. 3
figure 3

Comparisons of HLA allele frequencies in the TWB cohorts with the external Taiwanese datasets. Four-digit allele frequencies of HLA-A (A), HLA-B (B), HLA-C (C), HLA-DPA1 (D), HLA-DPB1 (E), HLA-DQA1 (F), HLA-DQB1 (G), HLA-DRB1 (H) are shown in this figure. Top ten alleles of each gene in the TWB cohort after HLA imputation were extracted and other alleles were grouped into the “other” allele for visualization and comparison with the other datasets

Full size image

To further evaluate the performance of the HLA imputation and to characterize the HLA genetic architecture in the Taiwanese population, we analyzed HLA haplotypes and compared the frequencies of the TWB cohort with a Chinese population, which is genetically close to the Taiwanese population. As shown in the Supplementary table 6, the most common four-digit HLA haplotype of eight classical HLA genes was A*33:03-C*03:02-B*58:01-DRB1*03:01-DQA1*05:01-DQB1*02:01-DPA1*01:03-DPB1*04:01, with a frequency of 3.85% in the Taiwanese population. Further comparison of HLA haplotypes of five HLA genes shared by both populations showed that the frequencies of HLA haplotypes in our TWB cohort were similar with that in the Han Chinese population. However, the HLA haplotype frequencies were observed to be slightly divergent across different Han Chinese subpopulations. The distribution of the HLA haplotype frequencies in our TWB cohort were more similar to the southern Han Chinese subpopulation. In both Taiwanese population and southern Han Chinese subpopulation, A*33:03-C*03:02-B*58:01-DRB1*03:01-DQB1*02:01 and A*02:07-C*01:02-B*46:01-DRB1*09:01-DQB1*03:03 were identified as the most common HLA haplotypes (Supplementary table 7).

Phenome-Wide Association Study

To investigate the phenotypic landscape within the HLA region for identifying phenotypes associated with HLA variants, we conducted a PheWAS using the TWB dataset with imputed HLA alleles. A total of 109 phenotypes were included in the analysis. Among them, 55 phenotypes as binary traits were classified into 12 categories, namely immune-related disease, metabolic disease, cardiovascular disease, kidney disease, digestive disease, mental disease, neurologic disease, ophthalmic disease, gynecological disease, bone disease, respiratory disease, and symptom (Supplementary table 1). Furthermore, another 54 phenotypes as quantitative traits fall into 8 categories, which were anthropometric, metabolic, cardiovascular, kidney-related, liver-related, hematological, bone-related, and respiratory (Supplementary table 2).

In the PheWAS, a total of 18,136 GWS associations with HLA variants were identified (Supplementary table 1-2, Supplementary dataset 2). These associations were found across 26 phenotypes, approximately one quarter of the 109 phenotypes tested, demonstrating pleiotropic characteristics of the HLA region in the Taiwanese population. After conditional analysis, these phenotypes showed at least one independent signal (Table 1). As shown in Fig. 4A and Table 1, four of the 26 phenotypes were binary traits. All of them were categorized as immune-related diseases, including drug allergy, arthritis, asthma, and AS. The most significant signal was the association between the HLA-B variation on the amino acid 138 and AS (p = 1.94×10-62). By contrast, the other 22 HLA variant-associated phenotypes were quantitative traits and fall into diverse categories. Most of these phenotypes were in the respiratory examination category (n=8), and the others were based on hematological examination (n=5), liver function (n=4), metabolic examination (n=4), and kidney function (n=1) (Fig. 4B, Table 1). After the forward-type conditional regression analysis, additional five independent association signals of RBC (n=3), WBC (n=1), and platelet (n=1) were identified. As a result, our PheWAS identified a total of 18,136 significant association signals, including 31 independent signals and another two variants that were in complete LD with phenotype-associated top variants (Table 1). Among the independent and top signals, 18 associations between HLA variants and phenotypes were published elsewhere (Table 1, Supplementary table 8). For example, HLA-DQA1*06:01 is the top signal associated with asthma in the Taiwanese cohort, which is consistent with a previous study of Chinese asthma families [35]. The rest of 15 associations were novel and not reported by the GWAS Catalog or previous studies (Table 1, Supplementary table 8).

Table 1 Top and independent signals with genome-wide significance in the HLA region
Full size table
Fig. 4
figure 4

Associations of HLA variations with binary traits (A) and quantitative traits (B). The solid lines indicate the threshold of genome-wide significance (p-value < 5×10-8); the dashed lines indicate the threshold of suggestive significance (p-value < 10-6); AS, Ankylosing Spondylitis; T-CHO, Total cholesterol; TG, Triglyceride; HDL, High density lipoprotein cholesterol; SGOT, Serum glutamic oxaloacetic transaminase; SGPT, Serum glutamic pyruvic transaminase; AFP, Alpha-Fetoprotein; RBC, Red blood cell count; WBC, White blood cell count; HB, Hemoglobin; HCT, Hematocrit; VC, Vital capacity; IRV, Inspiratory Reserve Volume; IC, Inspiratory capacity; VC/HT, Vital capacity/Height ratio; FVC, Forced vital capacity; FEF75, 75% of Forced expiratory flow; FEF75/HT, 75% of Forced expiratory flow/height ratio; EXV/FVC, Extrapolated Volume/ Forced vital capacity ratio

Full size image

To understand the pattern of associations between HLA variants and phenotypes, we further summarized HLA variants of the GWS association signals by their variation types, genomic locations, and gene classes (Supplementary figure 2). As shown in Supplementary figure 2A, the majority of the GWS signals were biallelic SNVs for most phenotypes. Notably, allelic and amino acid variations were dominant HLA variants in arthritis, T-CHO, and HDL. Furthermore, most of the GWS biallelic SNVs and indels on the HLA region identified in our PheWAS were located on intronic or intergenic regions (Supplementary figure 2B). Finally, these HLA variant-associated phenotypes were categorized into distinct patterns according to the HLA class of their corresponding variants: (1) only class I HLA variations (arthritis, AS, T-CHO, HB, HCT and the VC/HT ratio); (2) only class II HLA variations (drug allergy, asthma, and HDL); (3) both class I and class II HLA variations (TG, creatinine, albumin, SGOT, AFP, RBC, WBC, and platelet) (Supplementary figure 2C).

Heritability estimates and genetic correlation of HLA variations

To further understand the genetic effect of the HLA region on each phenotype, we evaluated the phenotypic variances explained by the HLA variations through HLA region-wide heritability estimation. Furthermore, we constructed the HLA region-wide genetic correlation network of phenotypes to discover common HLA genetic features amongst the phenotypes. As shown in Fig. 5A, hematological traits, which exhibited the greatest number of independent signals, had a large proportion of heritability explained by HLA variants. These hematological traits comprised platelet (0.82%), RBC count (0.23%), and WBC count (0.21%). Furthermore, the immune-related traits, including AS (0.46%), asthma (0.16%), and drug allergy (0.14%), also showed high HLA region-wide heritability. Interestingly, several traits without any GWS signals also displayed a large proportion of their variances explained by HLA variants. For example, articulus ache, SOS, and SI showed the HLA region-wide heritability of 0.28%,0.16%, and 0.10%, respectively. In the HLA region-wide genetic correlation network (Fig. 5B), phenotypes were linked together by their shared features in the associations with HLA variants. Generally, phenotypes within the same category were clustered together. For example, WHR, body waistline, body buttocks, body weight, and BMI were all in the cluster of anthropometric phenotypes with strong positive correlations between each other. Furthermore, a cluster of bone-related traits (SI, T-score, and Z-score) and a cluster of liver traits (SGOT and SGPT) were observed. However, there were also some cross-category correlations. For instance, positive correlations between bone density indices and respiratory examinations were detected. Moreover, we found negative correlations between AS and albumin (rg= -0.64) and between glaucoma and platelet (rg= -0.82).

Fig. 5
figure 5

Heritability estimates (A) and genetic correlation network (B) among 109 phenotypes of HLA polymorphisms. In panel A, only the traits with heritability ≥ 0.10% were shown; in panel B, the traits with p-value < 0.05 after FDR correction in the pairwise genetic correlation tests were demonstrated; AS, Ankylosing Spondylitis; RBC, Red blood cell count; WBC, White blood cell count; T-CHO, Total cholesterol; SOS, Speed of sound; HDL, High density lipoprotein cholesterol; ERV, Expiratory reserve volume; VC/HT, Vital capacity/Height ratio; VC, Vital capacity; LDL, Low density lipoprotein cholesterol; SGOT, Serum glutamic oxaloacetic transaminase; SI, Stiffness index; WHR, Waist–hip ratio; BMI, Body mass index; FVC, Forced vital capacity; HB, Hemoglobin; HCT, Hematocrit; EXV/FVC, Extrapolated Volume/ Forced vital capacity ratio; TG, Triglyceride; DBP, Diastolic pressure; SGPT, Serum glutamic pyruvic transaminase

Full size image

Pharmacogenomics of HLA variations

To understand the burden of HLA-related adverse drug reactions (ADR) in the Taiwanese population, we investigated the proportion of carriers of selected HLA alleles with known relationship to ADRs in our HLA- imputed Taiwanese cohort. As shown in Table 2, 30.86% of the imputed cohort were predicted to be HLA-C*01:02 carriers. This allele is known to be associated with higher risk of methazolamide-induced Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN). Furthermore, approximately 20% of the Taiwanese populations were HLA-B*58:01 carriers, which is associated with various drug-induced severe cutaneous adverse reactions (SCARs).

Table 2 Carriers and Frequencies of ADR-related HLA alleles in TWB (n=59,448)
Full size table

Discussion

Here, we constructed an accurate HLA imputation reference panel of eight HLA genes and imputed HLA alleles within the Taiwan Han Chinese population. The HLA allele frequencies are consistent with the previous reports from HLA direct-typing in other Taiwanese cohorts [25,26,27]. In addition, the haplotype distributions are similar to that in the southern Han Chinese populations [28]. The results indicate the robustness of this reference panel. Furthermore, within a diverse range of phenotypes, we discovered GWS associations with HLA variants, revealing numerous novel findings. Also, we observed both within and across categories correlations amongst the phenotypes accounting for the HLA genetic variations. Finally, the carriers of pharmacogenomic related HLA alleles, including HLA-C*01:02, HLA-B*58:01, and HLA-B*15:02, were characterized in the Taiwanese general population.

Huang et al. nicely constructed a Taiwanese-specific HLA imputation reference panel in 2020 [20]. The imputed HLA allele frequencies in the current study are similar to the findings by Huang et al. However, the imputation accuracies here are slightly better than the previous study (96.8% - 100% compared to 95.9% - 99.8%). Huang et al. utilized the TWBv1 genotype data for reference panel construction, while the reference panel in this study was constructed based on the TWBv2 array genotype. TWBv2 is the latest genome-wide SNP array implemented in the Taiwan biobank project. TWBv2 arrays include more variants across the HLA region [12, 21], which may be attributable to the higher accuracies of the imputation models. In addition, the haplotype frequencies in the TWB HLA imputed cohort are closer to the southern Han Chinese comparing to the other Han Chinese origins [28]. The observations were supported by previous reports that nearly 80% of the TWB subjects are clustered into the southern Han Chinese ancestry. This is also aligned with the immigrant history of Taiwan. The Taiwanese Minnan and Hakka ancestors moved from southeastern coastline of China in the past few centuries [17]. These results indicated the robustness and the reliability of the HLA imputation reference panel constructed in this study.

We explored the phenotypic implications of the HLA genes using the imputed HLA variations, 26 out of 109 examined traits were highlighted as significantly HLA-associated. In addition, a total of 31 independent signals and two variants in complete LD with the top signal were identified. Importantly, the HLA association patterns exhibited a high degree of consistency with previous reports. For example, the significant associations with a multitude of hematological traits, including the RBC, WBC, HB, and HCT, are enriched within the class I HLA genes, especially the HLA-B and HLA-C genes. Furthermore, the associations with allergic diseases and asthma are concentrated within class I HLA genes. These findings align with the observations in the Japanese and Korean populations [15, 16]. Conversely, we reported novel associations between HLA variations with several phenotypes. For example, in addition to the well-known association between HLA-B*27 and AS [69], our study reveals novel findings of AS association with HLA-B pos 138 and HLA-B-H138. This indicates a risk effect of histidine at the HLA-B position 138 in AS. The other novel associations include the rs9266290 and rs9266292 with drug allergy, and a variety of SNVs with respiratory quantitative traits. This variance could potentially be due to the inconsistency in phenotype collection or the natural differences of HLA genetic architectures amongst distinct ethnicities.

Within the study, we estimated the HLA region-wide heritability for each phenotype. Interestingly, we found that HLA variations only explained less than 1% of the phenotypic variations. This implies the multifactorial nature of phenotypes. The phenotypic variations are attributed to a broader spectrum of genomic variations and the environmental factors. Surprisingly, articulus ache, SOS, and SI were found to have high HLA region-wide heritability. Thus, future study should focus on the associations between HLA variations and these phenotypes in a larger sample size.

In the HLA region-wide genetic correlation analysis, not only within-category correlations but also cross-category correlations were observed. For example, bone density indices are positively correlated with respiratory examinations, which is with the previous findings [70, 71]. Moreover, a negative correlation between AS and albumin was detected. The result was supported by a prior study indicating fibrinogen to albumin ratio as a potential biomarker for AS [72]. Finally, we identified a negative correlation between glaucoma and platelet. The lower serum platelet level and alteration of platelet-related parameters in glaucoma patients imply the platelet activation in the disease progression of glaucoma [73, 74]. Collectively, these findings further demonstrate the pleiotropic feature of the HLA genetic region. Further studies are required to dissect the role of HLA molecules in these cross-category traits.

Given the significant clinical implications of HLA variations in pharmacogenomics, we comprehensively evaluated the burden of ADR-related HLA alleles in the Taiwanese population. In this study, more than twenty percent of the subjects are predicted to carry the HLA-B*58:01 allele, a variant associated with allopurinol-induced SCARs in a variety of Asian and European populations [37,38,39, 75,76,77,78,79]. Furthermore, approximately eight percent of individuals were identified as HLA-B*15:02 carriers in the Taiwanese population. HLA-B*15:02 allele is associated with a higher risk of carbamazepine-induced SJS/TEN. Importantly, this association has been confirmed in the Asian populations [40, 48,49,50, 52, 75, 80,81,82,83,84]. Notably, National Health Insurance (NHI) in Taiwan covers the HLA-B*15:02 and HLA-B*58:01 genetic tests before using carbamazepine and allopurinol, respectively [75]. Another finding is that 35.86% of the Taiwanese subjects are predicted to carry HLA-C*01:02, a variant with increased susceptibility to methazolamide-induced SJS/TEN [85, 86]. Both HLA-C*01:02 and HLA-B*59:01 were reported to be associated with methazolamide-induced SJS/TEN. According to previous study, effect size of HLA-C*01:02 on methazolamide-induced SJS/TEN is not as profound as HLA-B*59:01 [67]. However, none of the subjects in the training set was identified to carry HLA-B*59:01. Future studies with rare HLA alleles in the reference panel may improve the imputation accuracy and call rate to identify this allele. Furthermore, previous study indicated multiple factors contributing to pharmacogenomic implications [8]. It will be helpful to investigate the interaction between the cocontributing factors and HLA variants in methazolamide-induced SJS/TEN.

Despite the robust construction of a Taiwanese-specific HLA imputation reference panel and the comprehensively phenotypic exploration of the HLA region, there are still some limitations here. First, the sequencing-based HLA typing data utilized for training the reference panel only contains genotypes of eight classical HLA genes. Therefore, non-classical HLA genes and non-HLA genes in the HLA region cannot be imputed. Second, although the high imputation accuracies were established in our reference panel, the imputed HLA genotypes were not perfect match to the sequencing-based HLA typing. Thus, HLA imputation strategy is useful for the large cohort investigation especially for the phenotype landscape of HLA variations in a population study, while it is not yet an alternative to the clinical sequencing-based HLA typing. On the other hand, the application of this reference panel relies on the existence of alleles in the training set, limiting the imputation for rare HLA alleles. Future studies with an expansion of the sequencing-based HLA typing data will improve the accuracy and call rate of the reference panel. Third, the binary phenotypes analyzed here were collected with self-reported questionnaires, which may lead to a recall bias and data missing . Moreover, the quantitative phenotypes were collected cross-sectionally from the urine and blood tests at subjects' entry to the TWB. Without longitudinal follow up, the biochemical testing may not be able to reflect the long-term health conditions. Finally, ADR information is not available in the TWB, hence, our assessment of ADR risk allele burden is limited to the estimated prevalence of carriers based on the imputation alleles. Combination of genomic data in the TWB and electronic health records of the subjects will provide more information of the phenotypic landscape for HLA variations.

Conclusion

We successfully developed a Taiwanese-specific HLA imputation reference panel of eight classical HLA genes. This panel is useful to gain insight into the genetic and phenotypic landscape of the HLA variations within the Taiwan Han Chinese population. Importantly, we identified the estimated burden of ADR-related HLA alleles in the Taiwanese population that will be helpful in pharmacogenomic implementation.

Availability of data and materials

No datasets were generated or analysed during the current study.

Abbreviations

PheWAS:

Phenome-wide association study

ADR:

Adverse drug reactions

HLA:

Human leukocyte antigen

MHC:

Major histocompatibility complex

RA:

Rheumatoid arthritis

LD:

Linkage disequilibrium

SBT:

Sequencing-based typing

TWB:

Taiwan biobank

NGS:

Next generation sequencing

MAF:

Minor allele frequency

HWE:

Hardy-Weinberg equilibrium

IBD:

Identity by descent

SNVs:

Single nucleotide variations

AFND:

Allele frequency net database

PCR-SSO:

PCR sequence specific oligonucleotide probe

PCR-SSP:

PCR sequence specific primer

PCs:

Principal components

HE:

Haseman-Elston

SJS/TEN:

Stevens-Johnson syndrome/toxic epidermal necrolysis

SCARs:

Severe cutaneous adverse drug reactions

NHI:

National health insurance

AS:

Ankylosing spondylitis

QTL:

Quantitative trait loci

BMI:

Body mass index

WHR:

Waist–hip ratio

T-CHO:

Total cholesterol

TG:

Triglyceride

HDL:

High density lipoprotein cholesterol

LDL:

Low density lipoprotein cholesterol

SGOT:

Serum glutamic oxaloacetic transaminase

SGPT:

Serum glutamic pyruvic transaminase

AFP:

Alpha-fetoprotein

RBC:

Red blood cell count

WBC:

White blood cell count

HB:

Hemoglobin

HCT:

Hematocrit

SI:

Stiffness index

SOS:

Speed of sound

VC:

Vital capacity

IRV:

Inspiratory reserve volume

IC:

Inspiratory capacity

VC/HT:

Vital capacity/height ratio

FVC:

Forced vital capacity

FEF75:

75% of Forced expiratory flow

FEF75/HT:

75% of Forced expiratory flow/height ratio

EXV/FVC:

Extrapolated volume/forced vital capacity ratio

DRESS:

Drug reaction with eosinophilia and systemic symptoms

HSS:

Hypersensitivity syndrome

References

  1. Shiina T, Inoko H, Kulski JK. An update of the HLA genomic region, locus information and disease associations: 2004. Tissue Antigens. 2004;64(6):631–49.

    Article  CAS  PubMed  Google Scholar 

  2. Hudson LE, Allen RL. Leukocyte Ig-Like Receptors - A Model for MHC Class I Disease Associations. Front Immunol. 2016;7:281.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Barker DJ, et al. The IPD-IMGT/HLA Database. Nucleic Acids Res. 2023;51(D1):D1053-d1060.

    Article  CAS  PubMed  Google Scholar 

  4. Trowsdale J, Knight JC. Major histocompatibility complex genomics and human disease. Annu Rev Genomics Hum Genet. 2013;14:301–23.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Shiina T, et al. The HLA genomic loci map: expression, interaction, diversity and disease. J Hum Genet. 2009;54(1):15–39.

    Article  CAS  PubMed  Google Scholar 

  6. Fan WL, et al. HLA Association with Drug-Induced Adverse Reactions. J Immunol Res. 2017;2017:3186328.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Gibson A, et al. Updates on the immunopathology and genomics of severe cutaneous adverse drug reactions. J Allergy Clin Immunol. 2023;151(2):289-300.e4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Jaruthamsophon K, et al. HLA Allele-Restricted Immune-Mediated Adverse Drug Reactions: Framework for Genetic Prediction. Annu Rev Pharmacol Toxicol. 2022;62:509–29.

    Article  CAS  PubMed  Google Scholar 

  9. Jeiziner C, et al. HLA-associated adverse drug reactions - scoping review. Clin Transl Sci. 2021;14(5):1648–58.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Tam V, et al. Benefits and limitations of genome-wide association studies. Nat Rev Genet. 2019;20(8):467–84.

    Article  CAS  PubMed  Google Scholar 

  11. Bastarache L, Denny JC, Roden DM. Phenome-Wide Association Studies. Jama. 2022;327(1):75–6.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Feng YA, et al. Taiwan Biobank: A rich biomedical research database of the Taiwanese population. Cell Genom. 2022;2(11):100197.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Ritari J, et al. HLA-disease association and pleiotropy landscape in over 235,000 Finns. Hum Immunol. 2022;83(5):391–8.

    Article  CAS  PubMed  Google Scholar 

  14. Karnes JH, Bastarache L, Shaffer CM, Gaudieri S, Xu Y, Glazer AM, Mosley JD, Zhao S, Raychaudhuri S, Mallal S, Ye Z, Mayer JG, Brilliant MH, Hebbring SJ, Roden DM, Phillips EJ, Denny JC. Phenome-wide scanning identifies multiple diseases and disease severity phenotypes associated with HLA variants. Sci Transl Med. 2017;9(389):eaai8708. https://doi.org/10.1126/scitranslmed.aai8708.

  15. Hirata J, et al. Genetic and phenotypic landscape of the major histocompatibilty complex region in the Japanese population. Nat Genet. 2019;51(3):470–80.

    Article  CAS  PubMed  Google Scholar 

  16. Kim C, et al. Phenome-wide association study of the major histocompatibility complex region in the Korean population identifies novel association signals. Hum Mol Genet. 2022;31(15):2655–67.

    Article  CAS  PubMed  Google Scholar 

  17. Chen CH, et al. Population structure of Han Chinese in the modern Taiwanese population based on 10,000 participants in the Taiwan Biobank project. Hum Mol Genet. 2016;25(24):5321–31.

    CAS  PubMed  PubMed Central  Google Scholar 

  18. Barone JC, et al. HLA-genotyping of clinical specimens using Ion Torrent-based NGS. Hum Immunol. 2015;76(12):903–9.

    Article  CAS  PubMed  Google Scholar 

  19. Naito T, Okada Y. HLA imputation and its application to genetic and molecular fine-mapping of the MHC region in autoimmune diseases. Semin Immunopathol. 2022;44(1):15–28.

    Article  CAS  PubMed  Google Scholar 

  20. Huang YH, et al. A high-resolution HLA imputation system for the Taiwanese population: a study of the Taiwan Biobank. Pharmacogenomics J. 2020;20(5):695–704.

    Article  CAS  PubMed  Google Scholar 

  21. Wei CY, et al. Genetic profiles of 103,106 individuals in the Taiwan Biobank provide insights into the health and history of Han Chinese. NPJ Genom Med. 2021;6(1):10.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Zheng X, et al. HIBAG–HLA genotype imputation with attribute bagging. Pharmacogenomics J. 2014;14(2):192–200.

    Article  CAS  PubMed  Google Scholar 

  23. Zheng X. Imputation-Based HLA Typing with SNPs in GWAS Studies. Methods Mol Biol. 2018;1802:163–76.

    Article  CAS  PubMed  Google Scholar 

  24. Gonzalez-Galarza FF, et al. Allele frequency net database (AFND) 2020 update: gold-standard data classification, open access genotype data and new query tools. Nucleic Acids Res. 2020;48(D1):D783-d788.

    CAS  PubMed  Google Scholar 

  25. Lai MJ, et al. Distributions of human leukocyte antigen-A, -B, and -DRB1 alleles and haplotypes based on 46,915 Taiwanese donors. Hum Immunol. 2010;71(8):777–82.

    Article  CAS  PubMed  Google Scholar 

  26. Wen SH, Lai MJ, Yang KL. Human leukocyte antigen-A, -B, and -DRB1 haplotypes of cord blood units in the Tzu Chi Taiwan Cord Blood Bank. Hum Immunol. 2008;69(7):430–6.

    Article  CAS  PubMed  Google Scholar 

  27. Chen PL, et al. Comprehensive genotyping in two homogeneous Graves’ disease samples reveals major and novel HLA association alleles. PLoS One. 2011;6(1):e16635.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Zhou F, et al. Deep sequencing of the MHC region in the Chinese population contributes to studies of complex disease. Nat Genet. 2016;48(7):740–6.

    Article  CAS  PubMed  Google Scholar 

  29. Mbatchou J, et al. Computationally efficient whole-genome regression for quantitative and binary traits. Nat Genet. 2021;53(7):1097–103.

    Article  CAS  PubMed  Google Scholar 

  30. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010;38(16): e164.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Yang J, et al. Concepts, estimation and interpretation of SNP-based heritability. Nat Genet. 2017;49(9):1304–10.

    Article  CAS  PubMed  Google Scholar 

  32. Csardi G, Nepusz T. The igraph software package for complex network research. Inter J Complex Syst. 2006;1695. https://igraph.org.

  33. Negrini S, Becquemont L. HLA-associated drug hypersensitivity and the prediction of adverse drug reactions. Pharmacogenomics. 2017;18(15):1441–57.

    Article  CAS  PubMed  Google Scholar 

  34. Pavlos R, Mallal S, Phillips E. HLA and pharmacogenetics of drug hypersensitivity. Pharmacogenomics. 2012;13(11):1285–306.

    Article  CAS  PubMed  Google Scholar 

  35. Guo X, Ni P, Li L. Association between asthma and the polymorphism of HLA-DQ genes. Zhonghua Jie He He Hu Xi Za Zhi. 2001;24(3):139–41.

    CAS  PubMed  Google Scholar 

  36. Kim SH, et al. HLA-B*5901 is strongly associated with methazolamide-induced Stevens-Johnson syndrome/toxic epidermal necrolysis. Pharmacogenomics. 2010;11(6):879–84.

    Article  CAS  PubMed  Google Scholar 

  37. Kang HR, et al. Positive and negative associations of HLA class I alleles with allopurinol-induced SCARs in Koreans. Pharmacogenet Genomics. 2011;21(5):303–7.

    Article  CAS  PubMed  Google Scholar 

  38. Ha Pham TT, et al. Allopurinol-induced severe cutaneous adverse reactions in Vietnamese: the role of HLA alleles and other risk factors. Pharmacogenomics. 2022;23(5):303–13.

    Article  CAS  PubMed  Google Scholar 

  39. Yu KH, Yu CY, Fang YF. Diagnostic utility of HLA-B*5801 screening in severe allopurinol hypersensitivity syndrome: an updated systematic review and meta-analysis. Int J Rheum Dis. 2017;20(9):1057–71.

    Article  CAS  PubMed  Google Scholar 

  40. Sukasem C, et al. Association between HLA-B Alleles and Carbamazepine-Induced Maculopapular Exanthema and Severe Cutaneous Reactions in Thai Patients. J Immunol Res. 2018;2018:2780272.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Wang Q, et al. Association between the HLA-B alleles and carbamazepine-induced SJS/TEN: A meta-analysis. Epilepsy Res. 2017;135:19–28.

    Article  CAS  PubMed  Google Scholar 

  42. Kazeem GR, et al. High-resolution HLA genotyping and severe cutaneous adverse reactions in lamotrigine-treated patients. Pharmacogenet Genomics. 2009;19(9):661–5.

    Article  CAS  PubMed  Google Scholar 

  43. Zhang FR, et al. HLA-B*13:01 and the dapsone hypersensitivity syndrome. N Engl J Med. 2013;369(17):1620–8.

    Article  CAS  PubMed  Google Scholar 

  44. Tempark T, et al. Dapsone-induced severe cutaneous adverse drug reactions are strongly linked with HLA-B*13: 01 allele in the Thai population. Pharmacogenet Genomics. 2017;27(12):429–37.

    Article  CAS  PubMed  Google Scholar 

  45. Chen WT, et al. The Function of HLA-B*13:01 Involved in the Pathomechanism of Dapsone-Induced Severe Cutaneous Adverse Reactions. J Invest Dermatol. 2018;138(7):1546–54.

    Article  CAS  PubMed  Google Scholar 

  46. Su SC, et al. HLA Alleles and CYP2C9*3 as Predictors of Phenytoin Hypersensitivity in East Asians. Clin Pharmacol Ther. 2019;105(2):476–85.

    Article  CAS  PubMed  Google Scholar 

  47. Lee HY, et al. The HLA allele marker for differentiating ASA hypersensitivity phenotypes. Allergy. 2009;64(9):1385–7.

    Article  CAS  PubMed  Google Scholar 

  48. Tangamornsuksan W, et al. Relationship between the HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis: a systematic review and meta-analysis. JAMA Dermatol. 2013;149(9):1025–32.

    Article  CAS  PubMed  Google Scholar 

  49. Genin E, et al. HLA-A*31:01 and different types of carbamazepine-induced severe cutaneous adverse reactions: an international study and meta-analysis. Pharmacogenomics J. 2014;14(3):281–8.

    Article  CAS  PubMed  Google Scholar 

  50. Cheung YK, et al. HLA-B alleles associated with severe cutaneous reactions to antiepileptic drugs in Han Chinese. Epilepsia. 2013;54(7):1307–14.

    Article  CAS  PubMed  Google Scholar 

  51. Chang CC, et al. Association of HLA-B*15:13 and HLA-B*15:02 with phenytoin-induced severe cutaneous adverse reactions in a Malay population. Pharmacogenomics J. 2017;17(2):170–3.

    Article  CAS  PubMed  Google Scholar 

  52. Locharernkul C, et al. Carbamazepine and phenytoin induced Stevens-Johnson syndrome is associated with HLA-B*1502 allele in Thai population. Epilepsia. 2008;49(12):2087–91.

    Article  PubMed  Google Scholar 

  53. Koomdee N, et al. Association of HLA-A and HLA-B Alleles with Lamotrigine-Induced Cutaneous Adverse Drug Reactions in the Thai Population. Front Pharmacol. 2017;8:879.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Sabourirad S, et al. Investigating the association of Lamotrigine and Phenytoin-induced Stevens-Johnson syndrome/Toxic Epidermal Necrolysis with HLA-B*1502 in Iranian population. Exp Dermatol. 2021;30(2):284–7.

    Article  CAS  PubMed  Google Scholar 

  55. Yuan J, et al. Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African, Asian, and European descent. Aids. 2011;25(10):1271–80.

    Article  CAS  PubMed  Google Scholar 

  56. Keane NM, et al. HLA Class I restricted CD8+ and Class II restricted CD4+ T cells are implicated in the pathogenesis of nevirapine hypersensitivity. Aids. 2014;28(13):1891–901.

    Article  CAS  PubMed  Google Scholar 

  57. McCormack M, et al. HLA-A*3101 and carbamazepine-induced hypersensitivity reactions in Europeans. N Engl J Med. 2011;364(12):1134–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Kim SH, et al. Carbamazepine-induced severe cutaneous adverse reactions and HLA genotypes in Koreans. Epilepsy Res. 2011;97(1–2):190–7.

    Article  CAS  PubMed  Google Scholar 

  59. Nicoletti P, et al. Shared Genetic Risk Factors Across Carbamazepine-Induced Hypersensitivity Reactions. Clin Pharmacol Ther. 2019;106(5):1028–36.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Kim BK, et al. HLA-A*31:01 and lamotrigine-induced severe cutaneous adverse drug reactions in a Korean population. Ann Allergy Asthma Immunol. 2017;118(5):629–30.

    Article  CAS  PubMed  Google Scholar 

  61. Shi YW, et al. Association between HLA and Stevens-Johnson syndrome induced by carbamazepine in Southern Han Chinese: genetic markers besides B*1502? Basic Clin Pharmacol Toxicol. 2012;111(1):58–64.

    Article  CAS  PubMed  Google Scholar 

  62. Wong CSM, et al. HLA-B*15:11 status and carbamazepine-induced severe cutaneous adverse drug reactions in HLA-B*15:02 negative Chinese. Int J Dermatol. 2022;61(2):184–90.

    Article  CAS  PubMed  Google Scholar 

  63. Martin AM, et al. Predisposition to nevirapine hypersensitivity associated with HLA-DRB1*0101 and abrogated by low CD4 T-cell counts. Aids. 2005;19(1):97–9.

    Article  CAS  PubMed  Google Scholar 

  64. Sousa-Pinto B, et al. Pharmacogenetics of abacavir hypersensitivity: A systematic review and meta-analysis of the association with HLA-B*57:01. J Allergy Clin Immunol. 2015;136(4):1092-4.e3.

    Article  CAS  PubMed  Google Scholar 

  65. Daly AK, et al. HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet. 2009;41(7):816–9.

    Article  CAS  PubMed  Google Scholar 

  66. Tangamornsuksan W, Lohitnavy M. Association between HLA-B*5901 and methazolamide-induced Stevens-Johnson syndrome/toxic epidermal necrolysis: a systematic review and meta-analysis. Pharmacogenomics J. 2019;19(3):286–94.

    Article  CAS  PubMed  Google Scholar 

  67. Jiang M, et al. Unique motif shared by HLA-B*59:01 and HLA-B*55:02 is associated with methazolamide-induced Stevens-Johnson syndrome and toxic epidermal necrolysis in Han Chinese. J Eur Acad Dermatol Venereol. 2022;36(6):873–80.

    Article  CAS  PubMed  Google Scholar 

  68. Ikeda H, et al. HLA class I markers in Japanese patients with carbamazepine-induced cutaneous adverse reactions. Epilepsia. 2010;51(2):297–300.

    Article  PubMed  Google Scholar 

  69. Bowness P. HLA-B27. Annu Rev Immunol. 2015;33:29–48.

    Article  CAS  PubMed  Google Scholar 

  70. Sarkar M, et al. Osteoporosis in chronic obstructive pulmonary disease. Clin Med Insights Circ Respir Pulm Med. 2015;9:5–21.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Zeng X, et al. Association of bone mineral density with lung function in a Chinese general population: the Xinxiang rural cohort study. BMC Pulm Med. 2019;19(1):239.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Liu M, et al. The role of fibrinogen to albumin ratio in ankylosing spondylitis: Correlation with disease activity. Clin Chim Acta. 2020;505:136–40.

    Article  CAS  PubMed  Google Scholar 

  73. Ma Y, et al. Association between Platelet Parameters and Glaucoma Severity in Primary Open-Angle Glaucoma. J Ophthalmol. 2019;2019:3425023.

    Article  PubMed  PubMed Central  Google Scholar 

  74. Ma Y, et al. Platelet Parameters and Their Relationships With the Thickness of the Retinal Nerve Fiber Layer and Ganglion Cell Complex in Primary Open-Angle Glaucoma. Front Neurol. 2022;13:867465.

    Article  PubMed  PubMed Central  Google Scholar 

  75. Wang CW, et al. An Updated Review of Genetic Associations With Severe Adverse Drug Reactions: Translation and Implementation of Pharmacogenomic Testing in Clinical Practice. Front Pharmacol. 2022;13:886377.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  76. Hung SI, et al. HLA-B*5801 allele as a genetic marker for severe cutaneous adverse reactions caused by allopurinol. Proc Natl Acad Sci U S A. 2005;102(11):4134–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  77. Kaniwa N, et al. HLA-B locus in Japanese patients with anti-epileptics and allopurinol-related Stevens-Johnson syndrome and toxic epidermal necrolysis. Pharmacogenomics. 2008;9(11):1617–22.

    Article  CAS  PubMed  Google Scholar 

  78. Lonjou C, et al. A European study of HLA-B in Stevens-Johnson syndrome and toxic epidermal necrolysis related to five high-risk drugs. Pharmacogenet Genomics. 2008;18(2):99–107.

    Article  CAS  PubMed  Google Scholar 

  79. Tassaneeyakul W, et al. Strong association between HLA-B*5801 and allopurinol-induced Stevens-Johnson syndrome and toxic epidermal necrolysis in a Thai population. Pharmacogenet Genomics. 2009;19(9):704–9.

    Article  CAS  PubMed  Google Scholar 

  80. Chung WH, et al. Medical genetics: a marker for Stevens-Johnson syndrome. Nature. 2004;428(6982):486.

    Article  CAS  PubMed  Google Scholar 

  81. Hung SI, et al. Genetic susceptibility to carbamazepine-induced cutaneous adverse drug reactions. Pharmacogenet Genomics. 2006;16(4):297–306.

    Article  CAS  PubMed  Google Scholar 

  82. Mehta TY, et al. Association of HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome among Indians. Indian J Dermatol Venereol Leprol. 2009;75(6):579–82.

    Article  PubMed  Google Scholar 

  83. Tassaneeyakul W, et al. Association between HLA-B*1502 and carbamazepine-induced severe cutaneous adverse drug reactions in a Thai population. Epilepsia. 2010;51(5):926–30.

    Article  PubMed  Google Scholar 

  84. Then SM, et al. Frequency of the HLA-B*1502 allele contributing to carbamazepine-induced hypersensitivity reactions in a cohort of Malaysian epilepsy patients. Asian Pac J Allergy Immunol. 2011;29(3):290–3.

    CAS  PubMed  Google Scholar 

  85. Qiumei G, et al. Association of HLA-C*01:02 with methazolamide-induced toxic epidermal necrolysis. BMJ Case Rep. 2022;15(12):e250048.

    Article  PubMed  Google Scholar 

  86. Yang F, et al. HLA-B*59:01: a marker for Stevens-Johnson syndrome/toxic epidermal necrolysis caused by methazolamide in Han Chinese. Pharmacogenomics J. 2016;16(1):83–7.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We thank Dr. Che-Mai Chang (Taipei Medical University, Taiwan) for the helpful suggestions during the manuscript preparation.

Funding

This work was supported by grants from the Ministry of Education, Taiwan (DP2-TMU-112-T-06), the National Science and Technology Council, Taiwan (NSTC112-2320-B-038-026-MY3) and Taipei Medical University (111-5812-001-400; Yusuke Nakamura Chair Professorship).

Author information

Author notes

  1. Wan-Hsuan Chou and Lu-Chun Chen contributed equally to this work.

Authors and Affiliations

  1. Department of Clinical Pharmacy, School of Pharmacy, Taipei Medical University, Taipei, Taiwan

    Wan-Hsuan Chou, Lu-Chun Chen, Henry Sung-Ching Wong, Ching-Hsuan Chao & Wei-Chiao Chang

  2. Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan

    Hou-Wei Chu

  3. Master Program in Clinical Genomics and Proteomics, School of Pharmacy, Taipei Medical University, Taipei, Taiwan

    Wei-Chiao Chang

  4. Integrative Research Center for Critical Care, Wan Fang Hospital, Taipei Medical University, Taipei, Taiwan

    Wei-Chiao Chang

  5. Department of Pharmacy, Wan Fang Hospital, Taipei Medical University, Taipei, Taiwan

    Wei-Chiao Chang

  6. Department of Pharmacology, National Defense Medical Center, Taipei, Taiwan

    Wei-Chiao Chang

Authors
  1. Wan-Hsuan Chou
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Lu-Chun Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Henry Sung-Ching Wong
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ching-Hsuan Chao
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Hou-Wei Chu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Wei-Chiao Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

WHC, LCC, HSCW, and WCC conceptualized, conceived, and designed the study. WHC, LCC, HSCW, and WCC analyzed data. WHC, LCC, and WCC interpreted results. WHC wrote the original manuscript. WHC, LCC, and WCC revised the manuscript. HSCW and CHC provided comments during data analysis. HWC collected TWB data. WCC provided the funding and edited the manuscript. All authors contributed to the article and approved the submitted version.

Corresponding author

Correspondence to Wei-Chiao Chang.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of Taipei Medical University (TMU-JIRB N201906005) and the Taiwan Biobank, Academia Sinica (TWBR10906-03).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

Supplementary figure 1. Frequencies of the two-, four-, and six-digit HLA alleles before and after HLA imputation. Supplementary figure 2. Distribution of variant types, variant annotation types of biallelic SNVs and indels, and HLA classes of the HLA polymorphisms with significant association with phenotypes. Supplementary table 1. Classification of the 55 binary traits included in the PheWAS. Supplementary table 2. Classification of the 54 quantitative traits included in the PheWAS. Supplementary table 3. Overview of common haplotype frequencies in Taiwanese. Supplementary table 4. Comparison of haplotype frequencies between Taiwanese and Han Chinese. Supplementary table 5. Novelty of the identified independent association signals with genome-wide significance in the entire HLA region.

Additional file 2:

Supplementary dataset 1. HLA allele frequencies in different Taiwanese cohorts.

Additional file 3.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chou, WH., Chen, LC., Wong, H.SC. et al. Phenomic landscape and pharmacogenomic implications for HLA region in a Taiwan Han Chinese population. Biomark Res 12, 46 (2024). https://doi.org/10.1186/s40364-024-00591-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s40364-024-00591-z

Keywords

Biomarker Research

ISSN: 2050-7771

Contact us