Causal links of α-thalassemia indices and cardiometabolic traits and diabetes: MR study

Abstract

Our study aimed to investigate if genetic variants around 16p13.3's HBA1 locus, associated with erythrocyte indices and HbA1c levels, predict α-thalassemia-related erythrocyte indices, cardiometabolic traits, and diabetes risk in Taiwanese individuals. We analyzed Taiwan Biobank data, including whole-genome sequencing from 1,493 participants and genotyping arrays from 129,542 individuals. First, we performed regional association analysis using whole-genome sequencing data to identify genetic variants significantly associated with erythrocyte indices, confirming their linkage disequilibrium with the α⁰ thalassemia --^SEA deletion mutation, a common cause of α-thalassemia in Southeast Asian populations. Deletion mutation sequencing further validated these variants' association with α-thalassemia. Subsequently, we analyzed genotyping array data, revealing associations between specific genetic variants and cardiometabolic traits, including lipid profiles, HbA1c levels, bilirubin levels, and diabetes risk. Using Mendelian randomization, we established causal relationships between α-thalassemia-related erythrocyte indices and cardiometabolic traits, elucidating their role in diabetes susceptibility. Our findings highlight genetic variants around the α-globin genes as surrogate markers for common α-thalassemia mutations in Taiwan, emphasizing the causal links between α-thalassemia-related erythrocyte indices, cardiometabolic traits, and heightened diabetes risk.

Figure 1.. Flowchart of inclusion and exclusion criteria used to screen Taiwan Biobank project participants.

Abbreviations: WGS, whole-genome sequence; GWAS, genome-wide association study; QC, quality control.

Figure 2.. Regional association study of RBC count, hemoglobin (Hb) level, mean corpuscular volume, and mean corpuscular hemoglobin around the chromosome16p13.3 regions in 1,493 Taiwan Biobank participants with whole-genome sequence.

P-value was obtained from a linear regression of each RBC parameter with a genetic variant, adjusted for age, sex, current smoking status, and body mass index.

Figure 3.. Genomic structure and linkage disequilibrium map of chromosome 16p13.3 variants and their association with α⁰ thalassemia --^SEA deletion mutation, various phenotypes, and diabetes mellitus.

Figure 4.. Association between chromosome 16p13.3 variants and α-thalassemia indices.

(A, B, C, D, E, F, G, H, I) Association between genetic variants ((A, B, C), NPRL3 rs191086839; (D, E, F), LUC7L rs372755452; (G, H, I), PGAP6 rs375498857) from chromosome 16p13.3 and microcytic hypochromic erythrocyte traits (MC-HC-E), anemia, and microcytic anemia (MC-HC-A) in 1,493 Taiwan Biobank participants with whole-genome sequencing data. P-values were examined using a chi-squared test.

Figure 5.. BreakDancer and PCR analysis with direct DNA sequencing for α⁰ thalassemia deletion mutations.

(A, B, C, D, E, F, G, H, I, J, K, L, M, N, O) Diagrams of α-thalassemia mutations for gene alignment (A, B), genotyping by BreakDancer v1.3.6 (C, D, E, F), PCR (G, H), direct DNA sequencing (I, J, K), and association results of chromosome 16p13.3 variants (M, N, O) and genotyping performed using BreakDancer v1.3.6 (L) with the α⁰ thalassemia deletion --^SEA detected through PCR. (A) Genes are represented as black boxes and pseudogenes as white boxes. The α-thalassemia mutations are represented as grey lines. (B) Diagrams illustrate the structure of the α-globin gene cluster on chromosome 16 with α-thalassemia mutations. (G, H) Agarose gels show representative results of PCR assays. (G) Sizes of amplified fragments are expressed in base pairs (bp). Lane N, negative control; lane M, 100-bp DNA ladder H3 RTU (GeneDireX, Inc.); (G) Lanes 1, 2, 5, and 7 indicate α-thalassemia --^SEA heterozygotes because of the presence of the deletion-specific 188-bp band and a 314-bp band obtained from the control DNA sequence. Lanes 3, 4, 6, and 8 indicate participants without deletion mutations that provided only the 314-bp band. (H) Lanes 1 indicates α-thalassemia --^THAI heterozygotes because of the presence of the deletion specific 411-bp band and a 314-bp band. Lanes 2–5 indicate α-thalassemia --^FIL heterozygotes because of the presence of the deletion specific 550-bp band and a 314-bp band. Lanes 6 indicates participants without deletion mutation.

Figure S1.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for total cholesterol in Taiwan Biobank participants with genome-wide association study array data.

Figure S2.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for high-density lipoprotein–cholesterol in Taiwan Biobank participants with genome-wide association study array data.

Figure S3.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for low-density lipoprotein–cholesterol in Taiwan Biobank participants with genome-wide association study array data.

Figure S4.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for HbA1c in Taiwan Biobank participants with genome-wide association study array data.

Figure S5.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for total bilirubin in Taiwan Biobank participants with genome-wide association study array data.

Figure S6.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for hematocrit in Taiwan Biobank participants with genome-wide association study array data.

Figure S7.. Regional-plot association studies for genetic variants at positions 0.06 and 0.68 Mb on chromosome 16p13.3 for hemoglobin in Taiwan Biobank participants with genome-wide association study array data.