Genetic modifiers of body mass index in individuals with cystic fibrosis

Abstract

To identify modifier loci underlying variation in body mass index (BMI) in persons with cystic fibrosis (pwCF), we performed a genome-wide association study (GWAS). Utilizing longitudinal height and weight data, along with demographic information and covariates from 4,393 pwCF, we calculated AvgBMIz representing the average of per-quarter BMI Z scores. The GWAS incorporated 9.8M single nucleotide polymorphisms (SNPs) with a minor allele frequency (MAF) > 0.005 extracted from whole-genome sequencing (WGS) of each study subject. We observed genome-wide significant association with a variant in FTO (FaT mass and Obesity-associated gene; rs28567725; p value = 1.21e-08; MAF = 0.41, β = 0.106; n = 4,393 individuals) and a variant within ADAMTS5 (A Disintegrin And Metalloproteinase with ThromboSpondin motifs 5; rs162500; p value = 2.11e-10; MAF = 0.005, β = -0.768; n = 4,085 pancreatic-insufficient individuals). Notably, BMI-associated variants in ADAMTS5 occur on a haplotype that is much more common in African (AFR, MAF = 0.183) than European (EUR, MAF = 0.006) populations (1000 Genomes project). A polygenic risk score (PRS) calculated using 924 SNPs (excluding 17 in FTO) showed significant association with AvgBMIz (p value = 2.2e-16; r² = 0.03). Association between variants in FTO and the PRS correlation reveals similarities in the genetic architecture of BMI in CF and the general population. Inclusion of Black individuals in whom the single-gene disorder CF is much less common but genomic diversity is greater facilitated detection of association with variants that are in LD with functional SNPs in ADAMTS5. Our results illustrate the importance of population diversity, particularly when attempting to identify variants that manifest only under certain physiologic conditions.

Keywords: BMI; GWAS; Mendelian disease; complex trait; cystic fibrosis; genetic modifier; nutritional status; population difference; rare variant; whole-genome sequencing.

Graphical abstract

Figure 1

Distribution of AvgBMIz across three study coordinating centers All three sites have similar variance while distributions correspond with age of subjects; UW had youngest with highest mean BMI and UNC had oldest with lowest mean BMI. N = 4,393.

Figure 2

Manhattan plots for AvgBMIz in all pwCF and PI only from primary analyses SNPs at a locus on chromosome 16 reached genome-wide significance in both all pwCF (n = 4,393) (A) and PI only (n = 4,085) (B) analyses. SNPs at one locus on chromosome 21 reached genome-wide significance in PI-only analysis. The single SNP on chromosome 7 that appears to exceed genome-wide significance in PI-only analysis was determined to be inaccurately typed. SNPs in other loci that exceeded suggestive significance are described in Table S5. Dashed lines represent the genome-wide significance level after Bonferroni correction.

Figure 3

All and population-specific locusZoom plot for SNP rs162500 in ADAMTS5 in pwCF who are pancreatic insufficient (A) locusZoom plot for 4,085 PI-only individuals using LD information from 1000 Genomes global PCs. (B and C) Represent population-specific GWAS for AvgBMIz score using population-specific PCs. (B) PCA group A individuals (n = 3,791) in CFGP using LD information from 1000 Genomes European population. (C) PCA group B individuals (n = 77) in CFGP using LD information from 1000 Genomes African population. Each point represents a SNP with color indicating correlation (r²) to rs162500. The remaining 8 SNPs with p values < 10e−05 in the PI-CF analysis are in high LD with rs162500 (−log10 p values on left y axis) when using LD information from 1000 Genomes European. The blue peaks depict areas with high recombination rates (i.e., hotspots) based on 1000 Genomes Project data (recombination rate on right y axis). The recombination hot spots observed between ADAMTS1 and ADAMTS5 and within ADAMTS1 suggest that rs162500 and the 8 SNPs in high LD with rs162500 (purple diamond) form a haplotype encompassing ADAMTS5.

Figure 4

Haploview plot in build GRCh38 for regions on chromosome 21 which compasses ADAMTS5 Haploview plots defining linkage disequilibrium (LD) among SNPs within the AvgBMIz associated region on chromosome 21 for (A) ADAMTS1-ADAMTS5 (26,635–26,970 kb) and for (B) ADAMTS5 (26,917–26,967 kb). Eight SNPs with a p value < 1e−05 for AvgBMIz in PI-CF individuals (n = 4,085) are depicted in green. Regions where SNPs display high LD and form haplotype blocks are illustrated with black lines. The relative physical position of each SNP is given in the upper diagram in green (key SNP rs162500 is bolded). Pairwise linkage disequilibrium (r²) between all SNPs is given below each SNP combination. SNPs with MAF >0.3, call rate >90% in 200 random samples plus 65 individuals with the alternative allele at rs162500 (LD measured by r²).

Figure 5

Locus zoom plots for the top hit SNP rs28567725 at FTO locus LD information from 1000 Genomes Project (GRCh38) all (A) and European (EUR) (B) samples as reference. Each point represents a SNP, and the color indicates its correlation with rs28567725 (purple diamond) (−log10 p values on left y axis). The blue peaks depict areas with high recombination rates (i.e., hotspots) based on 1000 Genomes Project data (recombination rate on right y axis). Locus zoom plot for top hit SNP (rs1421085) at FTO loci (n = 4,085) from GIANT (GRCh37) all ancestry study with LD from 1000 Genomes all ancestry (C), and GIANT European ancestry study with LD from 1000 Genomes EUR ancestry (D). Note: The shift in coordinates is due to the difference between GRCh37 and GRCh38.