Phenome-wide analysis ofAPOL1risk variants reveals associations between one combination of haplotypes and multiple disease phenotypes in addition to chronic kidney disease

Abstract

Background Infectious diseases are a major driving force of natural selection. One human gene associated with strong evolutionary selection isAPOL1. TwoAPOL1variants, G1 and G2, emerged in sub-Saharan Africa in the last 10,000 years, possibly due to protection from the fatal African sleeping sickness, analogous toPlasmodium-driven selection of the sickle-cell trait. As homozygosity for the HbS allele causes sickle cell anaemia, homozygosity for theAPOL1G1 and G2 variants has also been associated with chronic kidney disease (CKD) and other kidney-related conditions. What is not known is the extend of non-kidney-related disorders and if there are clusters of diseases associated with individualAPOL1genotypes. Methods Using principal component analysis, we identified a cohort of 10,179 UK Biobank participants with recent African ancestry. We conducted a phenome-wide association test between all combinations ofAPOL1G1 and G2 genotypes and conditions identified with International Classification of Disease phenotypes using Firth's bias-reduced logistic regression and a false discovery rate to correct for multiple testing. We further examined associations with chronic kidney disease indicators: estimated glomerular filtration rate (eGFR) and urinary albumin:creatinine (uACR). Results The phenome-wide screen revealed 74 (mostly deleterious) potential associations with hospitalisation for a range of conditions. G1/G2 compound heterozygotes were specifically associated with hospitalisation in 64 (86.5%) of these conditions, with an over-representation of infectious diseases (including COVID-19) and endocrine, nutritional, and metabolic diseases. The analysis also revealed complexities in the relationship betweenAPOL1and CKD that are not evident when the risk variants are grouped together: high uACR was associated specifically with G1 homozygosity; low eGFR with G2 homozygosity and G1/G2 compound heterozygosity; progression to end stage kidney disease was associated with G1/G2 compound heterozygosity. Conclusions Among 9,594 participants, stratifying individualAPOL1risk variant genotypes had a differential effect on associations with both kidney and non-kidney phenotypes. The compound heterozygous G1/G2 genotype was distinguished as uniquely deleterious in its association with a range of ICD-10 phenotypes. The epistatic nature of the G1/G2 interaction means that such associations may go undetected in a standard genome-wide association study. These observations have the potential to significantly impact the way that health risks are understood, particularly in populations whereAPOL1G1 and G2 are common such as in sub-Saharan Africa and its diaspora.

APOL1 variants G1, G2 and N264K affect APOL1 plasma protein concentration: a UK Biobank study

“…Previously, phenome-wide analysis highlighted differences between the six observed combined APOL1 G1 and G2 genotypes (8). These differences extend to protein concentration (Table 6, Figure 3).…”

“…Olink APOL1 protein concentration data and complete APOL1 genotypes were available for 1,050 UK Biobank participants within our previously defined cohort of participants with recent sub-Saharan African ancestry (8). Olink data is expressed in the Normalized Protein eXpression (NPX) scale, enabling relative quantification of the same protein across multiple samples.…”

“…Recent laboratory and association studies have transformed our understanding of the human APOL1 gene: widening the scope of potentially detrimental effects in current populations as an evolutionary cost of surviving historic sleeping sickness epidemics (8,13,(20)(21)(22) as well as identifying an additional modifying variant that mitigates APOL1 impact both on trypanosomiasis and CKD (9)(10)(11)(12), and demonstrating significant phenotypic differences between the G1 and G2 variants (8). In studies examining the role of APOL1 in trypanosomiasis, these differences are well-established: G1 is protective against severe disease in T.b.…”

“…G2 (rs71785313) genotypes were imputed by the UK Biobank as previously described (17). By analysing genetic principal components, we previously identified a cohort of UK Biobank participants who had recent sub-Saharan African ancestry (8). Briefly, using principal components (PC) calculated by the UK Biobank genetic data of all UK Biobank participants, the large majority of participants with PC1 > 100 and PC2 > 0 described their ethnic background as 'Black or Black British' or 'Mixed'.…”

“…Using data from the UK Biobank, we previously highlighted differences in association between the two-variant APOL1 genotypes with kidney phenotypes, and demonstrated that one combined APOL1 genotype, G1/G2 (carried by millions of people worldwide), is deleteriously associated with 26 different conditions spanning human health (8). Our analysis exposed complexities in the relationship between APOL1 risk alleles and disease that are not evident when two-variant APOL1 genotypes (G1/G1, G1/G2, and G2/G2) are grouped together as a homogeneous risk category.…”