Genetic loci and prioritization of genes for kidney function decline derived from a meta-analysis of 62 longitudinal genome-wide association studies

Abstract

Estimated glomerular filtration rate (eGFR) reflects kidney function. Progressive eGFR-decline can lead to kidney failure, necessitating dialysis or transplantation. Hundreds of loci from genome-wide association studies (GWAS) for eGFR help explain population cross section variability. Since the contribution of these or other loci to eGFR-decline remains largely unknown, we derived GWAS for annual eGFR-decline and meta-analyzed 62 longitudinal studies with eGFR assessed twice over time in all 343,339 individuals and in high-risk groups. We also explored different covariate adjustment. Twelve genome-wide significant independent variants for eGFR-decline unadjusted or adjusted for eGFR-baseline (11 novel, one known for this phenotype), including nine variants robustly associated across models were identified. All loci for eGFR-decline were known for cross-sectional eGFR and thus distinguished a subgroup of eGFR loci. Seven of the nine variants showed variant-by-age interaction on eGFR cross section (further about 350,000 individuals), which linked genetic associations for eGFR-decline with age-dependency of genetic cross-section associations. Clinically important were two to four-fold greater genetic effects on eGFR-decline in high-risk subgroups. Five variants associated also with chronic kidney disease progression mapped to genes with functional in-silico evidence (UMOD, SPATA7, GALNTL5, TPPP). An unfavorable versus favorable nine-variant genetic profile showed increased risk odds ratios of 1.35 for kidney failure (95% confidence intervals 1.03-1.77) and 1.27 for acute kidney injury (95% confidence intervals 1.08-1.50) in over 2000 cases each, with matched controls). Thus, we provide a large data resource, genetic loci, and prioritized genes for kidney function decline, which help inform drug development pipelines revealing important insights into the age-dependency of kidney function genetics.

Keywords: acute kidney injury; chronic kidney disease; diabetes; gene expression.

Figure 1:. Eleven loci identified by GWAS for eGFR-decline unadjusted and/or adjusted for eGFR-baseline.

We conducted GWAS for eGFR-decline baseline-unadjusted and baseline-adjusted (n up to 343,339 or 320,737, respectively). Shown are association P-values versus genomic position, identified loci annotated by nearest gene: (A) association for eGFR-decline baseline-unadjusted identified one genome-wide significant locus for decline (P<5×10⁸) and two Bonferroni-corrected significant loci among the 263 lead variants for cross-sectional eGFR outside of UMOD-PDILT (red dots, P<0.05/263=1.90×10⁻⁴; known locus for decline marked in blue; novel loci for this phenotype in orange); (B) association for eGFR-decline baseline-adjusted identified 8 additional loci (novel loci marked in green; known loci or loci already identified in (A) marked in blue). Altogether, 11 loci were identified with genome-wide significance for eGFR-decline unadjusted and/or adjusted for eGFR-baseline.

Figure 2:. Relationship of SNP-effects on eGFR-decline baseline-unadjusted with baseline-adjusted effects for the 12 identified variants.

Shown are: (A) SNP-effects per year and allele for eGFR-decline baseline-unadjusted (“decline”) versus eGFR-decline baseline-adjusted in all studies (n_decline=343,339; n_decline-adj=320,737) and restricted to studies where baseline-adjusted results were computed rather than formula-derived (inserted panel, n=103,970); red line indicates identify line); (B) standardized SNP-effects per year and allele for eGFR-decline baseline-unadjusted (${\widehat{\beta }}_{DECLINE}/s{d}_{DECLINE}$, n=343,339) and per allele for cross-sectional eGFR on ln-scale (${\widehat{\beta }}_{BL}/s{d}_{BL}$, n=765,348 ); grey line indicates phenotype correlation line y=0.34*x (0.34=mean phenotype correlation across studies). For A&B: coding allele is the faster-decline allele (=cross-sectional eGFR-lowering allele). Color codes whether SNP was identified for decline baseline-unadjusted and/or baseline-adjusted. (C) Illustration of the SNP-effect for eGFR-decline baseline-adjusted (standardized to Y-scale) as a sum of the SNP-effect baseline-unadjusted (standardized) and the correlation-weighted SNP-effect on eGFR at baseline (standardized).

Figure 3:. Relationship of SNP-by-age interaction effects for eGFRcys with those of eGFRcrea and with SNP-effects for eGFR-decline for the 12 identified variants.

Shown are SNP-by-age interaction effect sizes per year and allele for cross-sectional eGFRcys (UK Biobank individuals independent from GWAS, n_SNPxage=351,601; main age effect modelled non-linearly, main SNP-effect linearly, age effect and SNP effect in interaction term linearly, age centered at 50 years) versus: (A) SNP-by-age interaction effects on cross-sectional eGFRcrea (n_SNPxage=351,462), (B) SNP-effects on eGFR-decline baseline-unadjusted per year and allele (n_decline=343,339). Coding allele is the faster-decline allele (=cross-sectional eGFR-lowering allele); color code as in Figure 2; red line indicates identity line; symbol types code significance of interaction term (P< 0.05/12). Among the 9 SNPs with genuine eGFR-decline association, 7 SNPs showed interaction for eGFRcrea or eGFRcys (all negative), and all 3 SNPs without genuine eGFR-decline association showed no interaction for eGFRcys (one with positive significant interaction for eGFRcrea).

Figure 4:. A concept for three classes of SNP-associations on cross-sectional eGFR distinguished by the presence and direction of the SNP-association with eGFR-decline.

Let A/a be the genotype group of individuals with, on average, lower cross-sectional eGFR compared to a/a (A=effect allele). Let’s further assume that eGFR-declines monotonously by age (approximated as linear decline) and that there is no “cross-over” between genotype groups. Shown are (left) a graphical scheme, (middle) the theoretical association, (right) the observed SNPs in line with the respective class. In the three graphical schemes, black lines illustrate mean eGFR-decline by genotype group; SNP-effects on eGFR for these individuals captured cross-sectionally at different ages are magenta. When a cross-sectional study captures individuals of relevant ages, the SNP-effects on eGFR should show an interaction by age for class II and class III SNPs (positive and negative, respectively). The 9 variants with genuine eGFR-decline association were class III, while the other 3 variants were class I.

Figure 5:. Data, analyses, and results in a nutshell.