APOL1 genetic variants in focal segmental glomerulosclerosis and HIV-associated nephropathy

Abstract

Trypanolytic variants in APOL1, which encodes apolipoprotein L1, associate with kidney disease in African Americans, but whether APOL1-associated glomerular disease has a distinct clinical phenotype is unknown. Here we determined APOL1 genotypes for 271 African American cases, 168 European American cases, and 939 control subjects. In a recessive model, APOL1 variants conferred seventeenfold higher odds (95% CI 11 to 26) for focal segmental glomerulosclerosis (FSGS) and twenty-nine-fold higher odds (95% CI 13 to 68) for HIV-associated nephropathy (HIVAN). FSGS associated with two APOL1 risk alleles associated with earlier age of onset (P = 0.01) and faster progression to ESRD (P < 0.01) but similar sensitivity to steroids compared with other subjects. Individuals with two APOL1 risk alleles have an estimated 4% lifetime risk for developing FSGS, and untreated HIV-infected individuals have a 50% risk for developing HIVAN. The effect of carrying two APOL1 risk alleles explains 18% of FSGS and 35% of HIVAN; alternatively, eliminating this effect would reduce FSGS and HIVAN by 67%. A survey of world populations indicated that the APOL1 kidney risk alleles are present only on African chromosomes. In summary, African Americans carrying two APOL1 risk alleles have a greatly increased risk for glomerular disease, and APOL1-associated FSGS occurs earlier and progresses to ESRD more rapidly. These data add to the evidence base required to determine whether genetic testing for APOL1 has a use in clinical practice.

Figures

Figure 1.

Distribution of haplotypes and risk alleles. Four APOL1 haplotypes were observed in the African American study group: the major haplotype, denoted as (+) and comprised of three nonrisk alleles; the G1GM haplotype with two missense risk alleles; the rare G1G+ haplotype with one missense risk allele; and the G2Δ6 haplotype with the 6 bp deletion risk allele. The risk allele and haplotype frequencies are shown for FSGS and HIVAN case and control groups in African Americans, including healthy blood donors and HIV+ individuals with no kidney disease.

Figure 2.

Distribution of age of onset of primary FSGS for APOL1 risk allele groups. Age of onset of primary FSGS is presented for African Americans with 0 or 1 APOL1 risk alleles (n = 63) and with 2 APOL1 risk alleles (n = 117). Height of bars represents the percentage of subjects of each group falling in each age group. As shown, APOL1-associated FSGS tends to present at a younger age and at a narrower onset age range, with 70% of cases presenting between 15 and 39 versus 43% of carriers of 0 or 1 risk alleles (P = 0.0009) (SD of presentation age in years, 11.8 versus 16.2, P = 0.004).

Figure 3.

APOL1 variants and kidney survival. The Kaplan-Meier plot depicts the kidney survival among 92 African American patients with primary FSGS, with the number of subjects at risk at each time point shown. The curves differed by the log rank test (P < 0.01). The hazard ratio for progression was 2.3 (95% CI 1.43, 4), and median survival was 5 yr for subjects with 2 APOL1 risks alleles, compared with 13 yr for subjects with 0 or 1 APOL1 risk allele. Survival curves were similar for 0 and 1 risk allele subjects (P > 0.05) and are therefore combined.

Figure 4.

Odds ratios for the effects of the APOL1 G1 and G2 alleles. Odds ratios with confidence intervals for association of different G1–G2 genotypes with combined FSGS/HIVAN cases compared with subjects carrying no G1 or G2 risk alleles (+/+). Heterozygous carriers of the risk alleles do not show significant association with FSGS/HIVAN cases, with the exception that combined +/ G1GM and +/ G1G+ was marginally significant (OR 1.8, P = 0.02). By contrast, dual heterozygous and homozygous individuals showed a consistent association. In this analysis, the rare G1G+ haplotype is distinguished from the common G1GM haplotype and G1G+ is able to complement G1GM, resulting in disease association for this genotype. A meaningful odds ratio for G1G+/G2, carried by one subject, could not be determined.

Figure 5.

Worldwide frequency distribution of APOL1 variants G1 and G2. Genotypes of G1 and G2 are determined for 51 populations in the Centre d'Etude du Polymorphisme Humain (CEPH) HGDP, for the HapMap Luhya population from Kenya (International HapMap Project [Phase II]), and for African Americans (AA) and European American (EA) controls in the NIH FSGS study cohort. The allele frequencies of G1 (red), G2 (orange), and wild-type alleles (light blue) in each population are depicted in pie charts overlaid upon a world map. Allele frequencies for G1 and G2 for each population are shown in Supplementary Table 2.