APOL1 and Progression of Nondiabetic Nephropathy

African Americans are disproportionately affected by most common causes of nephropathy and develop ESRD at rates approximately 3.5 times higher than those in European Americans.1 Population ancestry-based disparities in nondiabetic nephropathy have recently been attributed to two variants in the C-terminal domain of the apolipoprotein L1 gene (APOL1) on chromosome 22—G1: rs73885319 and rs60910145, encoding two highly correlated nonsynonymous amino acid changes, and G2: rs71785313, a two–amino acid deletion.2 These alleles are nearly absent in populations of European and Asian ancestry, and their major impact in African Americans is evident in the disease that has historically been labeled hypertension-attributed ESRD (representing the primary disorder focal global glomerulosclerosis FGGS with interstitial fibrosis and vascular changes), FSGS, and HIV-associated nephropathy (HIVAN) with odds ratios (ORs) of 7.3, 17, and 29, respectively.3 This spectrum has expanded to include nephropathy related to sickle cell disease4 and severe lupus nephritis with collapsing lesions or lupus nephritis–associated ESRD.5 For unclear reasons, hyperglycemia does not appear to trigger APOL1-associated nephropathy. This is a perplexing observation and one worthy of additional study. African Americans develop diabetic kidney disease far more often than do European Americans, yet diabetes-associated ESRD does not appear to reside in the APOL1 disease spectrum. Prior associations in African Americans with clinically diagnosed diabetes-associated ESRD probably reflected a subpopulation of diabetic patients with FSGS-related kidney disease who lacked renal histology. In this issue of JASN, Foster and colleagues report the effect of APOL1 mutations on risk of CKD in a community-based sample from the Atherosclerosis Risk in Communities (ARIC) study.6 In this report, 3067 African American participants without baseline CKD were evaluated for the development of CKD and progression to ESRD during follow-up examinations spanning 20 years. Among the 13.2% (n=404) of the cohort carrying two APOL1 nephropathy alleles, the incidence of CKD was 11.64 per 1000 person-years, associated with a 1.51-fold increased risk of CKD. Although there was a low overall incidence of ESRD in ARIC (3.38 per 1000 person-years), after adjustment for the presence of two APOL1 nephropathy alleles there was an association with a 1.92-fold increased risk of ESRD. Both observations remained nominally significant in nondiabetic individuals when stratified by disease status. Moreover, in the sample of ARIC participants who developed CKD during follow-up, progression to ESRD was significantly more rapid in those with two APOL1 nephropathy alleles. The authors concluded that APOL1 risk variants contributed to the development of CKD, progression to ESRD, and rapidity of nephropathy progression in African Americans. Stated limitations were that ARIC participants were older than those in other population-based reports and that CKD diagnosis was based on estimated GFR (eGFR), without consideration of proteinuria. Participants with early-onset nephropathy would have been excluded, thereby probably reducing the observed strength of APOL1 association with CKD. To place these results in context, the Dallas Heart Study examined the prevalence of APOL1-associated nephropathy in a large population-based sample of 1825 African Americans and 1042 European Americans.7 Among nondiabetic African Americans carrying two APOL1 nephropathy alleles, a nearly three-fold increase in risk of microalbuminuria and four-fold increase in risk of CKD (eGFR 30 mg/g was present in 23% of relatives, consistent with the expected familial aggregation of nondiabetic ESRD in African Americans. Despite the markedly higher frequencies of APOL1 risk alleles, presence of two APOL1 risk variants weakly predicted mild kidney disease. After allowing for differences in study design, it is evident that case-control studies evaluating patients with severe forms of nephropathy or ESRD demonstrated markedly higher ORs for APOL1 association than did population-based and family-based samples containing patients with milder kidney disease or albuminuria. Further evidence of an APOL1 contribution to disease severity and progression of nephropathy came from the report by Lipkowitz et al. in the African American Study of Kidney Disease and Hypertension (AASK) cohort.9 In comparison to “presence of kidney disease” in all AASK patients, in whom association with APOL1 risk variants was observed with an OR of 2.57, restricting analyses to AASK patients who developed advanced kidney disease during follow-up or had higher baseline levels of UACR markedly improved the association (OR, 4.61–6.29). This report definitively demonstrated that a large percentage of nondiabetic African Americans who have CKD, hypertension, and low-level proteinuria actually have primary kidney diseases in the FSGS spectrum with secondary hypertension—not hypertensive nephrosclerosis. Taken together, these reports more consistently implicate APOL1 with nephropathy progression and reveal stronger association with severe kidney disease, relative to its effect on the initiation of kidney disease. In fact, the APOL1 association with five causes of kidney disease, each with distinctive glomerular pathology (FSGS, HIVAN, FGGS, sickle cell nephropathy, and severe lupus nephritis) also supports a role for the progression of nondiabetic forms of kidney disease. With the information provided by this and other epidemiologic studies, it has become evident that the presence of two APOL1 nephropathy alleles is necessary but insufficient for development of kidney disease. A two-hit model whereby an additional environmental exposure or gene variant increases (or diminishes) risk for kidney disease appears most likely because not all individuals with two APOL1 nephropathy variants will develop CKD. Different second hits may contribute to the variable glomerular histologic findings (FSGS, FGGS, collapsing lesions), whereas interstitial and vascular changes appear relatively consistent across the spectrum. Strong support for this hypothesis comes from HIVAN or HIV-associated collapsing glomerulopathy. HIVAN exhibits the greatest African–European ancestry difference in incidence rate of any kidney disease, with up to 50-fold higher rates in African Americans. Further, approximately 50% of HIV-infected patients possessing two APOL1 nephropathy variants developed HIVAN before the current era of highly active antiretroviral therapy. This therapy changed the natural history of HIVAN; incidence rates are falling despite the presence of APOL1 risk genotypes in many currently infected individuals. This demonstrates that treating an environmental exposure, such as HIV, which maintains a renal reservoir of infection, reduces the risk for a genetically mediated kidney disease. This finding provides new hope for development of novel treatments to prevent nondiabetic nephropathy that relate to genetic susceptibility. Other viruses that infect cells in the kidney and lower urinary tract may also interact with APOL1 to alter the risk for nondiabetic CKD.10 Identification of genes that interact with APOL1 (gene/gene interactions) and APOL1/environment interactions will more precisely define an individuals’ risk for subsequent development of progressive nephropathy and improve the diagnostic accuracy of screening efforts in high-risk populations. APOL1 encodes ApoL1, a protein involved in the autophagic pathway. ApoL1 is present in podocytes, renal tubule cells, and glomerular endothelial cells. Although synthesized in podocytes, a role for cellular uptake of ApoL1 from the glomerular filtrate in podocytes and renal tubule cells or direct effects of circulating (free or complexed) ApoL1 protein on glomerular endothelial cells and renal tubule cells remains possible. The precise mechanism whereby APOL1 causes progression of nondiabetic kidney disease remains under study. However, epidemiologic reports such as the one by Foster et al. are adding to our understanding of which patients are at highest risk and how to best apply genetic screening to African ancestry populations. Disclosure None. This work was supported by National Institutes of Health grants R01 HL56266, R01 DK070941, and R01 DK084149 (B.I.F.).