APOL1 at ten years: Progress and next steps

Abstract

APOL1 kidney risk variants (RVs) were identified in 2010 as major drivers of glomerular, tubulointerstitial and renal microvascular disease in individuals with sub-Saharan African ancestry. In December 2020, the “APOL1 at Ten” conference summarized the first decade of progress and discussed controversies and uncertainties that remain to be addressed. Topics included trypanosome infection and its role in the evolution of APOL1 kidney RVs, clinical phenotypes in APOL1-associated nephropathy, relationships between APOL1 RVs and background haplotypes on cell injury and molecular mechanisms initiating disease, the role of clinical APOL1 genotyping, and development of novel therapies for kidney disease. Future goals were defined, including improved characterization of various APOL1 RV phenotypes in patients and experimental pre-clinical models; further dissection of APOL1-mediated pathways to cellular injury and dysfunction in kidney (and other) cells; clarification of gene-gene and gene-environment interactions; and evaluation of the role for existing and novel therapies.

Trypanosomes, evolutionary history

Trypanosoma are unicellular parasitic protozoa. Trypanosoma brucei ssp. causes potentially fatal human sleeping sickness as they move between the mammalian host and the tsetse fly. Human blood contains trypanolytic factors (TLF) comprised of haptoglobin related protein, apolipoprotein L1 (APOL1), APOA-1 and Immunoglobulin M¹. APOL1 is the main lethal factor in the TLF. After endocytosis of the TLF by trypanosomes, APOL1 is activated by a two-step process: activation at acidic pH followed by channel opening at neutral pH². APOL1 protein then creates cation selective channels in membranes leading to depolarization, a continuous influx of sodium and calcium, followed by chloride and subsequent osmotic swelling of the cell and internal organelles (Figure1).

Figure1.

APOL1 forms cytotoxic cation channels in a pH-dependent manner

These effects of APOL1 are inhibited in human-infective T.b. rhodesiense³ and T.b. gambiense^{4, 5}, each through a different resistance mechanism. This leads susceptible individuals to suffer from sleeping sickness, an acute infective form more common in East Africa caused by T.b. rhodesiense, or an indolent form more common in West Africa caused by T.b. gambiense. In turn, APOL1 G1 and G2 RVs modify susceptibility to African sleeping sickness. The G2 allele, even in the heterozygous state, confers protection from infection by T.b. rhodesiense, while G1 reduces severity of illness due to T.b. gambiense. However, homozygosity and compound heterozygosity (high risk genotypes: HR) increase susceptibility to nephropathy. The recent H3Africa study showed that the highest G1 frequencies were in West Africa, such as in the Côte d’Ivoire (43%) and the Fon population in Benin (34%)⁶. This is consistent with geographical correlation between the prevalence of APOL1 RVs and different forms of sleeping sickness. Interestingly, the frequency of G1 was markedly different between the Bantu vs Nilo-Sharan speaking Ugandans (14 vs 2%), in contrast to similar frequencies of G2 in these groups^6?.

In summary, the only currently known biological function of APOL1 is conferring protection against trypanosomiasis. To increase protection from a potentially deadly disease, the allele frequency RV APOL1 rapidly increased in Africa⁷. However, the newly evolved APOL1 RVs in homozygous form is associated with an increased long-term risk of kidney disease. Future studies should examine the role of the full spectrum of observed coding variants in APOL1 and their role in trypanosomiasis and kidney disease.

Clinical phenotypes

Familial aggregation of ESKD in African American families suggested an inherited basis for progressive nephropathy nearly 30 years ago^8–10. This concept was validated with the discovery of APOL1 association with a spectrum of non-diabetic forms of glomerulosclerosis in individuals with recent African ancestry^{11, 12}, including focal segmental glomerulosclerosis or FSGS (all varieties), collapsing glomerulopathy (collapsing FSGS), and the syndrome of solidified or diffuse glomerulosclerosis with low level proteinuria (often mis-labelled arterionephrosclerosis or hypertensive nephropathy)^13–15. As shown in Table 1, kidney disease associated with sickle cell disease, systemic lupus erythematosus, and allograft failure in kidneys transplanted from donors with APOL1 HR genotypes also reside in the APOL1-associated nephropathy spectrum^16–19.

Table 1.

APOL1-associated disease spectrum

Diagnosis	Manifestations, risk conditions, diagnosis	APOL1 association	Odds ratio
Primary FSGS	Nephrotic proteinuria, low serum albumin	Yes	17
Post-adaptive FSGS	Risk: Low birth weight, prematurity, increased body size Feature: normal serum albumin	Yes	ND
APOL1 FSGS	May mimic other forms	Yes	See other
Virus-associated FSGS	HIV-1, cytomegalovirus, Epstein—Barr virus, Sars-CoV-2, others	Yes: HIV-1	29 (USA), 89 (South Africa)
Drug-associated FSGS	Many medications	Interferon	ND
Hypertension-attributed ESKD or presumed arterionephrosclerosis	Hypertension, sub-nephrotic proteinuria and progressive loss of eGFR	Yes	7
Sickle cell nephropathy	Proteinuria, progressive renal disease	Yes	2.7–5.4
Pre-eclampsia	Hypertension, proteinuria	Yes	1.8, 1.9

One end of the disease spectrum, which shows the strongest association with APOL1 RV is collapsing glomerulopathy. It is seen with extremely high interferon levels, including interferon administration, HIV infection (HIV-associated nephropathy, HIVAN), systemic lupus, and SARS-CoV-2 infection (COVID-19 associated nephropathy, COVAN)^{20, 21}. This fits the concept that interferon drives APOL1 transcription, resulting in abundant APOL1 RV protein in kidney cells. It is possible that massive immune system activation related to SARS-CoV-2 may occasionally cause nephropathy in individuals with G1G0 genotype, as reported with HIVAN on the African continent and a recipient of a G1G0 deceased donor kidney²².

While kidney disease is the major manifestation of RV APOL1, they may also be associated with other phenotypes, such as sepsis and pre-eclampsia^{23, 24}. In three independent studies, the high-risk genotype was associated with increased odds of pre-eclampsia in African-Americans (odds ratio [OR] 1.4 to 3.3), two under a recessive model and one under a dominant model^{23, 25, 26}. In each study, risk was conferred by the genotype of the fetus. The mother’s country of origin (USA vs Haiti) and discordance between maternal and fetal APOL1 genotype seemed to modify the association. In two independent studies of black patients with nephrotic syndrome, a high risk (HR) APOL1 genotype was associated with increased odds of having been born preterm (OR 2.7, 9.6)^{27, 28}. Finally, ~50% of black children with nephrotic syndrome or proteinuric kidney disease in the Nephrotic Syndrome Study Network (NEPTUNE) and Chronic Kidney Disease in Children (CKiD) had an APOL1 HR genotype and these children had a significantly lower estimated glomerular filtration rate (eGFR) at presentation and more rapid decline of eGFR²⁷.

The phenotypic heterogeneity associated with APOL1 HR genotypes is poorly understood and future studies shall focus on deeper mechanistic understanding. Risk conferred by APOL1 in non-renal diseases appear much smaller than for collapsing FSGS, a role for APOL1 RVs in cardiovascular disease and hypertension independent of kidney disease has not yet been established^{24, 29, 30}. APOL1-associated nephropathy is an autosomal recessive disorder with incomplete penetrance. While, one APOL1 RV is sufficient to protect from trypanosomiasis, increased disease risk is predominantly observed in people carrying 2 RVs.

APOL1-induced cell injury in mammalian cells

APOL1-induced trypanosome killing has been studied for decades. However, APOL1-induced toxicity in mammalian cells remains poorly understood. Drs. Pollak, Olabisi, Scales, Raper and Susztak addressed aspects of cell injury^{5, 28, 31} and molecular mechanisms^{32, 33} using cell culture and animal models³⁴ (see Figure 2).

Figure 2.

Proposed mechanisms of APOL1-mediated cytotoxicity

Scales emphasized the critical role of APOL1 haplotypes for G1 and G2-mediated cytotoxicity in cultured podocytes stably expressing APOL1 RVs^{35, 36}. Specifically, the podocytotoxicity was dependent on the African APOL1 haplotype (E150/I228/K255), not simply the specific RV, even though there were no differences in localization or expression level between haplotypes³⁷. She demonstrated that critical thresholds of expression at the plasma membrane must be surpassed to achieve cytotoxicity. African G1/G2 APOL1 exits the endoplasmic reticulum and reaches the cell surface to cause toxicity (in agreement with Giovinazzo et al.³⁸), as indicated by the rescue effect of the ER-Golgi transport inhibitor Brefeldin A. Additionally, N-terminal splice variants of APOL1 lacking a fully functional signal sequence (isoforms B3 and C), which localize to the cytoplasmic face of the ER more than plasma membranes^{35, 36, 39}, were non-toxic⁴⁰. She emphasized that future publications should report the full sequence of the APOL1 variants used for in vitro and in vivo studies.

In a talk entitled “Re-envisioning the APOL1 cation channel,” Raper showed interesting results produced by studying the channel function of APOL1 reconstituted in an in vitro bilayer system^{4, 38}. Her team showed a population of active channels at the plasma membrane, resulting in an influx of Na⁺ and Ca²⁺ ions in human cells expressing APOL1 RVs. Interestingly, wild type G0 and RV APOL1 did not differ in ion-conducting properties but manifested differences in pH gating and membrane insertion³⁸. All variants of APOL1 traffic to the plasma membrane, en route they encounter acidification and neutralization along the secretory pathway, steps required for channel formation. However, while G1 and G2 are able to form cation channels when overexpressed, G0 does not. They hypothesize that G1 and G2 are activated at higher pH than G0, leading to channel formation (Figure 1). She suggested a model wherein APOL1-influenced channel activity is the upstream event causing cell death. Of particular importance, she also emphasized a potential public health issue due to the rapid evolution of Trypanosomal species in Africa in considering therapies which may only partially inhibit APOL1. These could lead to emergence of resistance mechanisms in parasites.

Susztak emphasized that conditional and inducible expression of G1 or G2 variants in mouse podocytes⁴¹ induced proteinuria, foot process effacement, global glomerulosclerosis and tubulointerstitial fibrosis; changes observed in patients with APOL1 HR genotypes^{42, 43}. These studies firmly established that APOL1 RVs are disease-causative. Furthermore, gene expression analysis by RNA sequencing of tissue from mouse models with podocyte-specific RV expression showed striking similarities to those in patients with APOL1-associated FSGS. Disease severity in mice correlated with APOL1 genotype and total APOL1 expression. Her team observed differences in intracellular trafficking of APOL1, including a defect in acidification of endocytic vesicles causing a defect in autophagy and downstream activation of the inflammasome^{40, 44, 45}.

Pollak noted that the G1 and G2 alleles are associated with different phenotypes in mice and humans, while G0 is generally less toxic in most but not all experimental systems⁴⁶. Higher APOL1 RV expression levels associate with increasing toxicity in humans and experimental models. In most cases, two RVs are required for toxicity^{37, 47}. The requirement for two RVs to cause disease and the recessive but apparent gain-of-function nature of the two RVs is poorly understood but may reflect a gene dosage effect, wherein by more APOL1 risk protein is made when two risk alleles are present.

Olabisi extended discussion of the two-RV requirement and the role of increased gene expression driving toxicity^{46, 48}. He concluded that carriage of the RVs and their dose contribute to toxicity in vitro, with no evidence for a dominant protective effect of G0 on G1/G2-induced toxicity in HEK-293 cells. In addition, in vitro studies support that podocyte toxicity is associated with a mitochondrial defect^49–51. This mitochondrial defect might mimic that observed in trypanosomes⁵².

While the exact molecular mechanism is not fully understood, several key themes emerged, such as the increased toxicity of RVs compared to reference G0, the dose-dependency of the toxicity, the critical role of background local haplotype and intracellular trafficking that together culminate into APOL1 channel activity and likely activation of multiple downstream pathways. In particular, we need to understand which injury mechanisms are important in which cells and under what conditions. There is the possibility that therapy directed against one or more injury pathways may slow, halt or reverse APOL1-associated nephropathy.

Biomarkers for predicting kidney disease

Although APOL1 RVs are major drivers of disease, only 15–20% of people with HR genotypes develop nephropathy that typically includes proteinuria. Of these, a small subset develop nephrotic syndrome, but most people with nephropathy lack heavy proteinuria. Thus, detection of sub-nephrotic proteinuria, possibly including microalbuminuria, and modest reductions in eGFR may be early biomarkers for intervention⁵³.

Model system studies established the key role of inflammation, especially viral infection and interferon, in precipitating disease. To improve risk stratification among APOL1 HR individuals, relationships between APOL1 and biomarkers of immune system activation and kidney injury have been studied. The growing list of biomarkers now includes soluble urokinase-type plasminogen activator receptor (suPAR), tumor necrosis factor receptor 1 (TNFR1), TNFR2 and kidney injury molecule-1 (KIM1). Inflammation is associated with an increase in APOL1 expression triggering nephropathy in APOL1 HR genotype individuals⁵⁴.

Inflammation and immune system activation increase plasma suPAR levels. Reiser et al. reported that a tripartite complex containing APOL1 protein, suPAR and integrin avb3 forms on the cell surface of podocytes. This complex can lead to cellular activation with podocyte detachment from the glomerular basement membrane, podocytopenia, proteinuria and glomerulosclerosis⁵⁵. Higher plasma suPAR levels are associated with development of CKD in African Americans from the Emory Cardiovascular Biobank and African American Study of Kidney Disease and Hypertension (AASK). The authors proposed an APOL1-environment interaction that could cause CKD in the setting of high suPAR levels related to inflammation or immune activation.

Nadkarni and colleagues demonstrated that plasma levels of kidney injury marker 1 (KIM1), tumor necrosis factor receptor (TNFR)1, and TNFR2 levels helped stratify individuals with APOL1 HR genotypes at risk for nephropathy⁵⁶. Among 498 HR APOL1 genotype participants in the BioMe Biobank Program in the Mount Sinai Health System 80 developed ESKD or sustained 40% decline in eGFR after 5.9 years. Baseline TNFR1, TNFR2, and KIM1 concentrations were independently associated with these composite renal outcomes and inclusion of these biomarkers significantly improved prediction. When the concentration of all three biomarker were elevated, the event rate was 40%, compared to 17% in participants who had an increase in only one or two biomarkers, and 7% with zero biomarkers. Machine learning algorithms incorporating electronic health records and biomarkers identified individuals with APOL1 HR genotypes at risk for nephropathy progression.

The absence of JC viruria presumably reflects a heightened immune response that prevented or eradicated viral infection; this heightened immune response could also be associated with immunologically driven APOL1 expression. For example, colonization of the kidney and lower urinary tract with JC polyoma virus occurs in ~30% of the general population. Individuals with JC polyoma viruria and APOL1 HR genotypes appear less likely to develop nephropathy compared to those lacking JC viruria. The heightened immune response in virus-free individuals presumably prevented or eradicated viral infection⁵⁷. This finding extends to non-APOL1-associated forms of CKD in African Americans, supporting a role for immune system activation in nephropathy historically attributed to hypertension or hyperglycemia⁵⁸.

Select environmental exposures may induce a more fulminant form of nephropathy in those with an APOL1 HR genotype. These include interferon administration and HIV-1 and SARS-CoV-2 infections triggering intense immune responses including interferon release^{20, 59}. Air pollution, fine particulate matter <2.5 μm (PM_2.5), might also be an environmental trigger for APOL1-associated nephropathy⁶⁰. PM_2.5 and APOL1 were independently associated with nephropathy in individuals with recent African ancestry in New York. However, the effect of PM_2.5 was stronger in individuals with an APOL1 HR genotype. If replicated, air pollution could prove to be an environmental factor that accentuates CKD risk. The mechanisms whereby exposure to air pollution increases risk are unclear. In contrast to these environmental exposures, major APOL1-second gene interactions have yet to be identified, although genetic modifiers with low allele frequencies likely exist⁶¹.

Role of APOL1 genotyping

The role of APOL1 genotyping in kidney transplantation and nephrology remains controversial as the majority of individuals with HR genotypes will not develop CKD. Genetic risk may have a negative psychological impact on some people specially because effective therapies are yet to come. However, novel therapies are on the horizon and instances exist where genotyping may prove helpful^{62, 63}. For example, APOL1 genotyping in those with recent African ancestry may improve compliance in patients with HIV infection or lupus at risk for nephropathy, guide physicians prior to administering interferon, provide a more definitive diagnosis/etiology in patients with FSGS or bland chronic kidney injury, detect non-diabetic nephropathy in individuals with diabetes mellitus, and assist with family planning. These require formal study, although prospective controlled trials involving provision of genetic information to subsets of participants will be challenging. APOL1 genotyping is required to determine eligibility for treatment trials using small molecule inhibitors and anti-sense oligonucleotide therapy and fortunately more such trials are likely. Widespread genotyping will likely only occur once effective therapies are available. Many transplant programs have incorporated genotyping in the living donor work-up and the APOL1 Long-term Kidney Transplant Outcomes (APOLLO) study is assessing transplant outcomes based on deceased and living kidney donor genotypes⁶³.

Current quandaries, therapeutics and salient questions for future research

Although there was strong consensus on many key issues, several important topics require further investigation. One such issue is whether or not there are differences between G1 and G2 variants. While initial studies indicated similar trypanolytic activities for the APOL1 G1 and G2 variants, recent studies indicate differences in SRA binding by the G1 and G2 variants that lead to differences in protection against T.b. rhodesiense and T.b. gambiense. This is in line with recently observed differences in allele frequencies in the H3Africa Study. Differences in development of kidney disease have also been observed. For example, individuals with heterozygous G1 allele showed increased incidence of ESKD, FSGS and HIVAN compared to G0/G0 individuals, and incidence was also higher in G1 risk allele subjects compared to G2. G1 was associated with younger ESKD age, a finding not present for G2 subjects. Future studies shall focus on understanding differences between G1 and G2 alleles, local background haplotype structure and a variety of other coding variants in APOL1.

APOL1-conferred protection from trypanosomiasis is due to its ion channel function. Mechanistic studies indicate that APOL1 can also function as an ion channel in mammalian cells (Figure 1). Channel activation patterns and differences between risk and reference allele are poorly understood. APOL1 undergoes complex intracellular trafficking, where changes in intracellular pH appear to be important for membrane insertion and channel function. In vitro reprogrammed and CRISPR-edited podocytes appear to be a promising new model system to understand these processes. Cell-type specific conditional inducible RV APOL1 transgenic mice or BAC transgenic animals have proven to recapitulate aspects of human disease and should be important for preclinical analyses. Ongoing clinical studies are targeting either the APOL1 channel function or the APOL1 dose, consistent with molecular studies.

Other critical issues include who should be treated and when should treatment start. Detection of biomarkers prospectively identifying high risk individuals remain critical for development of therapeutics. Early studies target the highest risk group, such as those with FSGS and heavy proteinuria. Another key question is whether the disease is still driven by APOL1 dose or channel activity at this stage. Furthermore, while these therapeutic modalities might be practical in areas where there is little risk for trypanosomiasis, it will be critical to understand their effect on parasitic infection. Finally, it will be important to determine whether patients with APOL1-associated nephropathy respond to SGLT2 inhibitors, a drug class with protective effects on kidney and cardiovascular disease.

Conclusions

Remarkable progress has been made in the past 10 years, since identification of APOL1 as the gene responsible for much of the increased risk of CKD among African Americans. We have a better understanding of the global distribution of populations affected by APOL1-associated nephropathy, clinical manifestations of disease and molecular pathways contributing to dysfunction in cultured cells, animal models, human cells, and tissues. Most exciting, new therapies are being developed and one is in a phase II trial stage.

The first ten years of this emerging field were characterized by productive collaborations across an array of disciplines, including molecular biology, protein chemistry, cell biology, animal models, pharmaceutical development, clinical research, and sociology. This comes at a time of renewed hope for preventing or slowing CKD progression in the clinic. Above all, the perspectives of patients, families, and communities was sought, considered, and acted upon in manuscripts and multi-disciplinary discussions. These interactions engendered efforts to communicate research advances and therapeutic options to the African American community and empower community members to craft appropriate messages and serve as messengers. Together, we believe novel therapies will emerge that change the narrative for many individuals currently destined to require renal replacement therapy leading to lives with preserved kidney function.