Evidence for a Dominant Negative Effect Conferred By the APOL1 G2 Sickle Cell Nephropathy Risk Allele in an in Vivo Model

BACKGROUND: Sickle cell disease (SCD) patients have a heterogeneous clinical course, and as patient survival has improved, end-organ damage has become an emergent clinical priority. End stage renal disease (ESRD), occurring in 5-18% of SCD patients, is a particular concern, because it is a major risk factor for early mortality (Platt et al. 1994). Detection of SCD nephropathy (SCDN) relies on relatively late-stage markers, namely proteinuria and reduced glomerular filtration rate (GFR). Consequently, at-risk SCD patients are not identified prior to end-organ damage. In non-SCD nephropathy, ESRD risk among African American individuals has been attributed to coding variants (termed G1 and G2) in apolipoprotein L1 (APOL1). The G1 allele consists of two nonsynonymous variants in perfect LD, rs73885319 and rs60910145 (encoding S384G and I384M), while the G2 variant consists of a six base pair deletion removing amino acids N388 and Y389 (~21% and ~13% allelic frequency in African Americans for G1 and G2, respectively). We demonstrated that these variants in APOL1 are strong predictors of risk for proteinuria in SCD (Ashley-Koch et al. 2011). Here, we use zebrafish as an in vivo model to both examine the role of apol1 in glomerular development and pronephric filtration and also to test the effects of APOL1 G1 and G2 expression in the developing kidney.

METHODS: A morpholino (MO) was designed by Gene Tools, LLC (Philomath, OR) to target the translation initiation site of zebrafish apol1. APOL1 G1 and G2 allelic constructs were synthesized from a wild-type (WT) APOL1 human open reading frame clone (GenBank: BC112943) using site-directed mutagenesis (Stratagene, QuikChange II), subsequently transcribed (mMESSAGE mMACHINE®, Life Technologies) into capped mRNA and co-injected with apol1-MO into zebrafish embryos. To assay glomerular filtration, 70 kDa FITC-conjugated dextran was injected into the cardiac venous sinus of 48 hour post-fertilization embryos. The eye vasculature of individual fish was imaged at two, 12, and 36 hours after dextran injection. The average fluorescence intensity was measured across the eye, and changes in intensity relative to the 2-hour post-injection measurements were calculated for comparison. Electron microscopy sections of five days post-fertilization embryos were cut on a Leica-Reichert Ultracut E ultramicrotome, and semi-thin sections (1.0μm) were stained and examined on a Phillips CM12 electron microscope.

RESULTS: As we showed previously (Anderson et al., ASH 2013), MO-induced suppression of apol1 in zebrafish embryos results in pericardial edema, glomerular filtration defects, and extensive podocyte loss. Importantly, complementation of apol1 morphants with WT human APOL1 mRNA rescues the observed kidney defects. However, we now show that neither APOL1 G1 nor G2 risk alleles ameliorate defects caused by apol1 suppression. Notably, injection of APOL1 G2 alone results in renal defects, as indicated by increased dextran clearance and the presence of microvillus protrusions in the urinary space. Injection of APOL1 G1 alone, however, does not induce noticeable kidney dysfunction. Furthermore, when APOL1 G2 injected embryos were titrated with increasing concentrations of human APOL1 WT mRNA, we observed a significant reduction of edema formation in developing embryos, suggesting a possible dominant-negative effect of the altered protein.

CONCLUSIONS: Unlike the WT APOL1 mRNA, neither the APOL1 G1 or G2 risk alleles could rescue kidney dysfunction due to knockdown of apol1 in zebrafish embryos, suggesting that these SCDN risk alleles impact the normal function of APOL1 in the kidney. Furthermore, development of edema with concomitant defects in glomerular ultrastructure in zebrafish embryos injected with APOL1 G2 mRNA alone suggest that this allele may act in a dominant-negative manner to induce kidney defects. Interestingly, it has been shown that APOL1 may cause toxic renal effects through programmed cell death pathways leading to glomerulosclerosis (Wan et al. 2008). Thus, apol1 suppression could result in the dysregulation of autophagic pathways, causing podocyte malformation and thereby affecting the susceptibility of the pronephros to glomerular injury. In summary, these data provide essential insight into the biological mechanisms by which APOL1 variants confer disease risk in human SCDN and other nondiabetic nephropathies.