GWAS Meta-Analysis of Glomerular Filtration Rate in Three Cohorts of Sickle Cell Disease Patients and In Vivo Functional Analysis Reveals Potential Nephropathy Candidate Genes

BACKGROUND: Sickle cell disease nephropathy (SCDN) is a common complication of sickle cell disease (SCD) associated with risk for early mortality (Platt et al., 1994; Elmariah et al, 2014). To identify potential genetic risk factors for SCDN, we performed genome-wide association studies (GWAS) for glomerular filtration rate (GFR) in three well-characterized SCD cohorts and performed in vivo functional analysis of one of the candidate genes in zebrafish.METHODS: Three previously described SCD cohorts were utilized in this analysis: Outcome Modifying Genes in Sickle Cell Disease (OMG-SCD) (Elmariah et al, 2014), Pulmonary Hypertension and Sickle Cell Disease with Sildenafil Therapy (Walk-PHaSST) (Machado et al, 2011) and Pulmonary Hypertension and the Hypoxic Response in Sickle Cell Disease (PUSH) (Minniti et al, 2009). Patients less than 16 years old were excluded from PUSH. GFR was estimated using the 'Modification of Diet in Renal Disease' (MDRD) study definition (Levey et al, 1999) and, in the OMG-SCD and Walk-PHaSST cohorts, dichotomized at the clinically relevant threshold of 90 ml/min/1.73m2. 1064 patients with complete data were included in the analysis (Table 1). Genotyping was performed using the Illumina Human610-Quad BeadChip (Illumina, San Diego, CA). Linear regression was utilized to test for association between each SNP and GFR, controlling for genome-wide principle components using PLINK (Purcell et al, 2008). Logistic regression was utilized for the analysis of GFR<90. A total of 479,389 SNPs in common across all three cohorts were tested in a meta-analysis using METAL (Willer et al, 2010). False discovery rate (FDR) p-values were generated using PROC MULTTEST in SAS (SAS Systems, Cary, NC). To target and functionally evaluate xylt1 in vivo, a morpholino (MO) was injected into zebrafish embryos at the 1-4 cell stage. A wild-type (WT) XYLT1 ORF clone (GenBank: BC156039) was transferred to the pCS2+ vector, and used as template to transcribe capped mRNA. XYLT1 mRNA was co-injected with xylt1-MO to rescue the phenotype and demonstrate assay specificity. Proteinuria was assessed in zebrafish models by quantification of GFP leakage as described (Zhou and Hildebrandt, 2012). Larva were scored at 5 days-post-fertilization (dpf) for cardiac or yolk-sac edema.RESULTS: For the meta-analysis of continuous GFR, the SNP with the most evidence for association was rs7553158 in the TNNI3K-FPGT locus (p= 7.1E-7). For the meta-analysis of GFR<90, the most significant association within a gene was rs7201659 in XYLT1 (p=1.7E-5). None of these findings met FDR significance. Nonetheless, knockdown of xylt1 in zebrafish embryos resulted in pericardial edema, a phenotype associated previously with glomerular filtration defects, which was confirmed by the detection of GFP leakage of an albumin equivalent protein in the surrounding media. Importantly, complementation of xylt1 morphants with WT human XYLT1 mRNA rescued significantly these filtration defects in developing embryos.DISCUSSION:The absence of mechanistic knowledge about SCDN has led to limited therapeutic options for patients. The present study seeks to identify genetic risk factors for SCDN and represents the largest GWAS of GFR in SCD patients to date. The GWAS was likely still underpowered, however, as none of the findings met FDR significance. Nonetheless, the most significant genes identified are excellent biologic candidates for kidney function. Fucose-1-Phosphate Guanylyltransferase (FPGT) converts GTP and β-l-fucose-1-phosphate to GDP-l-fucose, a process essential in the kidney to reutilize L-fucose from the turnover of glycolipids. It is also highly expressed in the porcine and human kidney (Pastuszak et al, 1998; GTEx, Broad Institute). Xylosyltransferase 1 (XYLT1, also known as XT-1) encodes the enzyme responsible for biosynthesis of heparin sulfate proteoglycans, which affect permeability of the glomerular basement membrane. Furthermore, the c.343G>T polymorphism in XYLT1 has been implicated in diabetic nephropathy (Schön et al, 2006). Here, we have identified putative novel nephropathy loci associated with kidney decline in the largest combined cohort of SCD patients to date. These results further our understanding of the underlying pathology of SCDN and have uncovered genetic markers that could be used for identification of at-risk SCD patients prior to the onset of kidney dysfunction.

Duke Scholars

Author Marilyn Jo Telen Medicine, Hematology

Author Allison Elizabeth Ashley-Koch Medicine, Nephrology