Pa g e 2
Significance statement (120 words summary)Increased albuminuria is a key manifestation of major health burdens, including chronic kidney disease and/or cardiovascular disease. Although being partially heritable, there is a lack of knowledge on rare genetic variants that contribute to albuminuria. The current study describes the discovery and validation, of a new rare gene mutation (~1%) in the CUBN gene which associates with increased albuminuria. Its effect multiplies 3 folds among diabetes cases compared to nondiabetic individuals. The study further uncovers 3 additional genes modulating albuminuria levels in humans. Thus the current study findings provide new insights into the genetic architecture of albuminuria and highlight novel genes/pathways for prevention of diabetes related kidney disease. P a g e 3
AbstractIdentifying rare coding variants associated with albuminuria may open new avenues for preventing chronic kidney disease (CKD) and end-stage renal disease which are highly prevalent in patients with diabetes. Efforts to identify genetic susceptibility variants for albuminuria have so far been limited with the majority of studies focusing on common variants.We performed an exome-wide association study to identify coding variants in a two phase (discovery and replication) approach, totaling to 33,985 individuals of European ancestry (15,872 with and 18,113 without diabetes) and further testing in Greenlanders (n = 2,605). We identify a P a g e 4 P a g e 6 P a g e 7 run (wherever required). Replication meta-analyses with p replication <0.05, were considered significant.
Combined (EUR/EUR-GL) Meta analysesThe combined meta-analyses was firstly performed with all individuals of European ancestry (Combined EUR) followed by pooling of Greenlandic data (Combined EUR-GL).Any SNP with a) p replication < 0.05 and b) p meta_EUR/EUR-GL < 5.0 × 10 -8 was considered overall significant while those with 5.0 × 10 -5 > p meta_EUR/EUR-GL > 5.0 × 10 -8 were considered suggestive.
Conditional analysesConditional analyses for novel SNPs identified in known loci (and/or in low LD, LD r2 <0.01) were performed to determine if the signal was independent (in discovery set). This was performed using the following linear model:trait ~ top identified SNP + secondary known SNP + other covariates. If the top SNP retained the association estimates and p value it was considered an independent signal.
Gene Aggregate TestsGene-based multi-marker association testing for rare and common (MAF > 0.0001) variants was performed using the Meta Analyses for SNP-Set (Sequence) Kernel Association Test (MetaSKAT) R package (https://cran.r-project.org/web/packages/MetaSKAT/index.html). At the study-specific level, the gene-based analyses were performed against the null model accounting for gender and ten principal components, (using the previously described SKAT-O method) 14 generating SKAT objects individually for each cohort with complete ExomeWAS data (discovery studies + MDCS study, n studies = 6)) which were then meta-analyzed.The ...