Genome-wide studies reveal factors associated with circulating uromodulin and its relationships to complex diseases
Yong Li, et al.
Yong Li, et al.
View: Text | PDF
Research Article Genetics Nephrology

Genome-wide studies reveal factors associated with circulating uromodulin and its relationships to complex diseases

Abstract

Uromodulin (UMOD) is a major risk gene for monogenic and complex forms of kidney disease. The encoded kidney-specific protein uromodulin is highly abundant in urine and related to chronic kidney disease, hypertension, and pathogen defense. To gain insights into potential systemic roles, we performed genome-wide screens of circulating uromodulin using complementary antibody-based and aptamer-based assays. We detected 3 and 10 distinct significant loci, respectively. Integration of antibody-based results at the UMOD locus with functional genomics data (RNA-Seq, ATAC-Seq, Hi-C) of primary human kidney tissue highlighted an upstream variant with differential accessibility and transcription in uromodulin-synthesizing kidney cells as underlying the observed cis effect. Shared association patterns with complex traits, including chronic kidney disease and blood pressure, placed the PRKAG2 locus in the same pathway as UMOD. Experimental validation of the third antibody-based locus, B4GALNT2, showed that the p.Cys466Arg variant of the encoded N-acetylgalactosaminyltransferase had a loss-of-function effect leading to higher serum uromodulin levels. Aptamer-based results pointed to enzymes writing glycan marks present on uromodulin and to their receptors in the circulation, suggesting that this assay permits investigating uromodulin’s complex glycosylation rather than its quantitative levels. Overall, our study provides insights into circulating uromodulin and its emerging functions.

Authors

Yong Li, Yurong Cheng, Francesco Consolato, Guglielmo Schiano, Michael R. Chong, Maik Pietzner, Ngoc Quynh H. Nguyen, Nora Scherer, Mary L. Biggs, Marcus E. Kleber, Stefan Haug, Burulça Göçmen, Marie Pigeyre, Peggy Sekula, Inga Steinbrenner, Pascal Schlosser, Christina B. Joseph, Jennifer A. Brody, Morgan E. Grams, Caroline Hayward, Ulla T. Schultheiss, Bernhard K. Krämer, Florian Kronenberg, Annette Peters, Jochen Seissler, Dominik Steubl, Cornelia Then, Matthias Wuttke, Winfried März, Kai-Uwe Eckardt, Christian Gieger, Eric Boerwinkle, Bruce M. Psaty, Josef Coresh, Peter J. Oefner, Guillaume Pare, Claudia Langenberg, Jürgen E. Scherberich, Bing Yu, Shreeram Akilesh, Olivier Devuyst, Luca Rampoldi, Anna Köttgen

×

Figure 5

B4GALNT2 p.Cys466Arg is a functional allele.

Options: View larger image (or click on image) Download as PowerPoint
B4GALNT2 p.Cys466Arg is a functional allele. (A) Homology-based model o...
(A) Homology-based model of the tridimensional structure of B4GALNT2 enzyme. The partial sequence of B4GALNT2 isoform 2 (Uniprot Q8NHY0-2) containing the glycosyltransferase domain (residues 254–464) was analyzed with iTasser. The top-scoring model is shown. The position of cysteine 406 (corresponding to position 466 in B4GALNT2 isoform 1, Uniprot Q8NHY0-1) in the reference sequence is shown. The effect of the p.Cys406Arg substitution was analyzed in Pymol with the Mutagenesis Wizard function. For each isoform, the most likely stereoisomer is shown. Visible red disks indicate significant contacts and bumps. The arginine substitution at position 406 is predicted to increase steric clashes, destabilizing protein structure. (B) Representative immunofluorescence analysis showing DBA signal (red) on the plasma membrane of unpermeabilized MDCK cells, transiently transfected with WT or p.Cys406Arg human B4GALNT2 (n = 3). (C) Representative immunofluorescence analysis showing WT B4GALNT2 (green) and DBA (red) in stably transfected MDCK cells. DBA signal is mostly evident in B4GALNT2-expressing cells (n = 3).