Large-scale whole-exome sequencing association studies identify rare functional variants influencing serum urate levels

Adrienne Tin, Yong Li, Jennifer A. Brody, Teresa Nutile, Audrey Y. Chu, Jennifer E. Huffman, Qiong Yang, Ming‐Huei Chen, Cassianne Robinson‐Cohen, Aurélien Macé, Jun Liu, Ayşe Demirkan, Rossella Sorice, Sanaz Sedaghat, Melody Swen, Bing Yu, Sahar Ghasemi, Alexander Teumer, Péter Vollenweider, Marina Ciullo, Meng Li, André G. Uitterlinden, Robert Kraaij, Najaf Amin, Jeroen van Rooij, Zoltán Kutalik, Abbas Dehghan, Barbara McKnight, Cornelia M. van Duijn, Alanna C. Morrison, Bruce M. Psaty, Eric Boerwinkle, Caroline S. Fox, Owen M. Woodward, Anna KöttgenNature Communications 2018 · DOI

TL;DR

This study leverages large-scale whole-exome sequencing to identify rare genetic variants that significantly influence serum urate levels, a key factor in gout development. They found that rare variants in urate transporters SLC22A12 and SLC2A9 are strongly associated with lower urate levels, and importantly, carrying these variants halves the risk of gout.

What Problem Does This Paper Address?

Gout, a painful inflammatory arthritis caused by elevated serum urate levels (hyperuricemia), affects a significant portion of the population and is often poorly managed. While previous genome-wide association studies (GWAS) have identified common genetic variants associated with serum urate, these variants typically have small effects. This leaves a substantial portion of the genetic contribution to serum urate and gout unexplained. The authors hypothesize that rare genetic variants, which can have larger individual effects, might play a crucial role in this missing heritability. They aim to uncover these rare variants and understand their functional impact on urate metabolism and gout risk.

Methods

The researchers employed a large-scale whole-exome sequencing (WES) association study approach. This involved analyzing the exomes (protein-coding regions of the genome) of a substantial number of individuals (up to 19,517) from European and African-American ancestries.

Key methodological aspects include:

  • Meta-analysis: Combining results from multiple cohorts to increase statistical power.
  • Burden tests and Sequence Kernel Association Test (SKAT-O): These are statistical methods specifically designed to test for associations with rare variants. Burden tests aggregate the effects of multiple rare variants within a gene, while SKAT-O is a more flexible approach that can detect associations even if variants have opposing effects.
  • Functional Validation: Selected rare variants identified in the association studies were further investigated through experimental methods:
  • Protein abundance assays: Using HEK293T cells to measure how much of the urate transporter protein is produced when specific variants are present.
  • Transport assays: Using Xenopus oocytes to measure the actual transport of urate by the transporter protein with different variants.
  • Protein structure mapping: Using computational models of the urate transporter to visualize where the identified variants are located and how they might affect protein function.

Datasets

The study utilized data from eight population-based cohorts comprising up to 19,517 European ancestry (EA) and African-American (AA) individuals. The specific cohorts included were:

  • Atherosclerosis Risk in Communities study (ARIC)
  • Cardiovascular Health Study (CHS)
  • Cilento study
  • CoLaus study
  • Erasmus Rucphen Family study (ERF)
  • National Heart, Lung, and Blood Institute GO Exome Sequencing Project (ESP)
  • Framingham Heart Study (FHS)
  • Rotterdam Study (RS)

The primary outcomes analyzed were serum urate levels, estimated glomerular filtration rate (eGFR), and urinary albumin-to-creatinine ratio (UACR). Gout prevalence was analyzed as a secondary outcome.

Key Insights and Ideas

  • Rare variants pack a punch: The study strongly suggests that rare genetic variants, particularly in urate transporters, have a disproportionately large impact on serum urate levels compared to common variants. This challenges the traditional focus on common variants in complex trait genetics.
  • URAT1 and GLUT9 are key players: The urate transporters SLC22A12 (URAT1) and SLC2A9 (GLUT9) emerged as critical genes, harboring multiple rare variants associated with serum urate. This highlights their central role in regulating urate homeostasis.
  • Loss-of-function variants are common and protective against gout: A significant finding is that many rare variants in SLC22A12 lead to a loss-of-function in the URAT1 transporter. Carriers of these variants have lower serum urate and, remarkably, a halved risk of developing gout. This provides a direct link between genetic variants, urate levels, and disease outcome.
  • Therapeutic potential of URAT1 inhibition: The experimental validation of loss-of-function variants in URAT1, and the known therapeutic effect of URAT1 inhibitors like lesinurad, strongly suggest that targeting URAT1 is a viable strategy for lowering serum urate and preventing gout.
  • Structural insights into GLUT9: Mapping variants onto the predicted structure of GLUT9 provided valuable insights into potential urate-binding sites and functional domains, paving the way for more targeted drug development.

Key Results

  • Aggregate association of rare variants: The study identified significant aggregate associations of low-frequency damaging variants in SLC22A12 (p = 1.3 × 10⁻⁵⁶) and SLC2A9 (p = 4.5 × 10⁻⁷) with serum urate levels.
  • Reduced gout risk: Carriers of rare SLC22A12 variants had a halved risk of gout (OR = 0.5, p = 4.9 × 10⁻³).
  • Functional validation of SLC22A12 variants: Three specific rare variants (R325W, R405C, and T467M) in SLC22A12 were confirmed as loss-of-function mutations through transport and protein abundance assays.
  • Structural mapping of SLC2A9 variants: Variants in SLC2A9 were mapped onto a predicted protein structure, revealing potential urate-binding residues and functional domains.

Limitations and Future Work

  • Model system limitations: The functional validation of variants was performed in cell lines and oocytes. While informative, these models may not perfectly replicate the complex in vivo environment of human kidneys. The interpretation of K536T's gain-of-function potential, for instance, requires caution due to the inability to assess protein abundance in human tissue.
  • Incomplete exome coverage: While WES covers most coding regions, gaps in capture can lead to missed variants.
  • Sample size for individual rare variants: Despite the large sample size, identifying rare variants with statistically significant individual associations remains challenging.
  • Comprehensive experimental testing: The sheer number of rare variants identified in SLC22A12 and SLC2A9 made comprehensive experimental testing of all of them infeasible.
  • Future Directions:
    • Further functional studies are needed to fully understand the mechanisms of identified variants, especially for those with less clear loss-of-function phenotypes.
  • Investigating common missense variants in known GWAS loci to pinpoint causal effects.
  • Leveraging larger, trans-ethnic datasets (e.g., from the TOPMed Program) to further explore rare variant associations.
  • Developing targeted therapies based on the identified functional domains and mechanisms of URAT1 and GLUT9.
  • Conducting genetic studies in populations with detailed information on urate-lowering therapies to gain complementary insights.

Summary updated

Abstract

Elevated serum urate levels can cause gout, an excruciating disease with suboptimal treatment. Previous GWAS identified common variants with modest effects on serum urate. Here we report large-scale whole-exome sequencing association studies of serum urate and kidney function among ≤19,517 European ancestry and African-American individuals. We identify aggregate associations of low-frequency damaging variants in the urate transporters SLC22A12 (URAT1; p = 1.3 × 10<sup>-56</sup>) and SLC2A9 (p = 4.5 × 10<sup>-7</sup>). Gout risk in rare SLC22A12 variant carriers is halved (OR = 0.5, p = 4.9 × 10<sup>-3</sup>). Selected rare variants in SLC22A12 are validated in transport studies, confirming three as loss-of-function (R325W, R405C, and T467M) and illustrating the therapeutic potential of the new URAT1-blocker lesinurad. In SLC2A9, mapping of rare variants of large effects onto the predicted protein structure reveals new residues that may affect urate binding. These findings provide new insights into the genetic architecture of serum urate, and highlight molecular targets in SLC22A12 and SLC2A9 for lowering serum urate and preventing gout.
BibTeX
@inproceedings{tin2018large,
  author = {Adrienne Tin and Yong Li and Jennifer A. Brody and Teresa Nutile and Audrey Y. Chu and Jennifer E. Huffman and Qiong Yang and Ming‐Huei Chen and Cassianne Robinson‐Cohen and Aurélien Macé and Jun Liu and Ayşe Demirkan and Rossella Sorice and Sanaz Sedaghat and Melody Swen and Bing Yu and Sahar Ghasemi and Alexander Teumer and Péter Vollenweider and Marina Ciullo and Meng Li and André G. Uitterlinden and Robert Kraaij and Najaf Amin and Jeroen van Rooij and Zoltán Kutalik and Abbas Dehghan and Barbara McKnight and Cornelia M. van Duijn and Alanna C. Morrison and Bruce M. Psaty and Eric Boerwinkle and Caroline S. Fox and Owen M. Woodward and Anna Köttgen},
  title = {{Large-scale whole-exome sequencing association studies identify rare functional variants influencing serum urate levels}},
  booktitle = {Nature Communications},
  year = {2018},
  doi = {10.1038/s41467-018-06620-4},
}
Figure