Genome-Wide Association Study of Creatinine Clearance Identifies New Loci for Kidney Function

This large-scale genome-wide association study of creatinine clearance in nearly 59,000 individuals identifies 16 genetic loci influencing kidney function, including two novel variants found in both sexes and two female-specific loci, thereby highlighting the advantages of using muscle-mass-independent measures to uncover sex-specific genetic architectures in chronic kidney disease.

The Gist

Imagine your kidneys are the master filtration plants of your body, constantly cleaning your blood and removing waste. For decades, scientists have tried to figure out how the "genetic blueprint" (your DNA) controls how well these plants work.

However, there was a big problem with the old way of measuring kidney health. It was like trying to judge the efficiency of a water filter by looking at the dirtiness of the water coming out, but forgetting that the amount of water depends on how much the factory workers (your muscles) are sweating.

The Problem: The "Muscle Mass" Confusion

Most previous studies used a metric called eGFR (estimated Glomerular Filtration Rate). This is a calculation based on creatinine, a waste product your muscles produce.

• The Analogy: Imagine you are judging a car's engine speed. If you have a huge, muscular engine, it produces a lot of exhaust (creatinine). If you have a tiny engine, it produces less. But if you only look at the exhaust to guess the engine speed, you might think the big engine is running faster than it actually is, just because it's bigger.
• The Issue: In humans, men usually have more muscle than women. So, standard tests often overestimate kidney function in muscular people and underestimate it in those with less muscle, making it hard to find the true genetic causes of kidney issues.

The Solution: A New "Pure" Measurement

The researchers in this paper decided to use a different tool: Creatinine Clearance (CrCl).

• The Analogy: Instead of just guessing based on exhaust fumes, they installed a direct flow meter on the pipe. They measured exactly how much waste the kidneys filtered out over 24 hours. This method ignores how much muscle you have and tells you exactly how well the kidney filter is working.

The Big Discovery: Finding New "Genetic Switches"

The team analyzed the DNA of nearly 60,000 people (mostly of European descent) from the Lifelines Cohort Study in the Netherlands. They looked for "switches" in the DNA that turned kidney function up or down.

Here is what they found:

1. Two Brand New "Switches" (The Novel Loci)
They found two genetic locations that no one had ever linked to kidney function before.

• Switch A (AGPAT4): This gene is like a lipid (fat) manager. It helps the body process fats. The study suggests that if this switch is "stuck," it might mess up the fat balance in the kidney cells, causing the filter to clog or slow down.
• Switch B (IGF2R): This is a growth regulator. It controls how cells grow and repair themselves. If this switch is off, the kidney cells might not repair themselves properly after damage.

2. The "Female-Only" Switches
The researchers noticed that some genetic switches only worked in women.

• The Analogy: Imagine a factory where the female workers have a special safety protocol that the male workers don't use.
• The Findings: They found two specific genetic spots that only affected kidney function in women. One involved a protein called GPC6, which helps build the "scaffolding" of the kidney filter. This explains why kidney disease might behave differently in men and women at a genetic level.

3. The "Muscle Trap" Revealed
When they compared their new "pure" results with the old "muscle-confused" results, they found something fascinating.

• Some genetic switches that looked like they controlled kidney function in the old studies were actually just controlling muscle size.
• The Analogy: It was like finding a switch that turned on a "muscle growth" lightbulb, and the old test thought it was turning on the "kidney filter." Now that they have the new flow meter, they can see that those switches were just about muscles, not the kidneys.

Why This Matters

This study is like upgrading from a blurry, distorted map to a high-definition GPS.

• For Doctors: It means we can eventually find better ways to predict kidney disease that aren't confused by how muscular a patient is.
• For Scientists: It opens the door to new medicines. Since they found that fat metabolism (AGPAT4) and growth factors (IGF2R) are key, doctors might one day treat kidney disease by targeting these specific pathways.
• For Everyone: It highlights that men and women might need different genetic approaches to kidney health, paving the way for more personalized medicine.

In short: By ignoring the "muscle noise," the researchers finally heard the true "kidney signal," discovering new genetic keys that could unlock better treatments for kidney disease in the future.

Technical Summary

1. Problem Statement and Rationale

Chronic Kidney Disease (CKD) is a major global health burden. While Genome-Wide Association Studies (GWAS) have successfully identified hundreds of loci associated with kidney function, they have predominantly relied on creatinine-based estimated Glomerular Filtration Rate (eGFRcrea).

• Limitation of eGFRcrea: This metric is influenced by non-renal factors, specifically muscle mass, which can lead to the over- or underestimation of true kidney function.
• Sexual Dimorphism: Kidney disease exhibits significant sex differences in prevalence and progression, yet the sex-specific genetic architecture of kidney function remains underexplored.
• Objective: The study aims to perform a large-scale GWAS using Creatinine Clearance (CrCl) as the primary phenotype. CrCl, calculated from 24-hour urine creatinine excretion and plasma creatinine, is considered a more accurate, muscle-mass-independent measure of glomerular filtration. The study also seeks to identify sex-specific genetic variants.

2. Methodology

• Study Population: Data were drawn from the Lifelines Cohort Study, a large, multi-disciplinary, three-generation population-based study in the Northern Netherlands.
  • Sample Size: 58,976 individuals of European descent.
  • Exclusions: Participants with missing phenotype data, inadequate 24-hour urine collections (defined by deviation from estimated volume), non-European ancestry, or first-degree relatives were excluded.
• Phenotypes:
  • Primary: Creatinine Clearance (CrCl), normalized to body surface area (BSA).
  • Secondary: eGFRcrea (using the CKD-EPI equation) for comparison with existing literature.
• Genotyping & QC: Genotyping was performed across three stages using different arrays (CytoSNP, GSA, and FinnGen Axiom). Quality control included standard filtering (MAF > 0.01, INFO > 0.4) and imputation against the 1000 Genomes Phase 3 reference panel.
• Statistical Analysis:
  • GWAS: Performed using linear mixed models (SAIGE and REGENIE) to account for relatedness and population stratification (adjusted for age, sex, and 20 principal components).
  • Meta-analysis: Fixed-effects inverse variance-weighting was used to combine results from the three genotyping subcohorts.
  • Sex-Specific Analysis: Stratified GWASs were conducted for males and females. Sex-difference testing was performed using a t-statistic formula with Bonferroni correction ($P < 0.004$).
  • Functional Follow-up:
    • eQTL/pQTL/PheWAS: Lookups in blood (eQTLGen) and kidney databases, Open Targets Genetics, and PheWAS catalogs.
    • Cell Type Prioritization: Used single-cell RNA-seq (scRNA-seq) data from human kidneys and stratified LDSC (CELLECT) to identify enriched cell types.
    • Colocalization & Fine-mapping: Performed using coloc and FINEMAP to assess shared causal variants.

3. Key Contributions

1. First Large-Scale CrCl GWAS: This is the first GWAS of its kind using a muscle-mass-independent phenotype (CrCl) in a large European cohort ($N \approx 59k$).
2. Novel Loci Identification: Discovery of two novel genetic variants associated with kidney function that were not previously identified in eGFR-based studies.
3. Sex-Specific Architecture: Identification of two female-specific loci, highlighting the importance of sex-stratified analyses in understanding kidney function genetics.
4. Phenotype Differentiation: Demonstration that CrCl and eGFRcrea capture distinct genetic signals, particularly regarding creatinine metabolism versus glomerular filtration.

4. Key Results

• Overall Associations:

  • Identified 16 independent loci with 21 genome-wide significant lead SNPs associated with CrCl.
  • SNP-based Heritability: Estimated at 12% for CrCl.
  • Genetic Correlation: High correlation between CrCl and eGFRcrea ($r_g \approx 0.73–0.77$), but moderate with eGFRcys ($r_g \approx 0.48$).

• Novel Loci (Sex-Combined):

  • rs146465192 (Chr 6): Located near RP1-249F5.3. Not significant in previous eGFR GWASs. Associated with plasma IGF2R levels (pQTL) and LDL cholesterol.
  • rs117014836 (Chr 6): Located near AGPAT4. Not significant in previous eGFR GWASs. Associated with blood expression of AGPAT4 (eQTL). AGPAT4 is involved in lipid metabolism (lysophosphatidic acid conversion), linking lipid dysregulation to kidney function.

• Sex-Specific Loci:

  • Female-Specific:
    • rs8002366 (GPC6): A novel locus associated with reduced CrCl in females. Associated with kidney GPC6 expression (eQTL). GPC6 encodes a heparan sulfate proteoglycan involved in the glomerular filtration barrier.
    • rs12908437 (IGF1R): Associated with reduced CrCl in females. Linked to blood IGF1R expression.
  • Male-Specific: No unique male-specific loci were identified; all male-significant SNPs were also significant in females or the combined model.

• Comparison with eGFRcrea:

  • Many known eGFR loci were replicated.
  • Divergence: The strongest eGFRcrea signal in the study (rs1346267, near GATM) was not genome-wide significant for CrCl. This suggests this locus influences creatinine production (metabolism) rather than glomerular filtration, validating the utility of CrCl in filtering out non-renal genetic noise.

• Cell Type Enrichment:

  • Heritability enrichment was observed in proximal tubule cells, connecting tubule cells, and epithelial progenitor cells, consistent with the physiological role of these cells in creatinine handling and filtration.

5. Significance and Implications

• Refining Kidney Function Genetics: By using CrCl, the study successfully disentangled genetic variants affecting true glomerular filtration from those affecting creatinine generation (muscle mass/metabolism). This provides a clearer picture of the genetic architecture of kidney disease.
• Biological Insights: The identification of AGPAT4 and IGF2R links lipid metabolism and IGF signaling pathways directly to kidney function, offering new potential therapeutic targets.
• Sexual Dimorphism: The discovery of female-specific loci (GPC6, IGF1R) underscores that genetic risk factors for kidney function differ by sex, which may explain epidemiological differences in CKD prevalence and progression between men and women.
• Clinical Relevance: The findings suggest that relying solely on eGFRcrea may miss critical genetic determinants of kidney health or misattribute metabolic signals to renal function. Future genetic studies and risk prediction models should consider muscle-mass-independent phenotypes like CrCl.

Limitations: The study is restricted to individuals of European ancestry, limiting generalizability. Additionally, the reliance on 24-hour urine collections limits the sample size compared to serum-based eGFR studies, and the scarcity of such data in other cohorts hinders external replication.