Kidney Week

Abstract: FR-PO641

Predicted Pathogenic Missense Variants in CLDN10 Are Associated with CKD

Session Information

• Genetic Kidney Diseases: Cohort Studies - Genetic Associations and Diagnoses
  October 25, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Genetic Diseases of the Kidneys

• 1202 Genetic Diseases of the Kidneys: Non-Cystic

Authors

• Chang, Alexander R., Geisinger Health, Danville, Pennsylvania, United States

Alexander R. Chang, MD, MS

• Moore, Bryn S., Geisinger Health, Danville, Pennsylvania, United States

Bryn S. Moore, MS

• Luo, Jonathan Z., Geisinger Health, Danville, Pennsylvania, United States

Jonathan Z. Luo

• Mallett, Andrew John, James Cook University, Townsville, Queensland, Australia

Andrew John Mallett, MBBS, PhD, FASN

• Biros, Erik, James Cook University, Townsville, Queensland, Australia

Erik Biros, PhD

• Mirshahi, Tooraj, Geisinger Health, Danville, Pennsylvania, United States

Tooraj Mirshahi, PhD

• Sayer, John Andrew, Newcastle University, Newcastle upon Tyne, Tyne and Wear, United Kingdom

John Andrew Sayer, MBChB, PhD

Background

CLDN10 encodes the tight junction protein claudin-10. and has been linked to a non-Bartter, non-Gitelman autosomal recessive salt-losing tubulopathy. Recently, heterozygous protein truncating variants (PTVs) in CLDN10 have been associated with higher cystatin C, creatinine, urea, and uric acid levels. In this study we examined whether predicted pathogenic missense variants were associated with estimated glomerular filtration rate (eGFR) and other renal phenotypes.

Methods

We used data from the MyCode/DiscovEHR study, a health system-based
cohort (n=174172) in central and northeast PA and the UK Biobank study (n=394,841).
In the MyCode cohort, predicted pathogenic missense variants were determined using Alpha Missense, a deep learning-based prediction tool. In UK Biobank, missense variants/low confidence (LC) loss of function variants were grouped together (www.genebass.org). Renal phenotypes available in UK Biobank included plasma cystatin C, creatinine urea, urate. For MyCode, renal phenotypes examined included creatinine-based estimated glomerular filtration rate (eGFR), eGFR <60, eGFR<30, dipstick proteinuria, and gout. Age and sex-adjusted gene burden analyses
for CLDN10 predicted pathogenic missense variants were examined using linear regression for eGFR and Firth logistic regression for categorical variables.

Results

In MyCode, there were 118 individuals heterozygous for 37 predicted pathogenic CLDN10 missense variants. Individuals heterozygous for predicted pathogenic CLDN10 missense variants had lower eGFR (-7.30 ml/min/1.73m², 95% CI: -10.96, -3.64; p=9.2E-5), and higher risk of eGFR <60 (OR 2.80, 95% CI: 1.75, 4.47; p=1.6E=-05). Associations were not significant for gout (OR 1.90, 95% CI: 0.93, 3.93; p=0.08), eGFR <30 (OR 1.27, 95% CI: 0.56, 2.87; p=0.6), or dipstick proteinuria (OR 1.28, 95% CI: 0.79, 2.07; p=0.3). In UK Biobank rare missense/LC CLDN10 variants were associated with cystatin C (p-value SKATO =8.66E-19), urea (p=1.88E-7), urate (p=6.25E-8), and creatinine (p=1.76E-7).

Conclusion

Predicted pathogenic missense variants in CLDN10 were associated with reduced eGFR and eGFR<60.

Funding

• NIDDK Support