Urinary Proteomics Identifies Cathepsin D as a Biomarker of Rapid eGFR Decline in Type 1 Diabetes

Abstract

Objective: Understanding mechanisms underlying rapid estimated glomerular filtration rate (eGFR) decline is important to predict and treat kidney disease in type 1 diabetes (T1D).

Research design and methods: We performed a case-control study nested within four T1D cohorts to identify urinary proteins associated with rapid eGFR decline. Case and control subjects were categorized based on eGFR decline ≥3 and <1 mL/min/1.73 m2/year, respectively. We used targeted liquid chromatography-tandem mass spectrometry to measure 38 peptides from 20 proteins implicated in diabetic kidney disease. Significant proteins were investigated in complementary human cohorts and in mouse proximal tubular epithelial cell cultures.

Results: The cohort study included 1,270 participants followed a median 8 years. In the discovery set, only cathepsin D peptide and protein were significant on full adjustment for clinical and laboratory variables. In the validation set, associations of cathepsin D with eGFR decline were replicated in minimally adjusted models but lost significance with adjustment for albuminuria. In a meta-analysis with combination of discovery and validation sets, the odds ratio for the association of cathepsin D with rapid eGFR decline was 1.29 per SD (95% CI 1.07-1.55). In complementary human cohorts, urine cathepsin D was associated with tubulointerstitial injury and tubulointerstitial cathepsin D expression was associated with increased cortical interstitial fractional volume. In mouse proximal tubular epithelial cell cultures, advanced glycation end product-BSA increased cathepsin D activity and inflammatory and tubular injury markers, which were further increased with cathepsin D siRNA.

Conclusions: Urine cathepsin D is associated with rapid eGFR decline in T1D and reflects kidney tubulointerstitial injury.

Figure 1

Associations of urine peptides (A) and urine proteins (B) with rapid eGFR decline in the JDRF Biomarkers Consortium discovery set. Results are from the full covariate model with adjustment for baseline age, sex, urine creatinine, diabetes duration, antihypertensive medication use, systolic blood pressure, hemoglobin A_1c, UACR, and eGFR. We calculated scaled protein PAR values by averaging scaled peptide PAR values corresponding to the same protein. FDR, false discovery rate; OR, odds ratio.

Figure 2

Meta-analysis of the association between cathepsin D and rapid eGFR decline in the JDRF Biomarkers Consortium for fully adjusted models. OR, odds ratio; TE, treatment effect; se, standard error.

Figure 3

Cathepsin D mRNA expression in tubules (A) but not glomeruli (B) is significantly correlated with VvINT in the Pima Indian cohort. mRNA expression levels are log₂ transformed.

Figure 4

Expression of CTSD mRNA in proximal tubular cell clusters of LDs and patients with DKD. A: Dot plots showing the expression of CTSD mRNA in proximal tubular cell clusters from living donor (blue) and DKD (pink) kidney. Color intensity indicates expression level, and the size of the dot indicates the percentage of cells expressing CTSD. B: Average expression levels of CTSD from proximal tubular cell clusters of LDs and patients with DKD. The P values shown in the figure are derived from the nonparametric Wilcoxon rank sum tests in the Seurat package. DTL, descending thin limb; LD, living donor; PT, proximal tubule.

Figure 5

Mouse proximal tubular epithelial cell studies. A: Increased intensity of cathepsin D and Lamp1 (lysosomal membrane marker) immunofluorescent staining is seen with 100 μg/mL and 200 μg/mL AGE-BSA compared with 5 mmol/L glucose, 25 mmol/L glucose, 100 μg/mL ctr-BSA, and 200 μg/mL ctr-BSA. Magnification ×40. B: Cathepsin D enzyme activity (measured in RFU) is increased with exposure to AGE-BSA compared with ctr-BSA. C and E: In the setting of 100 μg/mL AGE-BSA exposure, IL-1 and KIM-1 concentrations are significantly elevated in the media of MCT transfected with cathepsin D small interfering RNA (CatD-siRNA) compared with cells transfected with control small interfering RNA (ctr-siRNA). IL-1 was additionally significantly elevated with CatD-siRNA compared with ctr-siRNA in the setting of 100 μg/mL ctr-BSA exposure. D: The concentration of IL-6 is increased in the media of cells (AGE-BSA-MCT) transfected with CatD-siRNA vs. ctr-siRNA, though this difference does not meet statistical significance. Concentrations of 5.5 mmol/L glucose, 100 μg/mL AGE-BSA, and 100 μg/mL ctr-BSA were used unless otherwise indicated. **P < 0.01; ***P < 0.001; ****P < 0.0001.