Endogenous adenine mediates kidney injury in diabetic models and predicts diabetic kidney disease in patients

Abstract

Diabetic kidney disease (DKD) can lead to end-stage kidney disease (ESKD) and mortality; however, few mechanistic biomarkers are available for high-risk patients, especially those without macroalbuminuria. Urine from participants with diabetes from the Chronic Renal Insufficiency Cohort (CRIC) study, the Singapore Study of Macro-angiopathy and Micro-vascular Reactivity in Type 2 Diabetes (SMART2D), and the American Indian Study determined whether urine adenine/creatinine ratio (UAdCR) could be a mechanistic biomarker for ESKD. ESKD and mortality were associated with the highest UAdCR tertile in the CRIC study and SMART2D. ESKD was associated with the highest UAdCR tertile in patients without macroalbuminuria in the CRIC study, SMART2D, and the American Indian study. Empagliflozin lowered UAdCR in nonmacroalbuminuric participants. Spatial metabolomics localized adenine to kidney pathology, and single-cell transcriptomics identified ribonucleoprotein biogenesis as a top pathway in proximal tubules of patients without macroalbuminuria, implicating mTOR. Adenine stimulated matrix in tubular cells via mTOR and stimulated mTOR in mouse kidneys. A specific inhibitor of adenine production was found to reduce kidney hypertrophy and kidney injury in diabetic mice. We propose that endogenous adenine may be a causative factor in DKD.

Keywords: Chronic kidney disease; Diabetes; Fibrosis; Nephrology.

Figure 1. High urine adenine/creatinine levels identify patients with diabetes who are at high risk of end-stage kidney disease and mortality.

(A) Participants with diabetes in the CRIC cohort (n = 904) had urine adenine/creatinine ratios (UAdCRs) measured within 1 year of enrollment and followed for 10 years. The participants in the top tertile had the highest risk of end-stage kidney disease (ESKD) and all-cause mortality. (B) Participants from the SMART2D study (n = 309) had UAdCR measurements at the time of enrollment and were followed for 7 years. The participants in the top tertile for UAdCR had the highest risk for ESKD and all-cause mortality. A log-rank test was used to compare cumulative incidence curves in A and B. A P value of less than 0.05 was considered significant.

Figure 2. High urine adenine/creatinine tertile identifies end-stage kidney disease outcome in nonmacroalbuminuric patients with diabetes and empagliflozin-reduced urine adenine/creatinine ratio.

(A and B) The participants with the top urine adenine/creatinine ratio (UAdCR) tertile had a significant increase in risk of end-stage kidney disease (ESKD) from the (A) CRIC (n = 551) and (B) SMART2D (n = 309) studies. (C) Patients with T1 diabetes underwent treatment with empagliflozin for 8 weeks, which reduced UAdCR levels (n = 40 patients). A log-rank test was used to compare cumulative incidence curves in A and B, and a 2-sample t test was used to compare UAdCR levels in C. A P value of less than 0.05 was considered significant.

Figure 3. Spatial metabolomics identifies adenine in regions of pathology in nonmacroalbuminuric patients with diabetes.

(A) Adenine was localized to regions of normal glomeruli and vessels in the normal kidney. Yellow circles and red circles indicate the region of interest labeled on the AF image and adenine ion image, respectively. AF, autofluorescence. Scale bar: 200 μm (AF, adenine, PAS + adenine, and adenine); 50 μm (PAS). (B) In a diabetic kidney, adenine is diffusely increased across the tissue section and prominent in regions of sclerotic blood vessels, glomeruli with mild sclerosis, and regions of atrophic tubules and interstitial inflammation. Scale bar: 200 μm (AF, adenine, PAS + adenine, and adenine); 50 μm (PAS). (C) Quantitative assessment across healthy controls (n = 5 from the CROCODILE study) and diabetic samples (n = 8 T1D from CROCODILE and n = 8 T2D, 2 from CROCODILE and 6 from Kidney KPMP) demonstrates a statistically significant increase of adenine in kidney tissue sections. Two-tailed Student’s t test was used for the comparison. Data represent mean ± SEM.

Figure 4. Molecular pathways and events implicating the ribonucleoprotein biogenesis and mTOR pathways with adenine in DKD.

(A and B) The protein synthesis (ribonucleoprotein [RNP] biogenesis) pathway increased in proximal tubule cells of patients with DKD without proteinuria. Single-cell- transcriptomic data obtained from DKD kidney biopsies from the KPMP study were analyzed for differentially expressed genes in proximal tubules (PTs) of each DKD patient versus healthy reference tissue. Upregulated genes with an adjusted P ≤ 0.01 and ranked among the top 600 significant differentially expressed genes were subjected to pathway enrichment analysis using the Molecular Biology of the Cell Ontology (MBCO). Ranking for the RNP biogenesis pathway (a level-2 pathway canonically regulated by the mTOR pathway) is shown for 28 individual patients. Vertical dashed lines indicate P ≤ 0.01 for pathway ranking. (C) Up to the top 5, 5, 10, and 5 level-1 (dark red), level-2 (red), level-3 (blue), and level-4 (green) pathways, respectively, using MBCO are shown for patient 1 (P ≤ 0.01). See blue lines in A and B. Single-cell-transcriptomic data from patients with T2D (n = 6) with low albuminuria compared with cohort specific healthy samples was analyzed to identify upregulated pathways in PT cells. Note that the RNP biogenesis pathway is the top ranked level-2 pathway in both independent studies. (D–F) Cell culture studies in mouse proximal tubular cells demonstrated an increase in (D) fibronectin and (E) phospho-S6 kinase and (F) that mediation of fibronectin (FN) upregulation is blocked by rapamycin, indicating that mTOR mediates adenine effect. (G–K) Adenine administration to mice increases (G) serum soluble tumor necrosis factor-1 (sTNFR1) and (H) plasma kidney injury marker-1 (KIM-1) and (I) stimulates kidney and (J and K) matrix molecules in the kidney (n = 12 in control group and n = 7 in adenine treated group). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, 2-tailed Student’s t test was used for 2 group comparisons. One-way ANOVA was used for multiple group comparisons. Data represent mean ± SD.

Figure 5. Methylthioadenosine phosphorylase inhibitor ameliorates kidney injury in db/db mice with type 2 diabetes.

(A–C) Methylthio-DADMe-Immucillin-A (MTDIA) significantly reduced kidney (A) adenine levels, (B) kidney hypertrophy, (C) kidney KIM-1 levels, and (D) urine KIM-1 levels in diabetic mice. (E–G) MTDIA significantly reduced (E) diabetes-increased serum cystatin C and (F) partially reduced plasma creatinine and (G) albuminuria in diabetic mice. (H and I) Diabetes-induced kidney matrix protein levels were partially reduced by MTDIA. (J) Ribosomal S6 phosphorylation was partially reduced by MTDIA in the kidneys of db/db mice (n = 6 per group). *P < 0.05, **P < 0.01, ***P < 0.001, 2-tailed Student’s t test was used for 2 group comparisons. Data represent mean ± SD.