Expression of ADAM17 and serum PTH levels in dialysis patients with and without diabetes: Potential implications for...

The European COSMOS study, which included 6,797 haemodialysis patients from 20 countries, demonstrated that elevated PTH levels were associated with a higher risk of mortality in diabetic patients compared to non-diabetic ones. Elevated PTH levels have been linked to inflammatory processes. This study aimed to analyse, in epigastric arteries of dialysis patients, whether diabetes and elevated PTH levels might synergistically modulate the expression of inflammatory and vascular calcification (VC) markers through ADAM17 and/or its inhibitor, TIMP3.

Epigastric artery samples were examined from 62 chronic kidney disease patients, 31 non-diabetic and 31 diabetic with similar characteristics, obtained during kidney transplantation. The gene expression of ADAM17, TIMP3, inflammatory markers (TNFα, VCAM, and ICAM), α-actin, DKK1, and RUNX2, as well as protein expression of RUNX2 and vascular calcification, were evaluated. Analyses were performed based on the presence/absence of diabetes and serum PTH levels (using the median value as the cut-off). Differences between groups were assessed using Chi-squared tests for qualitative variables or Wilcoxon tests for quantitative variables.

In the epigastric arteries of diabetic patients, there was lower gene expression of TIMP3 (1.0 (0.8) U.R. vs 0.6 (0.5) U.R.; p = 0.028) and α-actin (1.0 (0.6) U.R. vs 0.7 (0.5) U.R.; p = 0.006), higher gene expression of VCAM (1.0 (0.9) U.R. vs 2.1 (2.8) U.R.; p = 0.011), TNFα (1.0 (1.1) U.R. vs 2.0 (1.9) U.R.; p = 0.008), and RUNX2 (1.0 (0.8) U.R. vs 2.4 (2.8) U.R.; p = 0.009), and a higher percentage of RUNX2-positive nuclei (14.9 (15.8) vs 31.1 (27.1); p = 0.038) compared to non-diabetic patients. Arteries from patients with PTH levels above the median exhibited higher gene expression of ADAM17 (1.0 (0.6) U.R. vs 1.8 (1.6) U.R.; p = 0.026). The increase in ADAM17 expression persisted when only diabetic patients were analysed (1.0 (0.5) U.R. vs 2.9 (2.9) U.R.; p = 0.038), but not in non-diabetic patients. The combination of elevated PTH and diabetes was associated with higher gene expression of RUNX2 (1.0 (0.8) U.R. vs 2.7 (3.3) U.R.; p = 0.019), a higher percentage of RUNX2-positive nuclei (13.3 (17.0) vs 32.9 (27.2); p = 0.039), and a higher Kauppila index (3.9 (5.0) vs 9.7 (7.3); p = 0.045).

Although changes in ADAM17 and TIMP3 gene expression were observed in diabetic patients, potentially related to increased synthesis and/or release of pro-inflammatory factors, these do not appear to explain the difference in mortality associated with elevated PTH levels between diabetic and non-diabetic patients observed in the COSMOS study. However, the increased mortality in diabetic patients with elevated PTH may be linked to greater VC progression mediated by RUNX2.