Extracellular matrix associated proteins and processing enzymes as urinary biomarkers of Chronic Kidney Disease

Background: Chronic Kidney Disease (CKD) is often progressive leading to End Stage Renal Disease (ESRD). Early detection of progressive CKD would almost certainly improve long term outcome. Current kidney function tests have serious limitations as they only can detect dysfunction after significant damage has occurred, provide limited help with diagnosis and are unhelpful in assessing early response to therapy or long term prognosis. Therefore there is a requirement for identification of early and reliable non invasive biomarkers of kidney disease and its progression. Regardless of the underlying disease, fibrosis drives movement to ESRD. Fibrosis is characterised by excessive accumulation of Extracellular Matrix (ECM). This thesis tests the hypothesis that mediators of the kidney fibrotic process are present in urine and would make good biomarkers of CKD and its progression. Aims: To assess urine samples from three animal models of kidney fibrosis (5/6th Subtotalnephrectomy (SNx), Diabetic Nephropathy (DN) and Chronic Allograph Nephropathy (CAN) as well as 325 human urine samples (292 CKD and 33 controls) to determine if urinary ECM molecules, Transglutaminase 2 (TG2), ε (γ-glutamyl) lysine, Hydroxyproline, Metalloproteinases 1, 2 and 9, TIMPs 1, 2 and 3, TIMP-1/MMP-1 complex, PAI-1 and MMP activity have potential to act as an early biomarker of kidney scarring and if any of them could predict the rate of progression better than Albuminuria. Experimental methods: Cation exchange chromatography was used to evaluate urine levels of ε (γ-glutamyl) lysine and Hydroxyproline. An in-house developed sandwich ELISA was used to analyse urine and blood levels of TG2 and commercially available immunoassays were used for determination of urine concentration of MMPs 1, 2 and 9, TIMPs 1, 2 and 3, TIMP-1/MMP-1 complex and PAI-1. MMP activity was measured by the EnzCheck Collagenase assay. One way ANOVA with Bonferroni correction was applied to estimate differences between groups. Area under the curve for Receiver Operating Characteristic (ROC) analysis was used to determine prediction accuracy. p < 0.05 was considered statistically significant. Results: TG2 and ε (γ-glutamyl) lysine excretion were significantly increased as renal scarring developed in SNx and CAN models. However, in DN, increased levels were only observed as a 24 hour excretion at 8 months of study only due to the increased proteinuric state. TG inhibitor treated animals had a reduction in the measured levels of both TG2 and of ε (γ-glutamyl) lysine in the SNx study. Urine Hydroxyproline had increased levels as renal disease scarring progressed. MMP-1 was significantly increased in SNx and DN animals, with MMP-9 excretion undetectable in all control samples. Amongst the TIMPs, TIMP-1 gave the most promising response in the early stages of SNx and DN with a 6.4 fold increase at 7 days post SNx, p-0.0025 and 25 fold increase at 4 months post STZ injection, p=0.0006. Whereas in CAN, the TIMP2: CR ratio was double that in F-L Allografts than in L-L Isografts (p=0.0095) as early as 8 weeks post-transplant and remained elevated up to 33 weeks. Urine PAI-1 excretion was within the lower limits of detection and therefore was not used in human samples. In the human study, ROC curve analysis demonstrated that the TG2/CR ratio (86.4%) and TIMP1/CR ratio (75.7%) are better predictors of patients with progressive CKD than ACR (73.5%) indicating their potential as non-invasive biomarkers of progressive kidney scarring. The ROC curve analysis of ε (γ-glutamyl) lysine / CR ratio (73.3%) and TIMP2/ CR ratio (71.2%) presented similar levels of accuracy of prediction as ACR, whereas all other candidates were inferior to ACR. Measurement of MMP-1 activity had a remarkable predictive value in detecting rapid progressive patients, as very low levels were found in this subgroup of patients. A combo ROC curve analysis using TG2CR, XLCR, TIMP1CR and ACR as prognostic tools of CKD progression generated a remarkable 87.7% risk prediction. Conclusions: The data presented in this thesis demonstrates a potential benefit of using ECM molecules (especially TG2, TIMP-1 and ε (γ-glutamyl) lysine) detected in urine as biomarkers of CKD progression. Using a combined biomarker panel containing ACR, TG2/CR, TIMP/CR and XL/CR ratios may allow the earlier and more reliable detection of patients with more aggressive CKD that requires stronger treatment strategies to give better long term outcomes and facilitate shorter clinical trials in CKD.