Introduction and hypothesis: Peritoneal dialysis (PD) is an important option for renal replacement therapy. Peritoneal sclerosis (PS) limits PD duration due to loss of ultrafiltration (UF) capacity, while about 3% of PD patients experience a condition termed encapsulating peritoneal sclerosis (EPS). In many fibrotic diseases reduced Extracellular matrix (ECM) breakdown due to lowered matrix metalloproteinase (MMP) activity occurs, often from over-expression of tissue inhibitors of MMP (TIMPs) that underlie fibrotic remodeling. Furthermore, recent application of 2D gel proteomics on peritoneal dialysis effluent (PDE) samples has identified several proteins that are elevated in patients with membrane damage. These observations have led to the hypothesis that: changes in proteins in PDE samples, in particular those associated with ECM breakdown have value as non-invasive biomarkers of PS and the switch to EPS. To test this hypothesis, PDE samples from 3 patient cohorts was analysed for ECM proteolytic activity. A range of ECM processing proteins and 3 proteins identified from previous proteomic studies of patients developing EPS (intellectin-1, dermatopontin and collagen α1 (I)) were analysed in PDE samples.
Methods: Three patient cohorts were studied: two were from Sheffield Kidney Institute (SKI) that consisted of 32 spot PDE samples (SKI-1) that included 1 EPS patient & 51 PDE & plasma samples collected during a peritoneal equilibrium test (PET) with multiple dwell times in patients who did not have EPS (SKI-2). The third cohort consisted of 209 samples from the Global Fluid Study (GFS) including sequential samples from 12 EPS & 42 matched controls patients. MMP activity was assessed using the ENZchek assay system. Plasmin activity was assessed by using cleavage of the V0882 substrate. TIMPs, MMPs, intelectin-1, dermatopontin, and collagen (α1) I were quantified by commercial ELISA in PDE and plasma samples. PDE cytology (macrophages, leukocytes, fibroblasts and mesothelial cells) was performed to determine if changes in any protein could be associated with changes in cell types. Clinical data were recovered from either the peritoneal dialysis database (PDDB) at Sheffield or the GFS archives. The analysis was performed using Microsoft Excel 2010 software, SPSS, and Graphpad prism (prism 5.01 for windows).
Results: Plasmin activity in PDE samples decreases with long duration of PD therapy. Minimal MMP activity was found in all PDE samples. In the SKI-1 cohort, MMP-1, -9, & -13 were almost undetectable with only MMP-2 & -3 being measurable with levels of ((mean±SD) 46±37 & 2.1±2.2 ng/mL respectively). In contrast TIMP-1 and TIMP-2 and to lesser extent TIMP-3 had significant levels in PDE samples from commencing PD (109±88, 17±12, and 0.28±0.33 ng/mL respectively). All TIMPs & MMP-2 were raised in the single patient who had a diagnosis of EPS. In samples from the GFS cohort, there was a rapid 6 fold increases in TIMP-1 within 100 days of the diagnosis of EPS, which when normalised to TIMP-2 levels was a good predictor of EPS. Calculation of the plasma to dialysate transfer rate by reference to that of circulating proteins with no peritoneal production and of known molecular weight (albumin, beta2microglobulin (B2M), transferrin, IgG, and creatinine) demonstrated that TIMPs & MMPs (especially TIMP-1 and MMP-2) have significant peritoneal production. Plasma levels for TIMP-1,-2, MMP-2,-3, and intelectin-1 (mean±SD) were 121±27, 85±16, 176±35, 11±5, and 374±136 ng/mL in healthy individuals respectively. Plasma levels in PD patients for TIMP-1,-2, MMP-2,-3, and intelectin-1 (mean±SD) were 297±78, 158±33, 309±112, 42±28, and 749 ±722 ng/mL respectively. None of the proteins identified by proteomics as predictors of EPS were able to be validated by ELISA. However TIMP-1,-2, MMP-2, intelectin-1, and collagen (α1) I in PDE samples had significant correlations with the loss of ultrafiltration and thus membrane damage. PDE cytology showed that peritoneal fibroblast and leukocyte numbers increase with time on PD, while peritoneal macrophage decreases with time on PD. There were no significant changes in mesothelial cells.
Conclusions: Negligible MMP activity in PDE samples results from high TIMP levels which could underlie the development of PS. The rapid increase in TIMP-1 within 100 days of EPS development offers value as a diagnostic tool or a late biomarker. Plasma levels of TIMP-1,2, MMP-2,3, and intelectin-1 are higher in patients on PD compare to healthy individuals. The increase in peritoneal fibroblasts may be a source of TIMP-1.
