Effects of peritoneal dialysis solutions on the secretion of growth factors and extracellular matrix proteins by human peritoneal mesothelial cells

Abstract

Objective: To compare the effects of different peritoneal dialysis solutions (PDS) on secretion of vascular endothelial growth factor (VEGF), transforming growth factor-beta1 (TGFbeta1), procollagen I C-terminal peptide (PICP), procollagen III N-terminal peptide (PIIINP), and fibronectin by cultured human peritoneal mesothelial cells (HPMC).

Design: Using M199 culture medium as control, commercial PDS containing 1.5% or 4.25% glucose and 40 mmol/L lactate [Dianeal 1.5 (D 1.5) and Dianeal 4.25 (D 4.25), respectively; Baxter Healthcare, Deerfield, Illinois, USA]; PDS containing 1.5% or 4.25% glucose with 25 mmol/L bicarbonate and 15 mmol/L lactate [Physioneal 1.5 (P 1.5) and Physioneal 4.25 (P 4.25), respectively; Baxter]; and PDS containing 7.5% icodextrin [Extraneal (E); Baxter] were tested. Growth-arrested and synchronized HPMC were continuously stimulated for 48 hours by test PDS diluted twofold with M199, TGFbeta1 1 ng/mL, or different concentrations of icodextrin. VEGF, TGFbeta1, and fibronectin secreted into the media were analyzed by ELISA, and PICP and PIIINP by radioimmunoassay.

Results: Dianeal 1.5, D 4.25, and P 4.25, but not P 1.5 and E, significantly increased VEGF secretion compared with control M199. D 4.25- and P 4.25-induced VEGF secretion was significantly higher than induction by D 1.5 and P 1.5, respectively, suggesting that high glucose may be involved in the induction of VEGF. Physioneal 1.5- and P 4.25-induced VEGF secretion was significantly lower than induction by D 1.5 and D 4.25, respectively, suggesting a role for glucose degradation products (GDP) in VEGF production. TGFbeta1 secretion was significantly increased by D 4.25 and E. Icodextrin increased TGFbeta1 secretion in a dose-dependent manner. All PDS tested significantly increased secretion of PIIINP compared with control. D 1.5- and D 4.25-induced PIIINP secretion was significantly higher than P 1.5, P 4.25, and E. Physioneal 4.25-induced PIIINP secretion was significantly higher than P 1.5, again implicating high glucose and GDP in PIIINP secretion by HPMC. There was no significant increase in PICP or fibronectin secretion using any of the PDS tested. Addition of TGFbeta1 1 ng/mL into M199 control significantly increased VEGF, PICP, PIIINP, and fibronectin secretion by HPMC.

Conclusions: The present study provides direct evidence that HPMC can secrete VEGF, TGFbeta1, and PIIINP in response to PDS, and that HPMC may be actively involved in the development and progression of the peritoneal membrane hyperpermeability and fibrosis observed in long-term PD patients. This study also suggests that both high glucose and GDP in PDS may play important roles in inducing VEGF and PIIINP production/secretion by HPMC.