Efficient in vitro lowering of carbonyl stress by the glyoxalase system in conventional glucose peritoneal dialysis fluid

Background. Reactive carbonyl compounds (RCOs) present in heat-sterilized peritoneal dialysis (PD) fluid have been incriminated in the progressive deterioration of the peritoneal membrane observed in long-term PD patients. The present study utilized the glyoxalase I (GLO I) system as a new approach to lower in vitro the peritoneal fluid content of RCOs such as methylglyoxal (MGO), glyoxal (GO) and 3-deoxyglucosone (3-DG). Methods. GO, MGO, and 3-DG solutions or conventional glucose PD fluids were incubated in vitro with various RCO lowering compounds. The evolution of GO, MGO, and 3-DG levels was monitored by high-performance liquid chromatography. The tested compounds included aminoguanidine and glutathione (GSH), alone or together with GLO I. The human GLO I gene was overexpressed in Chinese hamster ovary (CHO) cells, or ubiquitously in transgenic mice. Cell supernatant of the CHO transfectant and protein extracts of various organs of the transgenic mice were also tested. Results. Aminoguanidine incubated with MGO/GO/3-DG mixtures, promptly reduced RCO levels. GSH alone had a similar but milder and slower effect. Together with GLO I, it promptly decreased GO and MGO levels but was less efficient toward 3-DG. After incubation with glucose PD fluid, GSH together with GLO I had the same effect on MGO, GO, and 3-DG levels. Addition of transfected cell supernatant or tissue extracts overexpressing GLO I, together with GSH to either GO, MGO, or 3-DG solutions, promptly and markedly reduced GO and MGO but not 3-DG levels. Conclusions. GLO I together with GSH efficiently lowers glucose-derived RCOs, especially GO and MGO, both in conventional glucose PD fluids and in RCO solutions. The fact that genetically manipulated cells overexpressing GLO I activity have a similar effect suggests that maneuvers raising GLO I activity in peritoneal cells or in the peritoneal cavity might help prevent the deleterious effects of the peritoneal carbonyl stress in PD patients. The clinical relevance of this approach is yet to be documented.