Characterisation of glyoxalase 1 mutant mouse and glyoxalase 1 copy number alteration

Glyoxalase 1 (Glo1) of the glyoxalase system catalyses the metabolism of the reactive dicarbonyl metabolite, methylglyoxal, and thereby prevents potentially damaging glycation of protein and DNA. Glo1 is hypothesised to be a potential factor in the development of vascular complication of diabetes, such as diabetic nephropathy. The induction of diabetes in mice deficient in Glo1 provides a pre-clinical in vivo model to test this hypothesis. Glo1 mutant mice with putative Glo1 deficiency produced by the International Mouse Knockout Consortium (IMKC) were acquired from the European Mutant Mouse Archive. The initial aim of this study was to study the exacerbation of diabetic nephropathy by Glo1 deficiency in streptozotocin-induced diabetic mice, with an initial objective to confirm Glo1 deficiency in the IMKC Glo1 mutant mouse and subsequent objectives contingent on this. The preliminary studies were unable to confirm Glo1 deficiency in this mouse model and so a revised aim was to characterise the mechanism of compensatory Glo1 expression in the mutant mouse and explore similar occurrence in similar precursor mouse embryonic stem cells (ESCs) and related clinical application. Genotyping of Glo1 mutant mouse offspring by PCR revealed only heterozygotes and wild-type (WT) littermates, and no homozygotes without Glo1 wild-type alleles. Studies of the Glo1 mutant mouse revealed levels of Glo1 activity, protein and mRNA identical to those of wild-type control siblings. Other components of the glyoxalase system were also analysed – activity of glyoxalase 2, concentrations of methylglyoxal (MG) and D-lactate, and tissue protein content and urinary excretion of MG-derived glycation adduct MG-H1 and found no significant change in Glo1 mutant mice, with respect to WT controls. This suggested a functionally normal Glo1 and glyoxalase system in Glo1 mutant mice. Therefore, Glo1 mutant mice have a mutated Glo1 gene but with compensatory Glo1 expression identical to that of WT control. This provided a possible explanation for the unexpected normal phenotype of Glo1 mutant mice reported in the IMKC project. To explore the mechanism of compensatory Glo1 expression, Glo1 copy number was quantified by Taqman® method, normalizing response to transferrin receptor protein-1 (Tfrc). Glo1 mutant mice had 3 copies of Glo1 in all tissues analysed with amplification extending from 3’-end of exon 1 to the 5’-end of exon 6. Taqman copy number assay was established to detect and quantify mutant Glo1Gt(..)Lex and WT alleles. Most mutant mice contained two copies of Glo1 and one mutant copy of Glo1Gt(..)Lex – Glo1(+/+)Gt(..)1Lex. In some cases, however, 2 copies of both Glo1 and mutated Glo1Gt(..)Lex – Glo1(+/+)Gt(..)2Lex were found. Inheritance studies suggested a simple Mendelian inheritance with a WT allele accompanying the Glo1Gt(..)Lex mutant allele on arms of chromosome 17 such that Glo1 deficiency was prevented. This was indeed observed throughout the all breeding of the Glo1 mutant mice. I hypothesised that Glo1 copy number increase may have arisen in the mutant mice during gene trapping by copy number alteration (CNA) induced by increased methylglyoxal concentration, or dicarbonyl stress, in mouse ESCs. To explore and model this, mouse ESCs were cultured with exogenous 200 μM MG under atmospheres containing 20% oxygen - typical of most cell culture conditions, and 3% oxygen - typical of ESCs oxygen exposure in vivo. Incubation of ESCs for 12 days with MG induced CNV increase of Glo1 by up to 16% in both 20% and 3% oxygen atmospheres. Increase in Glo1 CNV at day 12 with MG treatment was associated with an increase in Glo1 protein. Therefore, functional low level CNA of Glo1 was induced by exposure to high levels of exogenous MG. No evidence was found for Glo1 CNA with dicarbonyl stress induced by Glo1 silencing or cell permeable Glo1 inhibitor. Finally, I hypothesised that GLO1 CNA may occur in clinical dicarbonyl stress, a severe example of which is patients with renal failure receiving haemodialysis - associated with ca. 5-fold increase in plasma MG concentration. DNA of peripheral mononuclear cells from healthy subjects and patients with renal failure receiving hemodialysis renal replacement therapy were examined. Human GLO1 copy number was not significantly different between the patients and the control subjects. This requires further investigation in this case and other examples of clinical dicarbonyl stress. From these studies I conclude that the IMKC Glo1 mutant mouse does not exhibit the Glo1 deficiency; rather, it maintains wild-type levels of Glo1 expression through Glo1 copy increase likely induced during gene trapping. Dicarbonyl stress in mouse ESCs in vitro induced low level Glo1 copy number increase – a model of Glo1 CNA in putative gene trapping associated dicarbonyl stress. It is unclear if GLO1 CNA occurs clinically. These findings reveal that focussed copy number alternation of GLO1 may provide a protective response to dicarbonyl stress in some circumstances