Glyoxal in hyperglycaemic ischemic stroke - a cohort study

Abstract

Background: Hyperglycaemia is frequent in acute ischemic stroke and denotes a bad prognosis, even in the absence of pre-existing diabetes. However, in clinical trials treatment of elevated glucose levels with insulin did not improve stroke outcome, suggesting that collateral effects rather than hyperglycaemia itself aggravate ischemic brain damage. As reactive glucose metabolites, glyoxal and methylglyoxal are candidates for mediating the deleterious effects of hyperglycaemia in acute stroke.

Methods: In 135 patients with acute stroke, we used liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) to measure glyoxal, methylglyoxal and several of their glycated amino acid derivatives in serum. Results were verified in a second cohort of 61 stroke patients. The association of serum concentrations with standard stroke outcome scales (NIHSS, mRS) was tested.

Results: Glucose, glyoxal, methylglyoxal, and the glyoxal-derived glycated amino acid N_δ-(5-hydro-4-imidazolon-2-yl)ornithine (G-H1) were positively correlated with a bad stroke outcome at 3 months as measured by mRS90, at least in one of the two cohorts. However, the glycated amino acids N_ε-carboxyethyllysine (CEL) and in one cohort pyrraline showed an inverse correlation with stroke outcome probably reflecting lower food intake in severe stroke. Patients with a poor outcome had higher serum concentrations of glyoxal and methylglyoxal.

Conclusions: The glucose-derived α-dicarbonyl glyoxal and glycated amino acids arising from a reaction with glyoxal are associated with a poor outcome in ischemic stroke. Thus, lowering α-dicarbonyls or counteracting their action could be a therapeutic strategy for hyperglycaemic stroke.

Keywords: Advanced glycation end-products; Glucose; Ischemic brain damage; Undernutrition.

Fig. 1

Schematic illustration of the connection of glucose metabolism, α-dicarbonyl production, and formation of related glycated amino acids [60]. While glyoxal and pyrraline can be generated directly from glucose by autooxidation, glycolysis is necessary to provide the methylglyoxal precursor molecules glyceraldehyde-3-phosphate and dihydroxyacetone phosphate. α-Dicarbonyls modify arginine or lysine residues resulting in the formation of G-H1 and CML or MG-H1, CEL and argpyrimidine through the reaction with glyoxal and methylglyoxal, respectively. Furthermore, glycated amino acids can be absorbed after digestion of food

Fig. 2

P-value heatmap of correlations and group differences of glucose, α-dicarbonyls, and glycated amino acids with diabetes-related parameters and neurological scores. Blood samples of four consecutive days after stroke have been analysed using LC-MS. (A) In patients of cohort 1, concentrations of the compounds were tested for an association with the history of diabetes or laboratory parameters of hyperglycaemia and diabetes (glucose, HbA1c, Spearman coefficient; diabetes mellitus, Mann Whitney U-test). (B) Correlations with neurological scores on admission (mRS, NIHSS) and later stages (mRS day 90, NIHSS discharge) in cohort 1 (Spearman coefficient). (C) Results were confirmed in cohort 2. mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale. White squares indicate no correlation, grey squares indicate parameters that were not measured (day 2–4, cohort 2) or for which a correlation with themselves (glucose) was meaningless

Fig. 3

Timeline of serum levels of patients in cohort 1 with a good or bad outcome. Concentrations of glucose, glyoxal and methylglyoxal were compared between good (mRS90_0 − 2) and bad outcome groups (mRS90_3 − 6). Mean ± SEM. *, p < 0.05, **, p < 0.01 (Mann Whitney U test)