Histidine containing dipeptides protect epithelial and endothelial cell barriers from methylglyoxal induced injury

Abstract

Integrity of epithelial and endothelial cell barriers is of critical importance for health, barrier disruption is a hallmark of numerous diseases, of which many are driven by carbonyl stressors such as methylglyoxal (MG). Carnosine and anserine exert some MG-quenching activity, but the impact of these and of other histidine containing dipeptides on cell barrier integrity has not been explored in detail. In human proximal tubular (HK-2) and umbilical vein endothelial (HUVEC) cells, exposure to 200 µM MG decreased transepithelial resistance (TER), i.e. increased ionic permeability and permeability for 4-, 10- and 70-kDa dextran, membrane zonula occludens (ZO-1) abundance was reduced, methylglyoxal 5-hydro-5-methylimidazolones (MG-H1) formation was increased. Carnosine, balenine (ß-ala-1methyl-histidine) and anserine (ß-ala-3-methyl-histidine) ameliorated MG-induced reduction of TER in both cell types. Incubation with histidine, 1-/3-methylhistidine, but not with ß-alanine alone, restored TER, although to a lower extent than the corresponding dipeptides. Carnosine and anserine normalized transport and membrane ZO-1 abundance. Aminoguanidine, a well-described MG-quencher, did not mitigate MG-induced loss of TER. Our results show that the effects of the dipeptides on epithelial and endothelial resistance and junction function depend on the methylation status of histidine and are not exclusively explained by their quenching activity.

Keywords: Barrier integrity; Carbonyl stress; Dipeptides; Ionic permeability; Transepithelial resistance; Zonula-occludens.

Fig. 1

Transepithelial and transendothelial resistance (TER). TER decreased dose-dependently upon exposure to methylglyoxal (MG, 200 µM) in human proximal tubulus cell (HK2, a) and human umbilical vein endothelial cells (HUVEC, c). The decrease of TER was prevented by co-incubation with carnosine (Carn) and anserine (Ans, both 70 mM) in both cell types (b,d). One-way ANOVA with Tukey´s test and correction for multiple comparison. ****p < 0.0001.

Fig. 2

Paracellular transport capacity for dextrans of different molecular sizes in epithelial and endothelial cells. Paracellular transport of 4 kDa (a), 10 kDa (b) and 70 kDa (c) fluorescein labelled dextrans across human proximal tubulus (HK2) and human umbilical vein endothelial (HUVEC) monolayers increased after 5 h of exposure to methylglyoxal (MG, 200 µM). This increase could be partly prevented by co-incubation with carnosine (Carn), anserine (Ans, both 70 mM) treated cells were comparable to medium treated controls. Two-way ANOVA followed by Dunnett’s test corrected for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 3

Zonula occludens (ZO-1) distribution in cell membranes of epithelial and endothelial cells. Methylglyoxal (MG, 200 µM) induces cell type specific ZO-1 cell membrane distribution. Representative immunostaining of ZO-1 showing the pattern at cellular membrane areas followed by quantification of membrane ZO-1 in polarized HK 2 (a,b) and HUVEC (c,d) grown on transwell filters upon exposure to MG and treatment with carnosine (Carn) and anserine (Ans, both 70 mM). MG resulted in discontinuous ZO-1 membrane distribution, this effect was prevented by Ans treatment. One-way ANOVA with Tukey´s test corrected for multiple comparisons. ***p < 0.001, ****p < 0.0001.

Fig. 4

Formation and quenching of the methylglyoxal 5-hydro-5-methylimidazolone (MG-H1). Methylglyoxal (MG) induces formation of the MG-H1 after 5 h of exposure in HK2 cells. Anserin (Ans) and carnosine (Carn, both 70 mM) quench MG-H1 in HK2. Representative western blot image and quantification of MG-H1 relative to ß-actin in HK2 cells after exposure to MG (a) and the reduction of MG-H1 after co-incubation with Ans and Carn (b). Effect of well described MG-H1 quencher aminoguanidine (AG) on transepithelial and transendothelial resistance (c). While in HK-2 cells (left) AG resulted in higher TER compared to MG treatment only, this effect was significantly lower compared to Ans. In HUVEC (right), AG had no effects. One-way ANOVA followed by Tukey’s test with correction for multiple comparisons. *p < 0.05, ****p < 0.0001.