Intermittent exposure of cultured endothelial cells to physiologically relevant fructose concentrations has a profound impact on nitric oxide production and bioenergetics

Abstract

Hyperglycaemia is known to induce endothelial dysfunction and changes in metabolic function, which could be implicated in diabetes-induced cardiovascular disease. To date, however, little is known about the impact of physiologically relevant concentrations of fructose on endothelial cells. A novel in vitro model was devised to establish the impact of substitution of a small proportion of glucose with an equal concentration (0.1 mM or 1 mM) of fructose on EA.hy926 endothelial cells during periodic carbohydrate "meals" superimposed on a normoglycaemic (5.5 mM) background. Parallel experiments were conducted using meals consisting of normoglycaemic glucose, intermediate glucose (12.5 mM) or profound hyperglycaemia (25 mM), each delivered for 2 h, with and without substituted fructose over 50 h. Outcome measures included nitrite as a surrogate marker of the mediator of healthy endothelial function, nitric oxide (NO), and a range of bioenergetic parameters using a metabolic analyser. Despite its relatively low proportion of carbohydrate load, intermittent fructose induced a substantial reduction (approximately 90%) in NO generation in cells treated with either concentration of fructose. Cell markers of oxidative stress were not altered by this treatment regimen. However, the cells experienced a marked increase in metabolic activity induced by fructose, irrespective of the glucose concentration delivered simultaneously in the "meals". Indeed, glucose alone failed to induce any metabolic impact in this model. Key metabolic findings were a 2-fold increase in basal oxygen consumption rate and a similar change in extracellular acidification rate-a marker of glycolysis. Non-metabolic oxygen consumption also increased substantially in cells exposed to fructose. There was no difference between results with 0.1 mM fructose and those with 1 mM fructose. Low, physiologically relevant concentrations of fructose, delivered in a pattern that mimics mealtime consumption, had a profound impact on endothelial function and bioenergetics in an in vitro cell model. The results suggest that endothelial cells are exquisitely sensitive to circulating fructose; the potential ensuing dysfunction could have major implications for development of atherosclerotic disease associated with high fructose consumption.

Fig 1. Representation of the experimental regimens for EA.hy926 cells (glucose only model–top panel; glucose+fructose–bottom panel).

The total treatment time was 50 h, incorporating 7 treatment periods at intervals designed to mimic popular mealtimes in human populations. Two parallel sets of experiments were undertaken: cells destined for metabolic analysis were cultured in specialized 96 well microplates for this purpose; analysis was undertaken at the end of the 50 hour treatment period. Cells destined for ROS/superoxide analysis were cultured in standard 96 well microplates. Supernatant was harvested after 48 h for nitrite analysis, while ROS/superoxide analysis was conducted on the cells after the final 2 h meal exposure.

Fig 2. Nitrite accumulation as a surrogate for NO generation.

Effect of intermittent glucose and glucose/fructose exposure on nitrite accumulation at NG, IG and HG test conditions (n = 12–14; ***P<0.001 Bonferroni post-test compared to glucose only control at equivalent concentrations).

Fig 3. Cellular bioenergetics.

Effect of intermittent glucose (G only)/glucose+fructose (G+0.1F, G+1F) exposure on cellular bioenergetics, described by (A) basal OCR, (B) complex V OCR, (C) maximal respiration, (D) proton leak, (E) NMOCR and (F) ECAR. Data are shown as mean ±SD for n = 6–8 separate experiments. Means are compared using 2-factor ANOVA with Bonferroni post-test. ***P<0.001; **P<0.01 for fructose-treated cells compared to glucose only treated cells with the same carbohydrate load. For proton leak, there was a significant impact of carbohydrate load (P<0.001) and of fructose (main effect, P<0.05). For D only: ⁺⁺⁺P<0.001 and ⁺⁺P<0.01 v equivalent bar at 5.5 mM load; ^###P<0.001 and ^##P<0.01 v equivalent bar at 12.5 mM load.

Fig 4. Energy map.

Energy map for cells treated with intermittent glucose and glucose/fructose. Dotted lines represent the direction of change induced by fructose substitution for glucose (0.1 mM–light symbols; or 1 mM–dark symbols) in cells treated with different glucose concentrations. Data are expressed as mean ± SEM (n = 8–12 for each mean depicted).

Fig 5. ROS generation.

Effect of intermittent glucose and glucose/fructose exposure on total ROS-RNS (A) and superoxide (B; n = 7–12; #P<0.05 compared to 5.5 mM glucose alone).