Increased production of reactive oxygen species in hyperglycemic conditions requires dynamic change of mitochondrial morphology

Abstract

Increased production of mitochondrial reactive oxygen species (ROS) by hyperglycemia is recognized as a major cause of the clinical complications associated with diabetes and obesity [Brownlee, M. (2001) Nature 414, 813-820]. We observed that dynamic changes in mitochondrial morphology are associated with high glucose-induced overproduction of ROS. Mitochondria undergo rapid fragmentation with a concomitant increase in ROS formation after exposure to high glucose concentrations. Neither ROS increase nor mitochondrial fragmentation was observed after incubation of cells with the nonmetabolizable stereoisomer L-glucose. However, inhibition of mitochondrial pyruvate uptake that blocked ROS increase did not prevent mitochondrial fragmentation in high glucose conditions. Importantly, we found that mitochondrial fragmentation mediated by the fission process is a necessary component for high glucose-induced respiration increase and ROS overproduction. Extended exposure to high glucose conditions, which may mimic untreated diabetic conditions, provoked a periodic and prolonged increase in ROS production concomitant with mitochondrial morphology change. Inhibition of mitochondrial fission prevented periodic fluctuation of ROS production during high glucose exposure. These results indicate that the dynamic change of mitochondrial morphology in high glucose conditions contributes to ROS overproduction and that mitochondrial fission/fusion machinery can be a previously unrecognized target to control acute and chronic production of ROS in hyperglycemia-associated disorders.

Fig. 1.

HG-induced ROS increase coincides with mitochondrial fragmentation. (A and B) HG incubation transiently increased ROS production. ROS was measured by oxidation of carboxy-dichlorodihydrofluorescein diacetate (Carboxy-DCF) at different times after HG incubation. Experiments were repeated more than five times and produced consistent results. Representative fluorescence quantification is shown in B. (C–F) Mitochondrial fragmentation in HG incubation. Clone 9 cells, whose mitochondria were labeled with red fluorescent protein, were incubated in 25 mM glucose. (C) Fragmented mitochondria were prevalent after 15 and 30 min in HG and normal tubular mitochondria after 60 min. (D) Cell counting is shown. (E) Analyses of mitochondrial morphology by form factor and aspect ratio. (E Left) Graphs plotted for form factor and aspect ratio of individual mitochondria in one cell are shown. (E Right) Enlarged images of a part of the corresponding cell used for the analysis, showing detailed mitochondrial morphology. (F) Average values of form factor and aspect ratio, as well as number of mitochondria, from five cells at each time point show transient fragmentation of mitochondria after incubation in HG conditions.

Fig. 2.

Roles of DLP1, glucose metabolism, and pyruvate transport in HG-induced ROS increase and mitochrondrial fragmentation. (A) Mitochondrial fragmentation in HG requires mitochondrial fission machinery. Cells transfected with DLP1-K38A (T) maintained elongated tubular mitochondria in HG media, whereas mitochondria in untransfected cells (U) were short and fragmented. (B–D) d-glucose, but not l-glucose, causes ROS increase and mitochondrial fragmentation. (B and C) DCF fluorescence increased in cells incubated in d-glucose, but not in l-glucose. (D) Both form factor and aspect ratio of mitochondria did not change in cells from l-glucose incubation. (E and F) Inhibiting mitochondrial pyruvate uptake prevents ROS increase in HG conditions. Preincubation with 0.1 mM CHC blocked ROS increase (E) but did not affect mitochondrial fragmentation (F).

Fig. 3.

Mitochrondrial fragmentation is necessary for HG-induced ROS overproduction and respiration increase. (A and B) ROS increase is not required for mitochondrial fragmentation in HG incubation. Cells pretreated with 100 nM carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP) blocked ROS increase (A) but mitochondria were still fragmented upon HG exposure (B). (C and D) Inhibition of mitochondrial fission prevented HG-induced ROS production. (C) Cells transfected with GFP-DLP1-K38A showed bright aggregates of GFP signals (Left, arrows and arrowheads). Under normal glucose conditions, ethidium fluorescence of transfected cells (Upper Right, arrows) was similar to that in untransfected cells. At 30-min exposure to 25 mM glucose, ethidium fluorescence remained low in transfected cells (Lower Right, arrowheads) whereas more intense ethidium fluorescence was observed in untransfected cells. (D) Quantification of fluorescence intensity is shown. (E and F) Inhibition of mitochondrial fission prevents respiration increase in HG conditions. (E) Oxygen consumption profiles of cells in normal and HG conditions. DLP1-K38A overexpression was induced by tetracycline (Tet-induced). (F) Respiration rates show that DLP1-K38A overexpression (Tet-induced) blocked HG-induced respiration increase (arrowheads).

Fig. 4.

Fission/fusion dynamics of mitochondria in HG-induced ROS increase. (A and B) Promoting mitochondrial fusion prevented HG-induced ROS increase. Whereas the ethidium fluorescence increased without tetracycline induction (A Upper and Control in B), Mfn2 overexpression by tetracycline greatly reduced ROS levels in HG (A Lower and Mfn2 in B). (C) Time-lapse images of mitochondria upon HG exposure. Tubular mitochondria (Upper Left) became fragmented in HG (Upper Right). Concurrent fissions occurred within tubules (Lower). Contraction and condensation of mitochondria are apparent (arrows).

Fig. 5.

ROS levels fluctuate during prolonged exposure to HG conditions. (A) ROS levels were measured in Clone 9 cells with DHE every 2 h after incubation with 50 mM glucose. Fluorescence intensity was normalized against the intensity at time 0 (F/F₀). The second phase of ROS increase was observed at 10-h incubation, and the ROS levels remained high until the 18-h time point, then decreased thereafter. (B and C) Fragmented mitochondria were prevalent in cells incubated for 10–18 h, and mitochondria returned to tubular morphology at 22–24 h in HG (gray bars in C). Mitochondria from cells incubated for 16 and 22 h are shown in B. Cells overexpressing DLP1-K38A maintained elongated tubular mitochondria during prolonged incubation with HG concentrations (C, solid bars). (D) Cells overexpressing GFP-DLP1-K38A (arrows in D) showed little change in ROS levels, which remained low throughout the extended incubation in HG (black bars) whereas ROS levels in untransfected cells (gray bars) increased and decreased.