Mitochondrial tubules are short and fragmented in sustained high glucose conditions. MitoTracker staining shows the tubular mitochondrial morphology in H9c2 cells in normal glucose conditions (A). Small and fragmented mitochondria were evident in cells incubated in 35mM glucose for 48 hours (B), whereas 30mM mannitol did not affect mitochondrial morphology (C). Scale bar: 20µm. Approximately half of the cell population contained fragmented mitochondria after high glucose incubation (D). Error bars represent SEM. (E–G) Computer-assisted morphology analyses of mitochondria. Individual mitochondria from cells incubated in the high glucose concentration for 48 hours have lower values for both form factor and aspect ratio (F: short and fragmented mitochondria) whereas many of those from normal glucose and high mannitol conditions show increased values (E and G: tubular mitochondria).

Inhibiting mitochondrial fission prevents high glucose-induced apoptosis. H9c2 cells were infected with Ad-DLP1-K38A and assessed for apoptosis after a 48-hour incubation in 35mM glucose. The number of annexin V-positive cells in the Ad-DLP1-K38A-infected culture in high glucose conditions was as low as that in normal glucose conditions (A). Cytochrome c immunofluorescence indicates cytochrome c in tubular mitochondria in normal glucose conditions (B), whereas it shows an increased cytosolic staining (cells marked by asterisks) along with fragmented mitochondria in cells incubated in the high glucose concentration (C). Cells infected with Ad-DLP1-K38A in high glucose conditions still show cytochrome c retained in elongated tubular mitochondria (D). Scale bar: 10µm. The Ad-DLP1-K38A infection normalized caspase activities in high glucose incubation (E). Error bars represent SEM. * indicates a statistically significant difference compared with the Control at p< 0.05.

Inhibiting mitochondrial fission in sustained high glucose incubation prevents ROS overproduction and MPT. (A–D) Control cells show low levels of ROS in normal glucose conditions (A), and elevated ROS in the 48-hour high glucose incubation (B). ROS levels did not increase in H9c2 cells infected with Ad-DLP1-K38A in the 48-hour incubation in high glucose conditions (C). Scale bar: 20µm. Quantification of ethidium fluorescence indicates that ROS levels of Ad-DLP1-K38A-infected cells in high glucose conditions were similar to those in normal glucose incubation (D). Error bars represent SEM. ** indicates a statistically significant difference compared with the Control at p<0.01. (E) Co-Q assays for the MPT show that H9c2 cells infected with Ad-DLP1-K38A in high glucose conditions exhibited calcein levels similar to those of cells in normal glucose conditions. (F, G) Inhibition of mitochondrial fission in primary cells from cardiovascular system prevented the ROS increase in sustained high glucose conditions. Ad-DLP1-K38A infection decreased ROS levels in BAEC (F) and SMC (G) incubated for 48 hours in both 20 and 35mM glucose. Error bars represent SEM.

Increased ROS levels in sustained high glucose conditions cause the MPT. (A, A’, B, B’) H9c2 cells were incubated with calcein-AM and CoCl₂ to assess the MPT in normal and high glucose conditions. Mitochondria were labeled with MitoTracker (A and B). Mitochondria in a majority of the cells in normal glucose conditions were calcein-positive (A’), whereas the mitochondrial calcein fluorescence was lost or decreased upon the MPT, which is more prevalent in cells incubated in 35mM glucose for 48 hours (B’). Scale bar: 20µm. (C) Quantification of calcein fluorescence in 40 randomly selected cells. High glucose incubation (HG) decreased calcein fluorescence. Bongkrekic acid (BA, 5µM), FCCP (100nM), or MnTMPyP (25µM) prevented the decrease of the mitochondrial calcein fluorescence in high glucose conditions. (D–F) Blocking the MPT or the ROS increase in high glucose incubation decreases apoptotic cell death. Trypan blue staining shows an approximately three-fold decrease of cell death in cells treated with BA, FCCP, or MnTMPyP after 48 hours in 35mM glucose (D). Annexin V cell surface labeling (E) and caspase activity (F) were also reduced by treating cells with BA, FCCP or MnTMPyP in high glucose condition. Error bars represent SEM.

The mitochondrial fission protein DLP1 is necessary for mitochondrial fragmentation in sustained high glucose conditions. (A) Immunoblotting of H9c2 cell lysates from uninfected (Un), control virus-infected (Ad-GFP), and DLP1-K38A virus-infected (Ad-DLP1-K38A) cells. While endogenous DLP1 was detected as a faint band in all three lysates, a dense band of the slightly larger DLP1 (asterisk) was found in Ad-DLP1-K38A-infected cells, indicating an overexpression of DLP1-K38A. (B) DLP1 immunofluorescence showed the presence of characteristic bright punctate aggregates, indicative of the DLP1-K38A expression (left panel). Ad-DLP1-K38A-infected cells show the elongated mitochondrial morphology due to inhibition of mitochondrial fission (right panel). Scale bar: 20µm. (C–F) MitoTracker staining of control-infected cells shows the tubular morphology in normal glucose conditions (C), but a fragmented phenotype after the 35mM glucose incubation for 48 hours (D). Cells infected with Ade-DLP1-K38A contained elongated mitochondria and no fragmented phenotype in high glucose incubation (E). Scale bar: 10µm. Less than 5% of Ad-DLP1-K38A-infected cells contained fragmented mitochondria after 30- and 48-hour incubations in high glucose conditions (F) Error bars represent SEM. (G–I) Computer-assisted quantitative analyses of mitochondrial morphology in cells from normal glucose (G), high glucose (H), and DLP1-K38A/high glucose (I) conditions. Mitochondria in cells expressing DLP1-K38A have higher values for both form factor and aspect ratio in the high glucose conditions, indicating the abolition of high glucose-induced mitochondrial fragmentation.

Sustained incubation of neonatal cardiomyocytes or H9c2 cells in high glucose conditions increases ROS production and cell death. (A) Neonatal cardiomyocytes were maintained in a normal glucose concentration (5.56mM). ROS levels and cell death were assessed at 30 and 48 hours after the cells were transferred to media containing 20 and 35mM glucose. ROS levels were measured by DHE oxidation (A, left panel). Quantification of fluorescence intensity showed approximately a two-fold increase of ROS in 35mM glucose compared to normal conditions at both 30- and 48-hour incubations. The increase of ROS in 20mM glucose was less than that in 35mM glucose. Trypan blue staining showed a 3–4 fold increase of cell death in 20 and 35mM glucose concentrations (A, middle panel). Cell surface labeling with annexin V was used to assess whether the cell death in high glucose incubation is possibly apoptotic (A, right panel). Cells incubated in 20 and 35mM glucose showed increased annexin V labeling. Error bars represent SEM. * and ** indicate statistically significant differences compared with the Normal Glc at the corresponding time points at p< 0.05 and p<0.01, respectively. (B) The cardiac myoblast cell line H9c2 shows increased ROS levels. H9c2 cells showing increased ethidium (upper panels) and DCF (lower panels) fluorescence in prolonged high glucose conditions (35mM Glc). Cells in normal glucose (5.56mM) had low ROS levels (left panels). Fluorescence quantification shows an approximately two-fold increase of the ROS level after 30- and 48-hour incubations in 35mM glucose (right panels). Scale bar: 20µm. (C–E) Prolonged high glucose incubation of H9c2 cells increases cell death. Proportions of trypan blue-positive cells (C) and annexin V-positive cells (D) were increased by approximately 5 and 10 folds at 30 and 48 hours, respectively, in 35mM glucose incubation. A significant increase of fluorescence resulting from caspase-mediated cleavage of the substrate rhodamine-110 bis-L-aspartic acid amide was observed in cells after a 48-hour incubation in 35mM glucose (E). Addition of 30mM mannitol instead of glucose increased neither cell death nor caspase activity. Error bars represent SEM.