C-Peptide Activates AMPKα and Prevents ROS-Mediated Mitochondrial Fission and Endothelial Apoptosis in Diabetes

Human umbilical vein endothelial cells (HUVECs) were isolated from human umbilical cord veins as previously described (17). Cells were maintained in M199 culture media supplemented with 20% FBS, 3 ng/mL basic fibroblast growth factor, 5 units/mL heparin, 100 IU/mL penicillin, and 100 μg/mL streptomycin. Cells between passages 3 and 7 were incubated overnight in low-serum medium (M199 supplemented as described above except with 5% FBS, 1 ng/mL basic fibroblast growth factor, and 25 mmol/L HEPES).

Intracellular ROS generation was measured using H₂DCFDA staining (Molecular Probes, Eugene, OR) as previously described (17). Cells on 2% gelatin-coated coverslips were treated with 33 mmol/L d-glucose for 48 h and stained with 10 μmol/L H₂DCFDA for 10 min in serum-free, phenol red-free media. For measurement of mitochondrial ROS, cells were incubated with 5 μmol/L MitoSox Red (Molecular Probes) for 30 min in low-serum media. Cells were stained with 2 μmol/L DAF-FM diacetate (Molecular Probes) for 30 min in low-serum medium for measuring intracellular NO production. Labeled cells were immediately analyzed by confocal microscopy (Fluoview-300; Olympus). Single-cell fluorescence intensities were determined for 30 cells per experiment. The levels of intracellular and mitochondrial ROS and intracellular NO were determined by comparing fold change in fluorescence intensity of treated cells versus control cells.

Cells were incubated with 33 mmol/L d-glucose for 48 h, and entire mitochondria were visualized in live cells by staining with 200 nmol/L Mito Tracker Red CMXRos (Molecular Probes) for 30 min at 37°C. Cells with a predominantly intact network of tubular mitochondria were identified as normal. Cells with disrupted and predominantly spherical mitochondria were identified as having mitochondrial fission. Ninety cells from three independent experiments were used to calculate the percentage of cells undergoing mitochondrial fission.

∆Ψ_m was monitored by confocal microscopy as previously described (33). Cells were incubated with 33 mmol/L d-glucose for 48 h and stained with 2 μmol/L of JC-1 (Molecular Probes) for 30 min at 37°C. Fluorescence intensities were determined at the single-cell level using confocal microscopy. Data are expressed as a normalized ratio of the fluorescence intensity of the monomeric form (green) to the J-aggregate form (red).

Cell death was assessed by 3,3′-dihexyloxacarbocyanine iodide (DiOC₆)/propidium iodide (PI) double staining using confocal microscopy as previously described (34). Briefly, cells were incubated with 33 mmol/L d-glucose for 72 h and stained with 50 nmol/L DiOC₆ and 10 µg/mL PI for 20 min. Stained cells were analyzed by confocal microscopy and differentiated as viable cells (DiOC₆^bright/PI⁻), early apoptotic cells (DiOC₆^dim/PI⁻), and late apoptotic cells (DiOC₆^dim/PI⁺). Total cell death (defined as the sum of early and late apoptotic cells) and frequency distributions were expressed as percentages of the total cell number.

To suppress AMPKα expression, HUVECs were transfected with AMPKα1/2-specific small interfering (si)RNA (Santa Cruz Biotechnology, Santa Cruz, CA) according to the manufacturer’s instructions. Briefly, subconfluent cells were transfected overnight with the AMPKα siRNA duplex, which was synthesized to target the coding sequence of human AMPKα1/2 mRNA. A noncoding siRNA (Dharmacon, Lafayette, CO) was used as a control. Transfection was performed using siLentFect Lipid Reagent (Bio-Rad Laboratories, Hercules, CA) for 24 h.

Six-week old male C57BL/6 mice were obtained from KOTECH (Pyungtaek, Korea). Experiments were performed in accordance with the guidelines of the Kangwon Institutional Animal Care and Use Ethics Committee. Diabetes was induced in mice by a single injection of streptozotocin (150 mg/kg body wt i.p.) as previously described (17). Sufficient hyperglycemia was observed 2 days postinjection as determined by measuring blood glucose and glucosuria. Mice with nonfasting blood glucose levels >16 mmol/L, polyuria, and glucosuria were considered diabetic. One week after the streptozotocin injection, an Alzet mini-osmotic pump 2004 (DURECT, Cupertino, CA) containing C-peptide (American Peptide Company, Sunnyvale, CA) in PBS with a delivery rate of 35 pmol/min/kg was subcutaneously implanted into each diabetic mouse in one group. The other diabetic and control mice received sham operations. During the continuous subcutaneous C-peptide perfusion, serum C-peptide levels were measured using a C-peptide Enzyme Immunoassay Kit (RayBiotech, Norcross, GA). During the whole course of these experiments, the vital statistics of each mouse were recorded.

Aorta from control (n = 8), diabetic (n = 8), and C-peptide–supplemented diabetic (n = 7) mice were dissected and cut longitudinally to open the endothelium for immediate staining. For intracellular ROS measurements, aortic segments were quickly transferred to serum-free M199 media and incubated at 37°C with H₂DCFDA for 10 min. For assessment of mitochondrial fission, endothelial layers from each group were stained at 37°C for 30 min with 200 nmol/L MitoTracker Red CMXRos. Stained aortic segments were mounted on glass slides and quickly observed by confocal microscopy.

Proteins were extracted from HUVECs and aortic segments of mice with lysis buffer containing 50 mmol/L HEPES (pH 7.5), 1 mmol/L EDTA, 150 mmol/L sodium chloride, 1% Triton X-100, 10 μg/mL aprotinin, 10 μg/mL leupeptin, 0.1 mmol/L phenylmethanesulfonylfluoride, 25 mmol/L β-glycerophosphate, and 2 mmol/L sodium orthovanadate. Protein extracts were resolved by SDS-PAGE and transferred to polyvinylidene difluoride membranes. Protein expression was analyzed with monoclonal antibodies against phospho-AMPKα, AMPKα, dynamin-related protein 1 (Drp-1) (Cell Signaling Technology, Beverly, MA), and fission-1 protein (Fis-1) (Santa Cruz Biotechnology). Protein bands were visualized with a chemiluminescence reagent (Pierce, Rockford, IL).

Data processing was performed using Origin 6.1 (OriginLab, Northampton, MA) and expressed as mean ± SD of at least three independent experiments. Statistical significance was determined using ANOVA. A value of P < 0.05 was considered statistically significant.

In HUVECs, high glucose (33 mmol/L) significantly increased the levels of intracellular and mitochondrial ROS (Fig. 1A and B). However, osmotic control l-glucose had no effect on the ROS levels (Fig. 1B). To understand the effect of cytosolic ROS on mitochondrial ROS levels, we inhibited PKC and NADPH oxidase that mediate high glucose–induced cytosolic ROS generation in endothelial cells. The PKC inhibitors, GF109203X and Ro-31-8220, attenuated high glucose–induced generation of intracellular and mitochondrial ROS in a dose-dependent manner (Fig. 1C). The well-established NADPH oxidase inhibitors, apocynin and diphenyliodinium (DPI), also prevented intracellular and mitochondrial ROS production in a dose-dependent manner (Fig. 1D). These results suggest that PKC-dependent activation of NADPH oxidase increases cytosolic ROS, which then promotes mitochondrial ROS generation under hyperglycemic conditions.

The superoxide dismutase 2 mimetic Mito-TEMPO inhibited high glucose–induced mitochondrial ROS levels in a dose-dependent manner (Fig. 1E). Interestingly, high glucose–induced intracellular ROS was also inhibited by mito-TEMPO in a dose-dependent manner at a rate similar to mitochondrial ROS inhibition (Fig. 1E). The role of mitochondrial ROS in the regulation of intracellular ROS was further investigated using Mdivi-1, a potent inhibitor of Drp-1–mediated mitochondrial fission (35). Mitochondrial fission is believed to be an important cause of hyperglycemia-induced endothelial damage as a result of overproduction of mitochondrial ROS (20). Mdivi-1 inhibited high glucose–induced production of mitochondrial and intracellular ROS in a dose-dependent manner (Fig. 1F). Thus, mitochondrial ROS regulates production of intracellular ROS in response to high glucose. Blocking the production of cytosolic ROS inhibits mitochondrial ROS generation, demonstrating a potential interregulation between cytosolic and mitochondrial ROS in response to high glucose in endothelial cells.

High-glucose treatment significantly induced mitochondrial fission in HUVECs as assessed by MitoTracker Red staining. Control cells showed an intact mitochondrial network. Cells bearing predominantly fragmented and spherical mitochondria were considered to have undergone mitochondrial fission (Fig. 2A and B). High glucose–induced mitochondrial fission was prevented by PKC inhibitors GF109203X and Ro-31-8220 and NADPH oxidase inhibitors apocynin and DPI (Fig. 2C). These inhibitors also blocked cytosolic ROS generation in response to high glucose in endothelial cells (Fig. 1C and D). However, l-glucose had no effect on mitochondrial fission (Fig. 2B). High glucose–induced mitochondrial fission was also normalized using mito-TEMPO and Mdivi-1 (Fig. 2D). Thus, high glucose–induced mitochondrial fission is mediated by both cytosolic and mitochondrial ROS in endothelial cells.

High glucose stimulated ∆Ψ_m collapse in endothelial cells as assessed by JC-1 staining (Fig. 2E), whereas l-glucose had no effect (Fig. 2F). The PKC inhibitors, GF109203X and Ro-31-8220, and the NADPH oxidase inhibitors, apocynin and DPI, rescued high glucose–induced ∆Ψ_m collapse (Fig. 2F). Furthermore, high glucose–induced ∆Ψ_m collapse was normalized by mito-TEMPO and Mdivi-1 (Fig. 2G). These results demonstrate that high glucose–induced ∆Ψ_m collapse is mediated by cytosolic and mitochondrial ROS in endothelial cells.

C-peptide inhibited high glucose–induced generation of intracellular and mitochondrial ROS in a dose-dependent manner, with maximal effect at 1 nmol/L (Fig. 3A). As intracellular and mitochondrial ROS both mediate high glucose–induced mitochondrial fission and ∆Ψ_m collapse (Fig. 2), we studied the effects of C-peptide on mitochondrial fission and ∆Ψ_m collapse in response to high glucose. C-peptide prevented high glucose–stimulated mitochondrial fission (Fig. 3B). Consistently, high glucose–induced expression of fission proteins Drp-1 and Fis-1 was inhibited by C-peptide (Fig. 3C–E). Additionally, C-peptide rescued high glucose–induced ∆Ψ_m collapse (Fig. 3F). However, heat-inactivated C-peptide had no inhibitory effect on high glucose–induced generation of intracellular and mitochondrial ROS (Fig. 3A), mitochondrial fission (Fig. 3B), or ∆Ψ_m collapse (Fig. 3F). Thus, C-peptide prevents high glucose–induced mitochondrial fission and ∆Ψ_m collapse by inhibiting ROS generation.

C-peptide stimulated phosphorylation of AMPKα and reversed high glucose–induced dephosphorylation of AMPKα (Fig. 4A and B). However, heat-inactivated C-peptide had no significant effect on AMPKα phosphorylation (data not shown). AMPKα1/2-specific siRNA suppressed AMPKα expression in a dose-dependent manner (Fig. 4C). AMPKα knockdown reversed C-peptide–mediated inhibition of high glucose–induced intracellular and mitochondrial ROS production (Fig. 4D). Furthermore, the AMPK inhibitor, compound C, dose-dependently reversed C-peptide–mediated prevention of high glucose–stimulated ROS generation (Fig. 4E). AMPK functions as a physiological suppressor of NADPH oxidase and ROS production in endothelial cells (29,30). Thus, it is likely that AMPKα-mediated inhibition of NADPH oxidase is an important mechanism by which C-peptide prevents production of intracellular and mitochondrial ROS in response to high glucose.

We then used AMPKα1/2 siRNA and compound C to investigate the contribution of AMPK to the altered mitochondrial dynamics stimulated by high glucose. AMPKα1/2 siRNA reversed C-peptide–mediated inhibition of mitochondrial fission in the presence of high glucose (Fig. 4F). Similarly, compound C prevented C-peptide–mediated inhibition of high glucose–induced mitochondrial fission (Fig. 4G). The essential role of AMPKα in C-peptide–mediated prevention of ∆Ψ_m collapse induced by high glucose was also found using AMPKα1/2 siRNA and compound C (Fig. 4H and I). Taken together, C-peptide inhibits high glucose–induced generation of intracellular and mitochondrial ROS through activation of AMPKα, resulting in prevention of mitochondrial fission and ∆Ψ_m collapse in endothelial cells.

Additionally, we further investigated whether NO production by C-peptide acts upstream of AMPK. C-peptide significantly elevated the level of intracellular NO at 0.5 h, with maximal effect at 2 h (Fig. 4J), and the elevated level returned back to the basal level at 12 h (data not shown). l-N^G-nitro-l-arginine methyl ester (l-NAME) prevented C-peptide–induced NO production, but compound C had no effect (Fig. 4J). C-peptide activation of AMPKα was blocked by l-NAME, and the NO donor S-nitroso-N-acetylpenicillamine reversed high glucose–induced AMPKα dephosphorylation at 48 h (Fig. 4K and L).

AICAR and metformin each inhibited high glucose–induced generation of intracellular and mitochondrial ROS in a dose-dependent manner (Fig. 5A and B). AICAR and metformin also rescued high glucose–mediated inhibition of AMPKα as demonstrated by its phosphorylation at Thr¹⁷² (Fig. 5C and D). High glucose–stimulated mitochondrial fission was significantly inhibited by AICAR and metformin (Fig. 5E). Consistently, these AMPK activators inhibited high glucose–induced expression of Drp-1 and Fis-1 (data not shown). ∆Ψ_m collapse induced by high glucose was significantly reversed by AICAR and metformin (Fig. 5F). Additionally, the ROS scavenger Trolox significantly prevented high glucose–stimulated mitochondrial fission and ∆Ψ_m collapse (Fig. 5E and F). Thus, AICAR and metformin mimic C-peptide by preventing high glucose–induced mitochondrial fission and ∆Ψ_m collapse through AMPKα-mediated inhibition of intracellular and mitochondrial ROS generation.

We used the DiOC₆/PI double-staining assay (34) to exam whether C-peptide prevents high glucose–induced apoptosis by activating AMPK signaling. C-peptide prevented high glucose–induced apoptosis of endothelial cells (Fig. 6A). d-glucose induced the endothelial cell apoptosis in a dose-dependent manner, which was inhibited by C-peptide (Supplementary Fig. 1). However, apoptotic cell death was neither activated by l-glucose nor inhibited by heat-inactivated C-peptide (Fig. 6A). Endothelial cell apoptosis was significantly inhibited by AMPK activators AICAR and metformin, the NADPH oxidase inhibitor apocynin, and the mitochondrial ROS scavenger mito-TEMPO (Fig. 6A). Thus, C-peptide prevents high glucose–induced endothelial cell apoptosis through activation of AMPK and inhibition of ROS generation. Moreover, inhibition of high glucose–induced apoptosis by Mdivi-1 (Fig. 6A) provides further evidence that Drp-1–dependent mitochondrial fission is important for high glucose–induced endothelial cell apoptosis.

The key role of AMPKα in C-peptide–mediated protection of endothelial cell death was further elucidated using AMPKα 1/2 siRNA and compound C (Fig. 6B and C). Transfection of AMPKα 1/2 siRNA or treatment with compound C significantly increased apoptosis (Fig. 6B and C). Moreover, C-peptide–mediated inhibition of high glucose–stimulated endothelial cell apoptosis was reversed by AMPKα 1/2 siRNA or compound C (Fig. 6B and C). Further, compound C reversed the C-peptide inhibition of dose-dependent high glucose–induced apoptosis (Supplementary Fig. 1). Thus, C-peptide protects endothelial cells from apoptosis by AMPKα-mediated inhibition of ROS generation and prevention of mitochondrial fission in hyperglycemia.

The role of AMPKα in C-peptide–mediated prevention of hyperglycemia-induced ROS generation and mitochondrial fission was further investigated in the aortas of streptozotocin diabetic mice. Five-week diabetic mice showed loss of body weight, increased rate of food and water consumption, and severe hyperglycemia (30.41 mmol/L) with glucosuria compared with nondiabetic controls (Supplementary Table 1). However, these parameters were not improved in diabetic mice supplemented with C-peptide. Serum C-peptide levels significantly decreased in untreated diabetic mice (P < 0.001) and were fully restored to the normal physiologic range (1.54 nmol/L) in diabetic mice supplemented with C-peptide using osmotic pumps (Supplementary Table 1).

Phosphorylation of AMPKα was decreased in the aortas of diabetic mice and was normalized by C-peptide replacement therapy (Fig. 7A and B). C-peptide supplementation also inhibited hyperglycemia-induced generation of intracellular ROS in aortic endothelial cells of diabetic mice (Fig. 7C and D). Furthermore, we assessed the preventive effect of C-peptide on mitochondrial fission. Hyperglycemia induced fragmentation and disorganization of mitochondria in the aortas of diabetic mice. C-peptide supplementation prevented hyperglycemia-stimulated mitochondrial fission (Fig. 7E). Consistently, we recently reported that C-peptide prevents hyperglycemia-induced apoptosis in the aortas of diabetic mice (17). Thus, consistent with our in vitro findings, C-peptide–mediated protection against hyperglycemia-induced apoptosis is due to AMPKα-dependent prevention of ROS generation and mitochondrial fission in aortic endothelium.

The development of vascular complications is associated with C-peptide deficiency in diabetes. We recently reported that C-peptide protects against ROS-mediated endothelial cell apoptosis; however, the mechanism by which C-peptide exerts ROS inhibition is not understood. In this article, we elucidated a new mechanism by which C-peptide prevents hyperglycemia-induced vasculopathy. As shown in Fig. 8, C-peptide inhibits high glucose–induced generation of intracellular and mitochondrial ROS through activation of AMPKα, which results in suppression of mitochondrial fission and ∆Ψ_m collapse and protects against endothelial cell apoptosis in response to hyperglycemia.

We targeted PKC-dependent NADPH oxidase and mitochondria because they are reported to be the two main sources of ROS generation in response to hyperglycemia in endothelial cells (12,14). Cytosolic NADPH oxidase and mitochondria played essential roles in amplifying intracellular ROS production in hyperglycemia. Inhibitors of the PKC-dependent NADPH oxidase pathway prevented high glucose–induced mitochondrial ROS generation, suggesting that cytosolic ROS facilitates mitochondrial ROS generation. Similarly, mitochondrial and intracellular ROS generation was inhibited by mito-TEMPO and Mdivi-1. Our results demonstrated the interregulation between cytosolic and mitochondrial ROS generation in response to high glucose in endothelial cells. Consistent with our findings, there was a recent report (36) of a vicious cycle in which mitochondrial superoxide stimulates cytosolic NADPH oxidase in a feed-forward fashion in response to angiotensin II in endothelial cells. Thus, it is likely that hyperglycemia triggers a positive feedback loop of intracellular ROS amplification in endothelial cells, in which cytosolic ROS production by PKC-dependent NADPH oxidase potentiates mitochondrial ROS increase and release, resulting in a further increase in cytosolic ROS.

We further investigated the potential link between ROS-dependent mitochondrial fission and apoptosis of endothelial cells in hyperglycemia. We found a prominent increase in mitochondrial fission upon treatment with high glucose in HUVECs. Blocking cytosolic ROS generation, inhibiting mitochondrial fission with Mdivi-1, or enhancing mitochondrial antioxidant activity with mito-TEMPO prevented mitochondrial fission. These results demonstrate that both cytosolic and mitochondrial ROS can enhance mitochondrial fission. We also found that ROS production and its downstream mitochondrial fission contributed to ∆Ψ_m collapse, which is considered to be an early stage of apoptosis (37). High glucose–induced ∆Ψ_m collapse was successfully prevented by blocking cytosolic and mitochondrial ROS or by inhibiting mitochondrial fission. Thus, intracellular ROS-mediated mitochondrial fission induces ∆Ψ_m collapse, which triggers apoptosis of endothelial cells during hyperglycemia (Fig. 8).

Our studies identify C-peptide–mediated activation of AMPKα as a novel mechanism that protects against hyperglycemia-induced vascular damage. AMPKα is considered to be an emerging therapeutic target for preventing diabetes complications (38–40). AMPK activation depends on phosphorylation of the α catalytic subunit at Thr¹⁷² and AMP binding to the γ subunit, whereas ATP promotes dephosphorylation of AMPK (41). Previous studies have shown that AMPKα is dephosphorylated and has diminished activity in diabetes (31,32,42). In this article, we demonstrated that C-peptide activates phosphorylation of AMPKα. Further, we showed that high glucose–stimulated dephosphorylation of AMPKα was reversed by C-peptide in HUVECs. AMPKα siRNA or the AMPK inhibitor, compound C, also inhibited C-peptide–mediated prevention of high glucose–induced apoptosis of endothelial cells by inhibiting ROS generation, mitochondrial fission, and ∆Ψ_m collapse. Additionally, the beneficial role of C-peptide against intracellular ROS amplification and mitochondrial fission through AMPKα activation was elucidated in the aortas of streptozotocin diabetic mice supplemented with C-peptide. Furthermore, AICAR and metformin mimicked the preventive effect of C-peptide on high glucose–induced ROS generation, AMPK dephosphorylation, mitochondrial fission, ∆Ψ_m collapse, and endothelial cell apoptosis. Thus, it is likely that AMPKα is essential for C-peptide–mediated prevention of hyperglycemia-induced vascular complications.

It would be interesting to elucidate the mechanism by which C-peptide activates AMPKα. It is reported that C-peptide stimulates endothelial NO synthase (43). NO can act as an endogenous activator of AMPK in vascular endothelial cells (44). In the current study, C-peptide elevated the level of intracellular NO. Compound C had no significant effect on the C-peptide–induced formation of NO, whereas l-NAME inhibited the NO formation, indicating that AMPK is not involved in the C-peptide–stimulated NO production. C-peptide activated AMPKα at 48 h, and the activation was significantly inhibited by l-NAME. Additionally, S-nitroso-N-acetylpenicillamine recovered high glucose–induced AMPKα dephosphorylation. Thus, it is possible to propose that NO production can contribute to the C-peptide stimulation of AMPKα phosphorylation in endothelial cells; however, it is necessary to elucidate the mechanism by which early NO formation can induce the late activation of AMPKα in response to C-peptide.

The major upstream kinase activating AMPK is the tumor suppressor liver kinase B1 (LKB1), which is essential for the AMPK activation by AICAR (45). In our study, AICAR mimicked C-peptide by preventing high glucose–induced ROS generation, mitochondrial fission, ∆Ψ_m collapse, and cell death, suggesting that C-peptide activation of AMPKα might involve LKB1. Additionally, protein phosphatase 2C might be involved in the C-peptide activation of AMPK, since increase in intracellular AMP levels promotes AMPKα phosphorylation by inhibiting its dephosphorylation by the protein phosphatase 2C (46). However, Ca²⁺/calmodulin-dependent protein kinase kinase β is unlikely involved in the C-peptide activation of AMPK because C-peptide has no effect on the level of intracellular Ca²⁺ in endothelial cells (17).

In conclusion, our data indicate that cross-talk between PKC-dependent NADPH oxidase and mitochondrial ROS generation results in a positive feedback loop that amplifies intracellular ROS production in endothelial cells. Amplification of intracellular ROS generation mediates mitochondrial fission, which then leads to ∆Ψm collapse and apoptosis of endothelial cells in diabetes. Importantly, our data support C-peptide supplementation as a new therapeutic strategy for preventing diabetic vascular complications. C-peptide–mediated activation of AMPKα inhibited hyperglycemia-induced intracellular ROS production, mitochondrial fission, and endothelial cell apoptosis. Supplementation with C-peptide may offer increased benefit beyond the limits of currently available pharmacological agents, such as AICAR and metformin. Thus, C-peptide supplementation should be tested in combination with insulin therapy as a new strategy for preventing vascular complications in type 1 diabetes and late-stage type 2 diabetes.

This work was supported in part by Korea Research Foundation of Korea grants (2008-05943 and 2013-008193) and by the Ministry of Health and Welfare through the National R&D Program for Cancer Control (1020420).

No potential conflicts of interest relevant to this article were reported.

M.P.B. researched data and wrote the manuscript. Y.-C.L. researched data. Y.-M.K. contributed to discussion. K.-S.H. designed and supervised experiments and edited the manuscript. K.-S.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.