Expression of the mitochondrial calcium uniporter in cardiac myocytes improves impaired mitochondrial calcium handling and metabolism in simulated hyperglycemia

Abstract

Diabetic cardiomyopathy is associated with metabolic changes, including decreased glucose oxidation (Gox) and increased fatty acid oxidation (FAox), which result in cardiac energetic deficiency. Diabetic hyperglycemia is a pathophysiological mechanism that triggers multiple maladaptive phenomena. The mitochondrial Ca²⁺ uniporter (MCU) is the channel responsible for Ca²⁺ uptake in mitochondria, and free mitochondrial Ca²⁺ concentration ([Ca²⁺]_m) regulates mitochondrial metabolism. Experiments with cardiac myocytes (CM) exposed to simulated hyperglycemia revealed reduced [Ca²⁺]_m and MCU protein levels. Therefore, we investigated whether returning [Ca²⁺]_m to normal levels in CM by MCU expression could lead to normalization of Gox and FAox with no detrimental effects. Mouse neonatal CM were exposed for 72 h to normal glucose [5.5 mM glucose + 19.5 mM mannitol (NG)], high glucose [25 mM glucose (HG)], or HG + adenoviral MCU expression. Gox and FAox, [Ca²⁺]_m, MCU levels, pyruvate dehydrogenase (PDH) activity, oxidative stress, mitochondrial membrane potential, and apoptosis were assessed. [Ca²⁺]_m and MCU protein levels were reduced after 72 h of HG. Gox was decreased and FAox was increased in HG, PDH activity was decreased, phosphorylated PDH levels were increased, and mitochondrial membrane potential was reduced. MCU expression returned these parameters toward NG levels. Moreover, increased oxidative stress and apoptosis were reduced in HG by MCU expression. We also observed reduced MCU protein levels and [Ca²⁺]_m in hearts from type 1 diabetic mice. Thus we conclude that HG-induced metabolic alterations can be reversed by restoration of MCU levels, resulting in return of [Ca²⁺]_m to normal levels.

heart disease caused by diabetes mellitus (DM), including diabetic cardiomyopathy, is an important medical problem that is complex in origin and difficult to treat in a precise way. Outcomes after myocardial infarction are worse in patients with DM than in nondiabetic patients (30). Key features of diabetic cardiomyopathy are abnormal cardiac myocyte (CM) contraction and abnormal fuel flux with decreased glucose oxidation (Gox) and markedly increased fatty acid oxidation (FAox), which is linked to decreased energetic efficiency (7, 25, 39, 41, 44). The mechanisms that contribute to these metabolic abnormalities are incompletely investigated.

Diabetic hyperglycemia results in a number of pathophysiological changes in the vascular system, but investigations of its role in CM function are limited. Exposure of CM to elevated extracellular glucose results in impaired CM contractility and Ca²⁺ flux (38). These effects are related to increased protein O-GlcNAcylation caused by elevated glucose concentration (9, 23). In addition, hyperglycemia elicits an increase in reactive oxygen species production due to increased input of reducing equivalents into the mitochondrial electron transport chain (3, 32). We previously demonstrated that exposure of CM to excessive glucose concentrations results in reduced mitochondrial free Ca²⁺ concentration ([Ca²⁺]_m) (23, 46). Again, the mechanisms involved in this effect remain incompletely investigated.

Several studies have shown that mitochondrial Ca²⁺ uptake in the CM is decreased in DM (15, 43, 49). However, the contribution of decreased [Ca²⁺]_m to decreased cardiac function in DM is unclear. The influence of [Ca²⁺]_m on the activity of specific mitochondrion-based dehydrogenases is well established (13, 14, 27). Nevertheless, the detailed mechanisms regulating mitochondrial Ca²⁺ handling are only recently being elucidated (2, 37). Recently, genes coding for the proteins of the mitochondrial Ca²⁺ uniporter (MCU) complex (MCUC), such as MCU, have been identified (2, 37). MCU is an integral membrane pore protein with two transmembrane domains that conducts mitochondrial Ca²⁺ currents (2, 12). Prior work using cell culture demonstrated that an increase in [Ca²⁺]_m stimulates the activity of three mitochondrial dehydrogenases (13, 14, 17, 27), including the pyruvate dehydrogenase (PDH) complex (PDC), which mediates Gox (42). Positive effects of [Ca²⁺]_m on mitochondrial complex III and V function and ATP production have also been reported (1, 11, 50). PDC activity mediates cardiac Gox, which is diminished in DM hearts, in contrast to the increase in FAox (5).

In the present study we investigated whether the high glucose (HG)-induced decrease in [Ca²⁺]_m was due to impairment of mitochondrial Ca²⁺ uptake. In addition, we investigated the role of MCU and whether reestablishment of mitochondrial Ca²⁺ uptake and [Ca²⁺]_m by MCU transgene expression could modulate fuel flux in neonatal mouse CM exposed to HG. Furthermore, we performed an analysis of MCU protein levels and [Ca²⁺]_m in hearts from type 1 diabetic (T1D) mice that confirmed data obtained in primary cultures.

MATERIALS AND METHODS

All animal protocols were approved by the University of California, San Diego, Institutional Animal Care and Use Committee and conform to the Guide for the Care and Use of Laboratory Animals, as outlined by the National Institutes of Health.

T1D mouse model.

T1D was induced by streptozotocin in C57BL/6NHsd mice as previously described (16, 21, 48, 52). Streptozotocin (50 mg/kg body wt ip) was injected for 5 consecutive days. T1D is documented by measured fasting glucose >200 mg/dl. These mice were nonketotic and euthyroid, as previously reported (53).

CM culture and adenoviral transduction.

Mouse neonatal CM were isolated using the Worthington neonatal CM isolation system (Worthington Biochemical, Lakewood, NJ) following the manufacturer's instructions. Cells were plated onto gelatin-coated culture dishes or glass chamber slides. Culture medium consisted of 4:1 Dulbecco's modified Eagle's medium (DMEM)-M199, 10% horse serum, 5% fetal bovine serum, and 1% penicillin-streptomycin-amphotericin B (Fungizone). Final glucose concentration for plating was 20 mM. After 24 h in culture, myocytes were treated with different glucose concentrations. CM were transduced with MCU at multiplicity of infection (MOI) of 5 and empty (control) viruses. Only for mitochondrial Ca²⁺ measurements, CM were transduced with Adv-mpericam. Studies were performed 48 h after viral transduction. Combination of Adv-MCU with Adv-mpericam does not alter expression of MCU protein levels (not shown).

CM exposed to simulated hyperglycemia.

Cells were allowed to adhere to the plates for 24 h before change to basic experimental culture medium (4.5:1 DMEM-M199, 2% fetal bovine serum, and 1% penicillin-streptomycin-amphotericin B) supplemented with glucose at physiological concentration [normal glucose (NG): 5.5 mM glucose + 19.5 mM mannitol] or elevated glucose concentration (HG: 25 mM glucose) and maintained in these conditions for 72 h. Viral transduction was performed 24 h after HG exposure.

Construction of adenoviral vectors.

We cloned a mouse gene coding for MCU (GenBank accession no. NM 001033259) into a replication-deficient adenoviral vector under control of the promoter-enhancer region of the human cytomegalovirus (Adv-MCU). An empty adenovirus without transgene (Adv-control) was used in the control groups. This approach is described elsewhere (45).

Mitochondrial Ca²⁺ measurement.

Mitochondrial pericam has been constructed and described by Nagai et al. (31). We used the ratiometric pericam, which has a bimodal excitation spectrum peaking at 415 and 494 nm. Pericam was targeted specifically to the mitochondrial matrix (mitochondrial pericam) by fusion to a cytochrome oxidase sequence. We cloned the gene coding for mitochondrial pericam into a replication-deficient adenoviral vector under control of the promoter-enhancer region of the human cytomegalovirus (Adv-CMVmpericam). Cells were transduced with Adv-CMVmpericam at MOI of 20. For mitochondrial Ca²⁺ measurements in adult CM from T1D mice, we cloned mitochondrial pericam into an AAV9.45 backbone (AAV-MitoPericam). AAV-MitoPericam was injected into the jugular vein in different groups of mice to express this mitochondrial Ca²⁺ sensor in vivo. The transgene was expressed for ≥4 wk before measurement of mitochondrial Ca²⁺ in isolated adult CM.

Pericam fluorescence.

The pericam fluorescence method is described elsewhere (47). Briefly, CM cultured on a chambered coverglass were perfused with HEPES-buffered medium containing 1.8 mM Ca²⁺ at room temperature. Chambers were mounted in a Nikon Diaphot epifluorescence microscope equipped with a ×100 Fluor (oil-immersion) objective interfaced to a dual-excitation (E1 and E2) lamp system (Photon Technologies International, Birmingham, NJ) set at 410 nm for E1 and 485 nm for E2. Fluorescence emission was filtered at 510 nm and directed to a photomultiplier tube. Additionally, an aperture mechanism allowed collection of fluorescence from a selected portion of the field that was always positioned over the peripheral, nonnuclear, regions of individual cells. Data were collected from the emission channel at a rate of 20 Hz, and the ratio of intensity of excitation at 410 nm to intensity of excitation at 485 nm (410/485) was calculated to provide relative comparisons of [Ca²⁺]_m between experimental treatments. [Ca²⁺]_m was determined as described elsewhere (19). For calibration of the mitochondrial pericam signal, ionomycin in Ca²⁺-free Tyrode solution with EGTA was used to obtain the minimum ratio (R_min) and digitonin in Tyrode solution with 2 mM Ca²⁺ was used to obtain the maximum ratio (R_max). [Ca²⁺]_m was determined as described elsewhere (19).

Analysis of mitochondrial Ca²⁺ transients.

Basal [Ca²⁺]_m was obtained from paced myocytes by averaging diastolic and systolic [Ca²⁺]_m, which are the minimum and maximum pericam ratios, respectively, that were measured during contraction. Ca²⁺ uptake is the maximum slope of the upstroke. Ca²⁺ release is the maximum slope during Ca²⁺ decline. We were able to inhibit the upstroke of the transient with Ru360 [MCU-specific inhibitor (29)], confirming that it corresponds to MCU-mediated mitochondrial Ca²⁺ uptake (not shown).

Western blot analysis.

Heart tissue or CM were homogenized with a Polytron homogenizer in lysis buffer. Protein content was measured by Bradford assay (Bio-Rad) and adjusted for equal loading. Protein extracts (20 μg) were separated on 4–12% Bis-Tris-HCl-buffered polyacrylamide gels (Invitrogen, Carlsbad, CA) and subjected to Western blotting. Antibodies for MCU, MCUb, MICU1, MICU2, and MCUR1 were purchased from Sigma-Aldrich (St. Louis, MO); essential MCU regulator (EMRE), PDH E1α, and actin from Santa Cruz Biotechnology (Santa Cruz, CA); and phosphorylated PDH E1α (S293) antibody from Abcam (Cambridge, MA). The blots were also probed with a mouse monoclonal porin antibody (voltage-dependent anion channel; Cell Signaling Technology, Danvers, MA) or actin antibody (Santa Cruz Biotechnology) as internal control to ensure equivalent protein loading and protein integrity. Horseradish peroxidase-conjugated secondary antibodies were used. Signals on the films were digitized on a 350 dots-per-inch scanner and analyzed using ImageJ (National Institutes of Health).

Immunoblot detection of overall protein O-GlcNAcylation.

We used a method described elsewhere for immunoblot detection of overall protein O-GlcNAcylation (22, 23). O-GlcNAc modification of proteins was detected in Western blots with the antibody RL2 (Thermo Fisher Scientific) raised against separate specific O-GlcNAc-dependent epitopes (20). To control for changes in total protein, we stained the membrane with Ponceau S.

MCU promoter activity.

MCU promoter activity was measured with a promoter-luciferase reporter construct. Briefly, the −2 kb region upstream of the MCU translation initiation codon was amplified from wild-type adult mouse heart cDNA using KOD polymerase (EMD Millipore) and cloned in front of the modified firefly luciferase gene in pGL3-basic (Promega). The expression cassette was removed by restriction digestion and inserted between the EcoRI sites of pENTR1a. LR Clonase II (Thermo Fisher) was used to recombine the resulting vector with pAd/CMV/V5-DEST, which was transfected into 293A cells to produce adenovirus (Adv-MCU-pro/Luc). CM were transduced with Adv-MCU-pro/Luc and treated with NG and HG. At 72 h posttransduction, cells were harvested and used for luciferase assay, as previously reported (18).

Metabolic fuel flux.

Fuel flux was determined in CM as described previously (6, 34). CM were incubated for 2 h at 37°C in serum-free medium in sealed petri dishes with the radiolabeled substrates. ¹⁴CO₂ produced during the incubation period was trapped in hyamine hydroxide (Fisher Scientific). The reaction was terminated with H₂SO₄ after 2 h of incubation and further incubated for 1 h at 37°C. Hyamine solution was transferred to scintillation vials for radioactivity counting. Pilot experiments were performed to confirm that 2 h of incubation was in the linear portion of the assay, as reported by others (10). Gox was determined from ¹⁴CO₂ released from [U-¹⁴C]glucose. Final glucose concentration during the assay was 5 mM. FAox was measured from the amount of ¹⁴CO₂ released from [1-¹⁴C]palmitate. Final palmitate concentration was 0.15 mM.

Measurement of PDC activity.

PDC activity was measured as described previously (54). PDC activity was estimated from the amount of ¹⁴CO₂ released from [1-¹⁴C]pyruvic acid. Final pyruvic acid concentration was 20 mM.

8-Hydroxy-2′-deoxyguanosine immunofluorescence microscopy.

After treatment, cells were washed once with 1× PBS and fixed in 4% paraformaldehyde for 15 min at room temperature. Cells were then rinsed once in 1× PBS and incubated for 30 min at room temperature on a shaking platform in blocking-and-permeabilization buffer [20 mM glycine, 1% BSA (IgG-free), 3% normal goat serum (the same species as the secondary antibody), 0.1% Triton X-100, 0.05% Tween 20, and 1× PBS] before incubation overnight at 4°C with anti-8-hydroxy-2′-deoxyguanosine (8-OHdG; product no. GTX41980, Genetex) diluted 1:20 in 1:10 blocking-and-permeabilization buffer. As control for nonspecific detection of antigens by the secondary antibody, cells were incubated with only 1:10 blocking-and-permeabilization buffer. On the following day, cells were rinsed three times with 1× PBS-0.05% Tween 20 and then incubated at room temperature for 1 h with goat anti-mouse IgG (H+L) secondary antibody-Alexa Fluor 568 conjugate (catalog no. A11004, Thermo Fisher Scientific) diluted 1:200 in 1:10 blocking-and-permeabilization buffer. Cells were then rinsed five times with 1× PBS-0.05% Tween 20. One drop of mounting medium with 4′,6-diaminido-2-phenylindole (ProLong Diamond Antifade Mountant with DAPI, Thermo Fisher Scientific) was added to the cells, cover glasses were mounted, and edges were sealed with a nonfluorescent nail polish. Images were captured with a Delta Vision deconvolution microscope system (Applied Precision) at the University of California, San Diego, School of Medicine Light Microscopy Facility. A ×100 lens was used to acquire images of ∼20 serial optical sections, spaced 0.2 μm apart. The data sets were deconvolved using SoftWorx software (Applied Precision) on a Silicon Graphics Octane workstation. 8-OHdG levels were quantified as the number of positive (red) pixels as a percentage of the number of total pixels using the “Colour Pixel Counter” plug-in in ImageJ. Data were normalized by subtraction of the nonspecific signal detected in samples incubated with no primary antibody.

Terminal deoxynucleotidyl transferase dUTP nick end-labeling assay.

DNA fragmentation was assessed as a marker of apoptosis by the terminal deoxynucleotidyl transferase dUTP nick end-labeling (TUNEL) assay (TUNEL Apoptosis Detection Kit, EMD Millipore) following the manufacturer's instructions. Images were captured with a fluorescence microscope (model DM2500, Leica) equipped with a ×40 lens. Images were acquired with the software Leica Application Suite X Advanced Fluorescence 1.0.0.12269, and positive-stained nuclei were counted with ImageJ.

Mitochondrial membrane potential measurement.

Mitochondrial membrane potential (Δψ_m) was measured by JC-1, as described previously (28). CM were loaded with 0.5 mM JC-1 at 37°C for 15 min and washed three times with medium. The cells were visualized under a Nikon Diaphot epifluorescence microscope equipped with a ×40 Fluor objective interfaced to a Photon Technologies dual-emission system, with the excitation wavelength set to 485 nm via a monochromator. Fluorescence emission was split and directed to two photomultiplier tubes through 20-nm band-pass filters centered at 531 and 584 nm, respectively. In addition, an aperture mechanism allowed collection of fluorescence from a selected area, which was always positioned over the cytoplasmic region of individual cells. Data are shown as the ratio of intensity of red (F₅₈₄) to green (F₅₃₁) fluorescence (F₅₈₄/F₅₃₁). Loss of Δψ_m is indicated by a decrease in F₅₈₄/F₅₃₁.

Statistical analysis.

Values are means ± SE from at least three different experiments. Comparisons between means were analyzed, as appropriate, by Student's t-test or one-way ANOVA followed by Tukey's post hoc test. P < 0.05 was considered statistically significant.

RESULTS

Reduced MCU protein levels and impaired mitochondrial Ca²⁺ handling in simulated hyperglycemia are restored by adenoviral MCU expression.

Since mitochondrial Ca²⁺ uptake is mainly mediated by MCUC, we proceeded to analyze the effect of HG on proteins that constitute the MCUC. In HG-treated CM, MCU protein levels were reduced 49% (Fig. 1A). MCU is the pore-forming unit of the MCUC. In contrast, other members of the MCUC were not affected by HG (Fig. 1B). Since the relative MCU gene expression was reduced 50% in HG-exposed CM (Fig. 2A), we sought to investigate if HG affects MCU promoter activity. Our results demonstrated that HG exposure inhibits MCU promoter activity by 47% (Fig. 2B). Subsequently, we used adenoviral gene transfer to increase MCU expression in HG-treated CM. Adv-MCU restored MCU protein levels toward NG levels (Fig. 3A). [Ca²⁺]_m was decreased 47% in CM exposed for 72 h to HG, which exceeds the physiological concentration of 5.5 mM (NG; Fig. 3B). Mitochondrial Ca²⁺ uptake and release were also decreased 44% and 39%, respectively, after HG treatment (Fig. 3B). MCU expression increased mitochondrial Ca²⁺ uptake and release toward NG levels in HG-exposed CM (Fig. 3B). Consequently, [Ca²⁺]_m returned to NG levels after MCU expression (Fig. 3B).

Fig. 1.

Influence of high-glucose (HG) treatment on expression of proteins in the mitochondrial Ca²⁺ uniporter (MCU) complex (MCUC). A: Western blot and densitometry data analysis showing significantly decreased MCU in myocytes exposed to 25 mM glucose (HG) vs. 5.5 mM glucose + 19.5 mM mannitol [normal glucose (NG)]. Values are means ± SE; n = 4 per group. *P < 0.05 vs. NG. B: Western blot showing that other members of the MCUC were not significantly affected by HG. Control for protein loading was confirmed with voltage-dependent anion channel 1 (VDAC1). AU, arbitrary units.

Fig. 2.

Simulated hyperglycemia decreased MCU gene transcriptional activity. A: relative MCU gene expression is reduced in HG. B: MCU promoter-luciferase reporter assay showing decreased MCU gene transcriptional activity. Cardiac myocytes (CM) were transduced with Adv-MCU-pro/Luc and treated with NG and HG. Cells were harvested at 72 h posttransduction and used for luciferase assay (see materials and methods). Values are means ± SE; n = 6 per group. *P < 0.01 vs. NG. RLU, relative light units.

Fig. 3.

MCU expression improves mitochondrial Ca²⁺ handling. A: Western blot and densitometry data analysis showing effect of Adv-MCU on MCU protein levels in HG (HG + MCU). MCU protein levels were restored by Adv-MCU expression in CM exposed to HG. Values are means ± SE; n = 3 per group. *P < 0.05 vs. NG and HG + MCU. B: MCU expression increased mitochondrial free Ca²⁺ concentration ([Ca²⁺]_m), mitochondrial Ca²⁺ uptake, and mitochondrial Ca²⁺ release in CM despite HG. Values are means ± SE; n = 25 cells per group. *P < 0.05 vs. NG and HG + MCU.

Reestablishing [Ca²⁺]_m improves Gox and FAox and reduces protein O-GlcNAcylation, despite simulated hyperglycemia.

To investigate if the increase in [Ca²⁺]_m after MCU expression affects mitochondrial metabolism, we used [¹⁴C]glucose as substrate for Gox and [¹⁴C]palmitate as substrate for FAox and measured ¹⁴CO₂ production. Gox was reduced 51% in HG-treated CM and returned to normal levels by MCU expression (Fig. 4A). FAox was increased 39% after HG treatment and returned to normal levels by MCU expression (Fig. 4B). We evaluated PDC activity by measuring ¹⁴CO₂ production using [¹⁴C]pyruvate as substrate. PDH activity was reduced 36% in HG (Fig. 5A). MCU expression returned PDH activity toward normal levels (Fig. 5A). Furthermore, the ratio of phosphorylated inactive PDH-to-total PDH was increased fourfold in HG, and MCU expression returned this value toward NG levels (Fig. 5B).

Fig. 4.

MCU expression restores glucose and fatty acid oxidation in HG-treated CM. A: diminished glucose oxidation level in HG was restored after Adv-MCU expression. B: increased fatty acid oxidation level in HG was restored to normal after Adv-MCU expression. Values are means ± SE; n = 9 per group from 3 experiments. *P < 0.05 vs. NG and HG + MCU.

Fig. 5.

MCU expression improves pyruvate dehydrogenase (PDH) activity. A: decreased PDH activity in HG was returned toward normal after Adv-MCU expression. PDH activity is estimated as the rate of ¹⁴CO₂ formation after pyruvate oxidation. Values are means ± SE; n = 9 from 3 separate experiments. *P < 0.05 vs. NG and HG + MCU. B: Western blot and densitometry data analysis showing that MCU restoration in CM in HG returns increased PDH phosphorylation to normal. Total PDH protein levels remain unchanged. Black lane in blot indicates that NG bands from the same blot were cut and aligned for comparison with HG and HG + MCU. Values are means ± SE; n = 4. *P < 0.05 vs. NG and HG + MCU. p-PDH, phosphorylated PDH.

Since MCU expression improved Gox, we hypothesized that less glucose could follow the hexosamine biosynthetic pathway (HBP), reducing protein O-GlcNAcylation in the CM. Therefore, we analyzed protein O-GlcNAcylation in our experiments. O-GlcNAcylation levels were increased in proteins of CM exposed to HG. As we expected, MCU expression reduced O-GlcNAcylation despite HG (Fig. 6).

Fig. 6.

Influence of MCU expression on HG-induced increase in protein O-GlcNAcylation. Overall protein O-GlcNAcylation analyzed with RL2 antibody was more prominent in CM exposed to HG than HG + MCU. MCU expression decreased O-GlcNAcylation despite HG. The same blot was stained by Ponceau S to show uniform protein loading. Values are means ± SE; n = 3. *P < 0.05 vs. NG and HG + MCU. **P < 0.05 vs. NG and HG.

MCU expression reduces the oxidative stress DNA damage and apoptosis despite hyperglycemia.

CM exposed to HG show increased levels of the oxidative stress biomarker 8-OHdG compared with CM in NG (Fig. 7A). MCU expression reduced 8-OHdG levels despite HG (Fig. 7A). In Fig. 7B, TUNEL assay shows more TUNEL-positive cells in CM treated with HG, indicating increased apoptotic cell death. MCU expression significantly reduced the number of TUNEL-positive cells despite HG (Fig. 7B). It has been reported that MCU overexpression sensitizes HeLa cells to apoptotic stimuli (12). Therefore, we investigated whether higher MCU protein levels could increase HG-induced apoptosis in CM. Adv-MCU at MOI of 20 was used, and MCU protein expression levels increased by 3.4-fold in HG (Fig. 8A). The number of TUNEL-positive cells after MCU overexpression was similar to that after HG exposure. We did not observe a synergistic effect of MCU overexpression on HG-induced apoptosis; however, the protective effect at low MCU expression levels was obliterated (Fig. 8B).

Fig. 7.

MCU expression reduces the oxidative stress marker 8-hydroxy-2′-deoxyguanosine (8-OHdG) in HG and reduces HG-induced apoptosis. A: 8-OHdG levels in CM treated for 72 h in NG (5.5 mM glucose + 19.5 mM mannitol) or HG (25 mM glucose) and transduced with Adv-MCU or Adv-control (empty). Values are means ± SE from 3 independent experiments. *P < 0.05 vs. NG and HG + MCU. B: analysis of images of CM stained for TUNEL assay. TUNEL staining was quantified using ImageJ as the number of positive nuclei as a percentage of the total number of nuclei. Values are means ± SE from 3 independent experiments. *P < 0.05 vs. NG and HG + MCU.

Fig. 8.

Influence of MCU overexpression on HG-induced apoptosis. A: Western blot and densitometry data analysis showing MCU protein levels in NG, HG, and HG + Adv-MCU at MOI of 20, which is 4 times the amount needed to normalize MCU expression levels in HG. Data show a 3.4-fold increase in MCU protein levels. Values are means ± SE; n = 4. *P < 0.05 vs. NG and HG + MCU. **P < 0.05 vs. NG and HG. B: analysis of images of CM stained for TUNEL assay shows similar TUNEL-positive cells in HG and HG + MCU. TUNEL staining was quantified using ImageJ as the number of positive nuclei as a percentage of the total number of nuclei. Values are means ± SE from 3 independent experiments. *P < 0.05 vs. NG.

Δψ_m is improved by MCU expression.

Mitochondrial dysfunction caused by HG concurs with decreased Δψ_m. Consistent with this, our results showed decreased Δψ_m in CM exposed to HG. In contrast, MCU expression improved Δψ_m in CM regardless of HG exposure (Fig. 9).

Fig. 9.

MCU expression improves mitochondrial membrane potential. Cells were loaded with JC-1 for analysis of membrane potential. Relative mitochondrial membrane potential reflects red-to-green fluorescence ratio. Values are means ± SE; n ≥ 30 cells per group from 3 independent experiments. *P < 0.05 vs. NG and HG + MCU.

MCU protein levels and [Ca²⁺]_m are reduced in hearts from diabetic mice.

To investigate if our findings in cultured CM subjected to simulated hyperglycemia are relevant to DM, hearts from T1D mice were analyzed for MCU protein levels. Figure 10A shows that MCU was markedly reduced by 60% in hearts of T1D mice. Furthermore, [Ca²⁺]_m in CM of T1D hearts was diminished compared with control values (Fig. 10B).

Fig. 10.

Diabetes mellitus leads to reduced MCU protein expression and diminished [Ca²⁺]_m in the heart. A: Western blot and densitometry data analysis showing MCU protein levels in hearts of 8-wk-old type 1 diabetic (T1D) mice. Values are means ± SE; n = 4 per group. *P < 0.05 vs. control. B: [Ca²⁺]_m in CM from control and T1D mice. Values are means ± SE; n = 30 cells per group from 3 mice. *P < 0.05 vs. control.

DISCUSSION

Several pathophysiological consequences in DM can be attributed to hyperglycemia. Our studies in isolated neonatal CM show that impaired mitochondrial Ca²⁺ handling is an important consequence. We have demonstrated for the first time that excessive glucose concentrations, simulating the levels in diabetic animals or humans, inhibited MCU expression and reduced mitochondrial Ca²⁺ uptake and release and [Ca²⁺]_m. In addition, expression of MCU restored these parameters to the normal range, despite elevated glucose concentrations. Furthermore, Gox and FAox were impaired in CM exposed to hyperglycemia and improved by MCU transgene expression. The effects of MCU expression were associated with reduction of protein O-GlcNAcylation levels, oxidative stress damage, and apoptosis. In addition, reduced Δψ_m in HG was improved by MCU expression. Our results in cultured CM are relevant to DM, since we demonstrated decreased MCU protein levels in T1D hearts as well as diminished [Ca²⁺]_m in CM of T1D mice.

Mitochondria accumulate significant amounts of Ca²⁺ from the cytosol. However, [Ca²⁺]_m is tightly regulated and modulates vital physiological mitochondrial functions, including oxidative phosphorylation, fuel metabolism, and cell death. Mitochondrial Ca²⁺ uptake is carried out by the MCU, a highly selective Ca²⁺ channel that moves Ca²⁺ across the mitochondrial inner membrane, a process that is dependent on Δψ_m (24). While MCU physiology has been studied for decades, a complete description of its molecular composition has remained elusive. Only recently, integrative genomics methods enabled discovery of the uniporter pore MCU and its regulatory subunits MCUb, MICU1, MICU2, MCUR1, and EMRE (2, 12, 35, 36, 40, 51). MCU is an integral membrane pore protein with two transmembrane domains that conducts mitochondrial Ca²⁺ currents (2, 8). Constitutive MCU knockout mice are viable and show decreased energy efficiency and maximal performance of skeletal muscle (33). Although DM may influence several members of the MCUC, our results suggest that the effect of excessive glucose decreasing mitochondrial Ca²⁺ is the result of a reduced level of MCU, the pore-forming subunit of the MCUC. We observed reduced mitochondrial Ca²⁺ uptake after HG exposure that can be explained by reduced MCU protein levels; however, we cannot rule out an effect of HG on MCU activity or the activity of other MCUC members.

Numerous studies have shown that mitochondrial Ca²⁺ uptake is decreased in DM (15, 43, 49), yet the contribution of decreased CM mitochondrial Ca²⁺ levels to decreased cardiac function in DM is unclear. We hypothesize that the levels of proteins in the MCUC will play an important role in this process. Prior work has demonstrated that an increase in mitochondrial Ca²⁺ stimulates the activity of mitochondrial dehydrogenases (13, 14, 17, 27, 42). Positive effects of mitochondrial Ca²⁺ on mitochondrial complex III and V function and ATP production also occur (1, 11, 50). PDH activity mediates cardiac Gox, which is diminished in DM hearts, in contrast to FAox, which is increased. Our results are consistent with these reports. In HG-exposed CM, we found reduced Gox, increased FAox, and reduced PDH activity, possibly due to diminished [Ca²⁺]_m. In support of this idea, MCU expression returned Ca²⁺ toward normal levels and restored Gox and the ratio of phosphorylated PDH to total PDH. In addition, FAox was reduced toward normal levels. We used palmitate to analyze FAox. Whether other different fatty acids or a combination of different substrates would lead to the same effects needs further investigation. MCU effects on Gox may be the result of the activator effect of [Ca²⁺]_m on PDH and, via PDH dephosphorylation, by activation of the PDH phosphatase (13). Consistent with this idea, we observed reduced PDH activity in HG associated with increased phosphorylated PDH levels. MCU expression returned these parameters toward NG values. Our results further support the idea of [Ca²⁺]_m as a metabolic regulator in CM and the role of abnormal mitochondrial Ca²⁺ handling in glucose toxicity that may be a pathophysiological mechanism in DM.

We have shown that hyperglycemia activates the HBP, leading to increased protein O-GlcNAcylation in the diabetic heart and HG-exposed CM (9, 22, 23). Since MCU expression improved Gox, we hypothesized that less glucose could follow the HBP, resulting in reduced protein O-GlcNAcylation in the CM. Our results confirm this idea, suggesting that not only hyperglycemia, but also impaired Gox, contributes to increase O-GlcNAcylation.

Previous work from our laboratory demonstrated that HG treatment reduced [Ca²⁺]_m in CM and that this effect was reversed by reducing protein O-GlcNAcylation or expressing transcription factor A (23, 46). Whether MCU is subjected to O-GlcNAcylation that modulates its activity is unknown. Regulation of MCU expression or activity under pathological states needs to be further investigated.

Simulated hyperglycemia induces apoptosis by Ca²⁺ overload and oxidative stress in cardiac cells (4, 26). Moreover, increased mitochondrial Ca²⁺ uptake by MCU overexpression sensitizes HeLa cells to apoptotic stimuli (12). We tested whether MCU expression could worsen HG-induced apoptosis in our experiments. We were surprised to find that restoration of MCU protein toward NG levels had a protective effect, reducing the number of apoptotic cells and decreasing the oxidative stress marker 8-OHdG. In addition, the antiapoptotic MCU effect disappeared with MCU overexpression (3.4-fold increase). We did not observe a synergistic effect of HG-induced apoptosis and MCU overexpression, possibly because HG is a less aggressive apoptotic stimulus than C2 ceramide or H₂O₂, which were used in HeLa cells (12).

Mitochondrial Ca²⁺ uptake is decreased in DM (15, 43, 49). Consistent with these results, we observed decreased MCU protein levels in hearts from T1D mice. In addition, CM from T1D mice showed reduced [Ca²⁺]_m. Further investigation is needed to determine whether restoration of MCU protein levels will reestablish mitochondrial Ca²⁺ handling and improve mitochondrial function and cardiac performance.

Our results demonstrate that exposure of CM to HG results in decreased mitochondrial Ca²⁺ levels due to decreased MCU protein levels. In addition, we observed decreased Gox and increased FAox in HG. MCU transgene expression increased mitochondrial Ca²⁺ uptake, [Ca²⁺]_m, and PDH activity and normalized fuel flux and protein O-GlcNAcylation with no detrimental effects. Altered mitochondrial Ca²⁺ handling with reduced MCU expression could be among the mechanisms that lead to mitochondrial dysfunction in DM.

GRANTS

This work was supported by National Institutes of Health Grant 5 P01 HL-066941-13, Department of Veterans Affairs Merit Review Award 5 I01BX001121-02, National Institute of Neurological Disorders and Stroke Grant NS-047101 to the University of California, San Diego, Neuroscience Microscopy Facility, and the P. Robert Majumder Charitable Foundation. J. Diaz-Juarez received support from UCMEXUS-CONACYT (CN 15-1489).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.D.-J., J.S., and W.H.D. developed the concept and designed the research; J.D.-J., J.S., F.C., B.T.S., T.D., and A.D. performed the experiments; J.D.-J., J.S., F.C., B.T.S., and T.D. analyzed the data; J.D.-J., J.S., F.C., B.T.S., and W.H.D. interpreted the results of the experiments; J.D.-J., J.S., and F.C. prepared the figures; J.D.-J. and J.S. drafted the manuscript; J.D.-J., J.S., F.C., B.T.S., and W.H.D. edited and revised the manuscript; J.D.-J., J.S., F.C., B.T.S., T.D., A.D., and W.H.D. approved the final version of the manuscript.