GSH or Palmitate Preserves Mitochondrial Energetic/Redox Balance, Preventing Mechanical Dysfunction in Metabolically Challenged Myocytes/Hearts From Type 2 Diabetic Mice

Male C57BLKS/J-lepr^db/lepr^db diabetic (db/db; stock number 000642) and nondiabetic C57BLKS/J-lepr⁺/lepr⁺ (+/+; stock number 000662) mice, 8–10 weeks of age, were obtained from The Jackson Laboratory (Bar Harbor, ME) and housed ad libitum with food and water. For all experiments, animals 10–12 weeks of age were used when db/db mice were severely hyperglycemic (28). At the time the mice were killed, wild-type (WT) versus db/db body mass (g), heart weight (mg), and blood glucose levels were on average 25 ± 0.6 versus 44 ± 1.4; 183 ± 10 versus 148 ± 6; and 220 ± 9 versus 523 ± 12 mg/dL, respectively. All procedures involving the handling of animals were approved by the Animal Care and Use Committee of the Johns Hopkins University and the Mount Sinai Hospital and adhered to the National Institutes of Health Public Health Service guidelines.

Isolation and handling of mitochondria were performed as described previously (29). In mouse heart mitochondria, respiratory control ratios (RCRs; state 3/state 4) of 5 or higher were obtained.

Respiration was assayed in freshly isolated mitochondria with a high-throughput automated 96-well extracellular flux analyzer (Seahorse XF96; Seahorse Bioscience, Billerica, MA) as described in detail elsewhere (30). In brief, a medium containing (in mmol/L) 137 KCl, 2 KH₂PO₄, 0.5 EGTA, 2.5 MgCl₂, 20 HEPES at pH 7.2 and 37°C, with 0.2% FA-free BSA was used. Mitochondria were assayed in polyethyleneimine-coated XF96 plates. After removing the solution of polyethyleneimine, the equivalent of 5–15 μg mitochondrial protein was transferred to each well. The status of the different complexes of the respiratory chain was evaluated with substrates of complex I, II, and IV. Maximal respiratory capacity was analyzed in the presence of 1 mmol/L ADP and 50 μmol/L dinitrophenol. Mitochondrial protein was determined using the bicinchoninic acid method, BCA protein assay kit (Thermo Scientific, Barrington, IL). NADPH, ΔΨ_m, and H₂O₂ were determined as described elsewhere (29), using the same medium as above for measuring respiration (excluding BSA).

The same batch of freshly isolated cardiomyocytes was assayed in parallel for contractility and two-photon imaging.

Experiments with intact cardiomyocytes were performed at 37°C in a thermostatically controlled flow chamber mounted on the stage of an upright microscope (Nikon E600FN) attached to a multiphoton laser scanning system, as described previously (31). Cells were loaded with fluorescent probes and imaged as previously described (29,31) (Fig. 2, legend).

Myocytes were isolated as previously described (32). After isolation, cells were placed in a perfusion chamber with a flow-through rate of 2 mL/min. Sarcomere shortening was assessed by real-time imaging, and whole-cell Ca²⁺ transients by Fura 2-AM fluorescence using an inverted fluorescence microscope (Nikon TE2000) and IonOptix (Myocam) software (32).

The functional activity of the thioredoxin reductase (TrxR2)/thioredoxin (Trx2) system was monitored in intact mitochondria through the redox status of Trx2 (33). To determine the proportion of reduced and oxidized Trx2, a redox Western blot was performed as described previously (33,34). The redox potential of the Trx2 redox couple was quantified through the Nernst potential as previously described (33).

After anesthesia, hearts were harvested and the aorta cannulated and retrogradely perfused with Krebs-Henseleit (KH) buffer warmed and gassed with 95% O₂ and 5% CO₂. Hearts were paced at 600 bpm (10 Hz, 4-ms duration, 4 V), and left ventricular (LV) function was monitored with a water-filled balloon connected to a pressure transducer coupled to a continuous data recording system. After stabilization, in addition to LV function parameters, baseline values of coronary perfusion pressure (CPP) were obtained through a side arm of the aortic cannula connected to a separate pressure transducer.

db/db and WT mice, 8–10 weeks of age (n = 13), of either sex were killed, and their hearts were rapidly excised, cannulated, perfused, and placed in a custom-designed imaging chamber. Perfusion pressure was maintained at ∼60 mmHg and temperature at 36.5 ± 1°C throughout the entire protocol. Unipolar electrocardiograms were recorded for rhythm analysis using silver electrodes (Grass Instruments). Hearts were stained with di-4-ANEPPS (20 μmol/L) for 5 min, allowing the high-resolution mapping of optical action potentials (APs), and therefore, the measurement of epicardial AP duration (APD), normalized upstroke velocity, and conduction velocity (CV) during steady-state pacing, as previously described (35,36).

Data were analyzed with the software GraphPad Prism (version 3; San Diego, CA) or MicroCal Origin. The statistical significance of the differences between groups and treatments was evaluated with two-way ANOVA using Tukey multiple comparison test or with a Student t test (small samples, paired t test with two-tailed P values), and the results were presented as mean ± SEM (95% CI).

Are changes in redox status occurring in WT and db/db myocytes at baseline and after metabolic and redox stress? To answer this question, WT and db/db myocytes were loaded with ROS or GSH sensors and imaged with two-photon laser scanning fluorescence microscopy (Fig. 2A). We compared NADH, GSH, and normalized ROS levels between groups at baseline and during exposure to stress produced by HG (30 mmol/L glucose) and HG + β-agonist isoproterenol (ISO, 10 nmol/L), with or without the FA palmitate (Palm). At baseline, NADH and GSH levels were comparable between WT and db/db myocytes (Gluc 5) (Fig. 2B and C), and exposing myocytes to HG or HG + ISO determined a progressive GSH oxidation (Fig. 2C) in both groups, without affecting NADH (Fig. 2B). However, stressing db/db myocytes either with HG or HG + ISO led to a more prominent decay of GSH levels compared with WT (Fig. 2C), with higher MitoSOX fluorescence (Fig. 2D). In fact, during exposure to HG + ISO, O₂^.− production in db/db myocytes was approximately twofold greater than that in WT cells (Fig. 2D), with more elevated, although nonsignificant, CM-DCF fluorescence in both WT and db/db myocytes (Fig. 2E).

To establish whether changes in substrate availability/utilization by the heart can affect cellular/mitochondrial redox balance, we used Palm in the range of serum concentrations found in type 2 diabetes (37). This FA restored redox balance in db/db cardiomyocytes (Fig. 2) in a concentration-independent manner (Supplementary Fig. 1). Under any of the conditions tested, Palm efficiently countered exaggerated ROS emission (15–60-fold, P < 0.001), increasing GSH levels (20–40%, P < 0.001) in both WT and db/db. NADH remained nearly all reduced (∼93%) (Fig. 2B). Together, these data indicate that stress conditions, such as the HG + ISO regimen, unleash the propensity of the db/db heart to build up oxidative stress, whereas the use of Palm may overcome some of these adverse events.

What sites are liable for altered redox state in type 2 diabetic hearts subjected to metabolic/energetic stress? To answer this question, we measured mitochondrial respiration, coupling of oxidative phosphorylation (OxPhos), NADH levels, ΔΨ_m, and rate of H₂O₂ emission as a function of respiration state and electron transport mode in isolated mitochondria from WT and db/db hearts. Mitochondrial respiration was first quantified in the presence of substrates of complex I, II, and IV and the coupling of OxPhos under each condition (Fig. 3A–D). The dynamic response of mitochondrial NADH and ΔΨ_m to 5 mmol/L glutamate/malate (G/M) or 5 mmol/L succinate (Succ) and the specific rate of H₂O₂ emission under forward (FET) and reverse electron transport (RET) were also investigated (Fig. 3 and Supplementary Fig. 2).

First, we found that the basic bioenergetic behavior of mitochondria from WT and db/db hearts was noticeably different. The RCR (state 3/4), a measure of the degree of coupling of OxPhos, was markedly decreased at the level of the three main respiratory complexes in db/db mitochondria (Fig. 3A). In db/db mitochondria, a trend to higher values of state 4 respiration was evident in both complex I and II (Fig. 3B and C). Thus, it is plausible that the overall mechanism of energy transduction is impaired in db/db mitochondria, including electron flow and ATP synthesis.

Mitochondria from db/db hearts exhibited significantly greater H₂O₂ emission levels, both under FET (Fig. 3E and F) and RET (Fig. 3G and H) modes of electron transport. ROS emission during state 4→3 transition decreased less in db/db mitochondria, especially during FET, the physiological mode of electron transport (Fig. 3E), and in the face of similar mitochondrial NADH and ΔΨ_m dynamic response to complex I and II substrates (Supplementary Fig. 2). In RET, the decrease in ROS emission observed during state 4→3 transition, both in WT and db/db mitochondria (Fig. 3G), is likely due to the pronounced ΔΨ_m depolarization elicited by ADP (Supplementary Fig. 2), to which ROS by RET is extremely sensitive (38).

The state 4→3 transition is a fundamental one in mitochondrial physiology, representing the shift from a highly energized (i.e., high ΔΨ_m) and reduced redox status of low respiration and ATP synthesis to high respiration and ATP synthesis in response to elevated ATP demand (Fig. 3B–D, E, and F) (29,39). Hence, we further investigated whether high ROS outflow during 4→3 transition could be localized at a site other than the respiratory complexes. To mimic the db/db phenotype, we inhibited the adenine nucleotide translocator (ANT) (with carboxyatractyloside) or the ATP synthase (with oligomycin) in WT mitochondria, measuring H₂O₂ levels in the presence of substrate and ADP. Mitochondrial preparations preincubated with either inhibitor failed to decrease ROS emission when ADP was added, during both FET (Fig. 3I and J) and RET (Fig. 3K and L). Thus, impairments in the state 4→3 transition both at the level of electron transport in the respiratory chain and of the ANT and/or ATP synthase are likely liable for the sustained rise in ROS levels in mitochondria from db/db hearts.

Is ROS scavenging capacity impaired in type 2 diabetes, and if so, where? First, we measured the protein expression levels of the main components of the mitochondrial antioxidant system (Fig. 1). Since no differences were found between WT and db/db either in isolated mitochondria (Supplementary Fig. 3) or at whole-tissue levels (not shown), we sought for changes in the abundance of the mitochondrial TrxR2/Trx2 system. Again, no differences were observed (Fig. 4A). Therefore, we asked whether changes in Trx2 redox status, i.e., oxidized/reduced (TrxSS/Trx[SH]₂), occurred under nonenergized and energized conditions (Supplementary Fig. 2), using G/M as a substrate (FET) (33). This approach revealed marked differences between WT and db/db (Fig. 4B). At baseline, only 34% of Trx2 was in its reduced status, both in WT and db/db (Fig. 4B, top right). However, substrate addition to energize mitochondria to state 4 led to significantly greater Trx(SH)₂ in WT versus db/db mitochondria. Trx(SH)₂ increased in parallel with NADPH and ΔΨ_m when G/M was added (Supplementary Fig. 2). In state 3, Trx2(SH)₂ levels remained as high as in state 4. Since the reducing power of Trx2 is determined by the ratio TrxSS/Trx(SH)₂, we calculated the latter and expressed it as redox potential. Upon substrate addition to WT mitochondria, the redox potential became more negative, thus more reducing, from −294 ± 5 to −325 ± 2 mV. This increment was less pronounced in mitochondria from db/db mice, from −294 ± 5 to −313 ± 2 mV (Fig. 4B, bottom right). The 12-mV difference in Trx2 redox potential accounts for a significant 2.3-fold rise in the TrxSS/Trx(SH)₂ ratio.

Finally, we investigated whether altered Trx2 redox could be reversed by treatment with exogenous GSH ethyl ester (GSHee). Preincubation of db/db mitochondria with GSHee caused a significant increase in Trx(SH)₂ (from 64 ± 3 to 78 ± 1%), equivalent to a twofold rise in the TrxSS/Trx(SH)₂ ratio (Fig. 4B, right). Noteworthy, these values are comparable to those found in WT Trx(SH)₂ levels and similar to those found in state 3. Thus, in addition to their energetic impairment, db/db mitochondria exhibit diminished ROS scavenging capacity via the Trx2 system.

Next, we tested whether the HG + ISO regimen, which unveiled major redox/energetic deficits in cardiac db/db mitochondria, would reverberate on mechanical function and Ca²⁺ transient of myocytes from WT and db/db hearts. Changes in the amplitude and kinetics of contractility and Ca²⁺ transients in WT and db/db myocytes were measured under normal (5 mmol/L) and HG conditions, in the presence or absence of ISO, with and without Palm at two different concentrations.

Fractional shortening (FS) and Ca²⁺ transient amplitude/kinetics were similar in WT versus db/db myocytes at 5 and 30 mmol/L glucose (Supplementary Tables 1 and 2). Next, we examined the effects of β-stimulation via ISO on db/db and WT myocytes. At 5 mmol/L glucose, ISO response was similar in both groups (Table 1). Conversely, HG + ISO unveiled marked differences between db/db and WT myocytes: in db/db, ISO-induced enhancement of contractility/relaxation was lower, coupled to altered Ca²⁺ kinetics (Table 2 and Fig. 5). These findings suggest a deficit in cAMP-dependent protein kinase–mediated enhancement in contractility and Ca²⁺ handling of db/db myocytes under HG.

Under HG, Palm had a significant negative influence on the amplitude of Ca²⁺ transient and sarcomere shortening in db/db myocytes at both concentrations tested (Supplementary Tables 1 and 2). Yet, its administration in db/db myocytes challenged with HG + ISO improved contractility and Ca²⁺ handling to a higher extent than that seen in WT (Table 2). Palm preserved the cardiomyocyte response to the β-stimulant under HG in myocytes from db/db hearts (Fig. 5 and Table 2) as well as in WT and db/db cardiomyocytes at 5 mmol/L glucose (Table 1). Thus, Palm is beneficial for db/db cardiomyocytes exposed to a concomitant energetic and redox challenge.

To strengthen the causative link between impaired contractility and the more oxidized redox status of db/db cardiomyocytes, and to assess whether a shift in substrates may affect their function under challenging conditions, we tested the impact of cell-permeable GSHee or Palm. Exogenous GSHee offset ISO-mediated changes in ROS and GSH levels of db/db cardiomyocytes exposed to HG (Supplementary Fig. 4); all functional indexes recovered to WT levels after GSH (Fig. 5A–D). Moreover, GSHee replenished the thiol intracellular pool by ∼20%, without significantly changing NADH levels (Supplementary Fig. 4, insets), countering, in turn, the rise in ROS levels (Supplementary Fig. 4). To do so, GSHee acted through the reconstitution of the Trx2 pool (Fig. 4), ultimately limiting ROS emission in mitochondria from HG/ISO-challenged db/db mice and preventing myocyte mechanical deficiency.

Even more so, Palm significantly improved both contractility and Ca²⁺ handling, beyond the levels required to fully offset HG + ISO adverse effects (Fig. 5A–D and Table 2). This beneficial impact was coincident with a more reduced redox status (Fig. 2), both under normal conditions and HG. Thus, in diabetes, FAs may have a role that goes beyond that of energetic providers (11,40); this contention is reinforced by the fact that the impact of GSHee and Palm on myocyte function is not additive (Fig. 5).

The combined HG + ISO regimen in the absence or presence of Palm uncovered major differences in db/db and WT myocyte redox conditions (Fig. 2) and electro-contraction (E-C) coupling impairment (Fig. 5) that were not apparent at baseline (Fig. 2, Table 1, and Supplementary Table 1). These alterations were underlined by mitochondrial energetic/redox alterations (Fig. 3). To evaluate the functional implications of these changes at the whole-heart level, we used Langendorff-perfused mouse hearts and high-resolution optical AP imaging. With the former approach, we evaluated possible changes in LV function and coronary vascular response (CPP) under HG (Fig. 6 and Supplementary Fig. 6) and euglycemia (EG) (Supplementary Figs. 5 and 7) in the absence or presence of ISO and Palm (Fig. 6A–D and Supplementary Figs. 5–7A–D). As in cardiomyocytes (Fig. 5 and Table 2), whole WT hearts exhibited better LV function versus db/db hearts when exposed to HG + ISO (Fig. 6A–D). Optical mapping revealed that although HG alone did not alter APD in paced WT and db/db hearts, the HG + ISO regimen uncovered important functional differences between groups. Under HG + ISO challenge, there was a significant decrease in both APD50 (P = 0.019) and 75 (P = 0.030) in db/db compared with WT hearts (Fig. 6E–H and Supplementary Fig. 8).

Notably, Palm (same concentration as in cardiomyocytes) significantly improved the functional performance of db/db hearts exposed to HG + ISO by decreasing CPP and increasing LV-developed pressure (LVDP) concomitantly with the rates of db/db heart contraction (dP/dt_max) and relaxation (dP/dt_min) (Fig. 6C and D), whereas this FA did not affect WT heart performance (Fig. 6A–D). Similar results were obtained at both Palm concentrations tested under HG (Fig. 6 and Supplementary Fig. 6). However, the LVDP, dP/dt_max, and dP/dt_min, but not CPP, of WT hearts were significantly decreased under EG + ISO at the two Palm concentrations tested (Supplementary Figs. 5 and 7). In essence, Palm improved LV functional performance of hearts from diabetic animals subjected to metabolic/redox challenge.

Here we show that energetically/redox-stressed cardiomyocytes from type 2 diabetic db/db mice display impaired mitochondrial state 4→3 energetic transition, accounting for uninhibited ROS emission, under both FET and RET modes. Deficits in local pools of mitochondrial GSH and Trx2 account for this amplified ROS emission. The resulting more-oxidizing intracellular environment correlates with altered E-C coupling in cardiomyocytes/intact hearts. These alterations are causative and not incidental because exogenous GSH rescues both mitochondrial/cellular redox balance and E-C coupling. Also, shifting substrates, i.e., the use of the FA Palm, reestablishes proper myocyte redox milieu while improving function in metabolically/redox-challenged whole hearts.

The balance between ROS production in the respiratory chain and the efficiency of the ROS scavenging systems determines mitochondrial H₂O₂ emission as a function of the redox environment (29). Here we assign to mitochondrial impairment in the key state 4→3 energetic transition, and defective mitochondrial GSH and Trx2 systems, a leading role in the excess ROS emission occurring in db/db myocytes, ultimately resulting in myocyte/whole-heart functional impairment. Uncoupling of OxPhos, along with the inability of mitochondria to decrease ROS levels during the state 4→3 transition under FET, plays a major role in the energetic/redox dysfunction. Recent data show that under states 4 and 3, mouse heart mitochondria divert a significant portion of total respiratory flow to ROS generation, which is continuously offset by GSH/Trx scavenging systems (30,33).

Indeed, HG + ISO provoked a concomitant rise in ROS levels and GSH depletion that was associated with contractile and Ca²⁺ handling failure. Mechanistically, the resetting of mitochondrial and cellular redox balance operated by exogenous GSH was mediated by the TrxR2/Trx2/peroxiredoxin 3 (Prx3) system, whose redox potential became more reduced in the presence of additional cytoplasmic GSH. The augmented scavenging capacity bestowed by GSHee could also indirectly increase SOD activity by lowering H₂O₂, which exerts a negative feedback on this enzyme (41) (Supplementary Fig. 4). Herein, intact ex vivo perfused db/db hearts exhibited altered vascular tone, impaired LV function, and significant shortening of APD compared with WT hearts. These deficits are all crucial for maintaining proper E-C coupling, and thus contraction, under both basal and increased-workload conditions. A possible mechanism responsible for APD shortening is the nonischemic activation of sarcolemmal ATP-sensitive (K-ATP) channels, reflecting energetic shortage when energy demand is high in db/db hearts. Impaired myocardial Ca²⁺ handling is known to be involved in diabetic cardiomyopathy (42). Our study shows that these alterations become particularly prominent when diabetic cells are subjected to redox/energetic stress, thus expanding previous observations showing that β-adrenergic response may be blunted in db/db myocytes (42,43).

The status of diabetic cardiomyopathy is maintained by the compounding effects of HG, mitochondrial energy deficit, and oxidative stress. However, no major contractile deficits are evident under HG alone, even in db/db hearts. This prompted us to use conditions of combined energetic/redox challenge (44). Since myocardial glucose metabolism (e.g., transport and glycolysis) is decreased and FA oxidation increased in diabetes (11), we analyzed the effect of the saturated FA Palm on intracellular redox, contractility, and Ca²⁺ handling in myocytes/hearts. Palm determined a transition from oxidized-to-reduced cellular redox status in db/db cardiomyocytes, abating ROS levels drastically. This effect was coupled to a marked GSH rise both in WT and db/db myocytes (Fig. 2B). Consequently, Palm significantly improved ISO-induced contractile reserve in db/db cardiomyocytes. Conversely, its absence was accompanied by marked deficits in Ca²⁺ handling and myocyte mechanical properties (Fig. 5A–D). Myocardial substrate use in patients with diabetes involves increased delivery of FAs (3,11,45). At the whole-heart level, Palm produced a substantial improvement in LV function in db/db hearts exposed to metabolic/redox stress, and a relative decrease in the WT. Previous reports have elegantly shown that, in the presence of 11 mmol/L glucose, Palm has a negative impact on LV function of db/db mice (25). Present evidence is consistent with these findings, and may help to mechanistically explain the decline in LV function and efficiency seen in db/db hearts with Palm + glucose (25). However, here we aimed to test a different question, i.e., how isolated db/db myocytes/hearts respond to increased metabolic/redox stress using the HG + ISO regimen. Although Palm partially worsened LV function in WT hearts, it definitely improved the performance of db/db hearts subjected to HG + ISO. Thus, our study suggests that in animals with type 2 diabetes, FAs are important for handling increased workload that demands higher energy supply while exposing cardiac tissue to the likelihood of ROS imbalance (46).

FAs are a major source of acetyl-coenzyme A for the tricarboxylic acid cycle, contributing 60–70% of myocardial ATP at physiological levels (40). However, the higher energetic budget provided by Palm (three times higher than from glucose) is produced up to two-thirds in the form of reducing equivalents (24 NADH and 8 FADH₂) from β-oxidation in mitochondria. This reducing power contributes electrons to antioxidant systems and the mitochondrial respiratory/energetic machinery (Fig. 1). The significantly lower levels of GSH found in db/db cardiomyocytes argue in favor of a more oxidized and constitutively compromised supply of reducing equivalents to the GSH/Trx systems through NADPH (Fig. 1). The rescue of Trx(SH)₂ to WT levels via GSHee supports this contention, suggesting that the achievement of a more balanced redox status could be mediated by the TrxR2/Trx2/Prx3 pathway, a major controller of H₂O₂ emission from mitochondria (30,33). The recovery of contractile activity observed in db/db cardiomyocytes preincubated with GSHee, along with a more-reduced redox status in cells, and of Trx2 in mitochondria, demonstrates for the first time that changes in mitochondrial pools of GSH and Trx2 can have a negative impact that propagates from mitochondria to the function of myocytes in challenged db/db hearts.

Concerning the role of FAs, one intrinsic limitation of our study is that we tested only the acute effects of Palm. In a more chronic stand, the upregulation of FA uptake is known to promote the accumulation of intracellular lipid, resulting in lipotoxic effects (47), likely due to deficits in lipase function, ultimately leading to cardiac disease (11,25). Notwithstanding, our findings suggest that, when properly used, FAs are important suppliers of reducing equivalents, particularly when db/db myocytes/hearts are exposed to HG and more workload.

This work was supported by National Institutes of Health Grant R01-HL-091923-01. C.G.T. was supported by International Society for Heart Research–European Section/Servier; V.C. by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil; and S.S. by the National Natural Science Foundation of China Grant 81100155.

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

C.G.T., V.C., B.A.S., C.X., and S.S. researched data and contributed to discussion. W.H.W., B.O., and R.C.S.-B. contributed to discussion and reviewed and edited the manuscript. S.C. and F.G.A. researched data, contributed to discussion, and reviewed and edited the manuscript. N.P. and M.A.A. researched data, contributed to discussion, and wrote the manuscript. M.A.A. 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.

Parts of this study were presented as a poster at the 54th Annual Meeting of the Biophysical Society, San Francisco, California, 20–24 February 2010; 55th Annual Meeting of the Biophysical Society, Baltimore, Maryland, 5–9 March 2011; and 56th Annual Meeting of the Biophysical Society, San Diego, California, 25–29 February 2012.