Isolated Mouse Adult Cardiomyocytes Display Minimal Mitochondrial ATP Demand and Maximal Reliance on Glycolysis

Graphical Abstract

Dear Editor,

Mitochondrial maladaptation has been recognized as a pathological mechanism in the development of heart failure (1–3). To elucidate and characterize mitochondrial maladaptation in cardiomyocytes, one must compare cellular bioenergetics between healthy and heart failure conditions. For this purpose, investigators have optimized a standardized protocol (4) and have reported differential cellular bioenergetic profiles between sex (5) and ventricles (6). However, a notable and consistent finding across these studies (4–6), often overlooked or insufficiently explored, was that mouse adult cardiomyocytes (ACMs) were unresponsive to oligomycin (ATP synthase inhibitor). As a result, these studies typically report only basal or maximal respiration rates, neglecting ATP-linked respiration and glycolytic capacity, thereby leaving important aspects of cellular bioenergetics unexplored.

To this end, we have comprehensively characterized the bioenergetic and glycolytic capacities of neonatal rat ventricular myocytes (NRVMs), human induced pluripotent stem cell derived cardiomyocytes (hiPSC-CMs) and ACMs and reported their oxygen consumption (OCR) and proton efflux rates (PER), respectively (4, 7). We found that, unlike NRVMs and hiPSC-CMs, blocking ATP synthase by oligomycin in ACMs minimally decreased their basal OCR (Figs. 1A – C). Concomitantly, we found that oligomycin addition significantly increased the basal PER in both NRVMs and hiPSC-CMs as a compensatory mechanism for ATP production (Figs. 1D – E). In strong contrast, the basal PER was minimally increased in ACMs (Fig. 1F). Additionally, a notable observation in ACMs was the remarkable similarity between their OCR and PER profiles compared to both NRVMs and hiPSC-CMs (Figs. 1C and F).

Figure 1. Isolated ACMs rely more on glycolysis than mitochondria for ATP production.

Oxygen consumption (OCR) and proton efflux rate (PER) traces of intact (A and D) NRVMs (20,000 cells/well on gelatin coated XF 96-well plates), (B and E) hiPSC-CMs (20,000 cells/well on gelatin coated XF 96-well plates) and (C and F) ACMs isolated from 10-week-old male C57BL/6J mouse as described previously (10) (4,000 – 6000 cells/well on laminin coated XF 96-well plates) in basal and after sequential injections of oligomycin (NRVMs: 2μM, iPSC-CMs and ACMs: 2.5μM), FCCP (NRVMs, iPSC-CMs and ACMs: 1μM), and rotenone/antimycin A (NRVMs: 0.5μM, iPSC-CMs: 1μM and ACMs: 10μM). Measurements performed using Seahorse XF Pro Analyzer were normalized to cell number counted using BioTek Cytation 5 Cell Imaging Multimode Reader. All intact bioenergetic assays were performed in XF DMEM or RPMI medium (pH 7.4) containing 10mM glucose, 1mM pyruvate, and 2mM glutamine. (G) ATP-linked (OCR: datapoint 6 subtracted from 3) and proton leak (OCR: datapoint 15 subtracted from 6) respiration; (H) basal (Glycolytic PER: datapoint 3) and oligomycin-stimulated (Glycolytic PER: datapoint 6 subtracted from 3) glycolytic capacities; (I) mitochondrial and glycolytic ATP production rates from NRVMs (n = 5 replicates), hiPSC-CMs (n = 14 replicates) and ACMs (n1 = 8, n2 = 10, n3 = 16, n4 = 12 replicates, each isolated from 4 different mice on separate days).

(J) OCR and (K) PER traces of intact ACMs isolated from 10–12-week-old male C57BL/6J mouse (2000 cells/well) in basal and after sequential injections of oligomycin in a dose-dependent manner (1, 2.5, 5 and 10 μM), 1μM FCCP, and 10μM rotenone/antimycin A. Measures were normalized to cell number. (L) Mitochondrial and glycolytic ATP production rates from ACMs (n1 = 8, n2 = 6, n3 = 7 replicates per condition, each isolated from 3 different mice on separate days).

(M) Representative confocal images of hiPSC-CMs and ACMs stained with TMRE under control conditions and following treatment with FCCP (iPSC-CMs: 4μM and ACMs: 1μM) or 2.5μM oligomycin. Scale bar = 25μm for hiPSC-CMs and 10μm for ACMs. (N) TMRE fluorescence was captured using excitation at 548nm and emission at 574nm. Mean TMRE fluorescence per cell was quantified using the ImageJ software for hiPSC-CMs (n1 = 10, n2 = 10 images, each containing 15–20 cells, obtained from two different batches cultured on separate days) and ACMs (n1 = 6, n2 = 7, n3 = 24 individual cells, each isolated from 3 different mice on separate days).

(O) Citrate synthase (CS) activity measured using 1.25μg of frozen whole-cell lysates prepared from hiPSC-CMs (n = frozen hiPSC-CMs from 2 batches) and ACMs (n = frozen ACMs from 3 mice). Reaction mixtures included the sample, 100μM DTNB and 0.3mM Acetyl-CoA, to which 0.5mM oxaloacetic acid was added and the increase in absorbance at 412nm was measured. Enzyme activities were normalized to protein levels.

(P) Extracellular acidification (ECAR) traces from 5μg of frozen whole-cell lysates prepared from hiPSC-CMs (n = frozen hiPSC-CMs from 2 batches) and ACMs (n = frozen ACMs from 3 mice) sustained by 5mM succinate/2μM rotenone before and after sequential injections of 2μM antimycin A, 20 mM ATP/1μM FCCP, and 5μM oligomycin. Measures were normalized to protein levels. The complex V ATP hydrolytic capacity was quantified as the difference in acidification (ECAR) rates between ATP/FCCP and oligomycin injections. All frozen bioenergetic assays were performed in MAS buffer (70mM sucrose, 220mM mannitol, 10mM KH₂PO₄, 5mM MgCl₂, 1mM EGTA, and 2mM HEPES, pH 7.2) containing 5mM succinate, 2μM rotenone and 100μg/ml of cytochrome c.

OCR traces of permeabilized (Q) hiPSC-CMs (20000 cells/well, n = 10 replicates from two batches) and (S) ACMs isolated from 10–12-week-old male C57BL/6J mouse (2000 cells/well, n1 = 12, n2 = 14, n3 = 24 replicates per condition, each isolated from 3 different mice on separate days) sustained by 5mM Pyruvate/0.5mM Malate or 80μM Palmitoyl-CoA/0.5mM L-Carnitine/0.5mM Malate before and after sequential injections of 4mM ADP, 10μM oligomycin, and 10μM rotenone/antimycin A. Measures were normalized to cell number. ATP-linked (OCR: datapoint 9 subtracted from 6) and proton leak (OCR: datapoint 12 subtracted from 9) respiration from (R) hiPSC-CMs and (T) ACMs. All permeabilized bioenergetic assays were performed in MAS buffer (70mM sucrose, 220mM mannitol, 10mM KH₂PO₄, 5mM MgCl₂, 1mM EGTA, and 2mM HEPES, pH 7.2) containing 0.2% BSA and 2nM Seahorse XF Plasma Membrane Permeabilizer (PMP).

Data are presented as mean ± SEM. P values were calculated by (G – I, L, R and T) multiple t tests, (N) two-factor ANOVA, corrected by post hoc ‘Holm-Sidak’s’ multiple comparisons test or (O and P) t tests. ns, non-significant; *P < 0.05; **P < 0.01; ***P <0 .001.

Follow-up analyses revealed that proton leak (oligomycin insensitive) was significantly higher than ATP-linked (oligomycin sensitive) respiration in ACMs, while ATP-linked respiration was higher in both NRVMs and hiPSC-CMs (Fig. 1G). FCCP is a mitochondrial uncoupler that mimics energy demand by operating the electron transport chain at maximal capacity via mitochondrial membrane potential depletion. The similarity in response of respiration upon FCCP stimulation in all three types of cardiomyocytes suggests that the oligomycin insensitivity observed in ACMs was due to reduced mitochondrial ATP demand and not cell viability (Figs. 1A – C). Since isolated ACMs have lower dependence on mitochondrial respiration for ATP production, we explored whether they depend on alternative pathways, such as glycolysis. To test this, we calculated the basal and oligomycin-stimulated glycolytic capacities and found that basal glycolysis in ACMs was more than 50 percent of their oligomycin-stimulated glycolytic capacity (Fig. 1H), suggesting ACMs are already relying maximally on glycolysis. Additionally, we found that glycolytic ATP production rate was significantly greater than mitochondrial-dependent ATP production rate in ACMs, whereas both NRVMs and hiPSC-CMs exhibited the opposite (Fig. 1I). Moreover, consistent with a prior report (4), increasing concentrations of oligomycin resulted in similar inhibition of ATP synthase but progressively reduced FCCP-induced OCR and PER in ACMs (Fig. 1J – L). Taken together, our findings indicate that isolated ACMs rely more on glycolysis than mitochondria for ATP production.

Oligomycin sensitivity is influenced by several factors, including proton leak, ATP demand, ATP synthesis rate, defects in substrate supply or electron transport chain (ETC) activities. Our systematic assessment provided evidence that reduced mitochondrial ATP demand in ACMs might be the fundamental reason for oligomycin insensitivity. Proton leak is often associated with lower mitochondrial membrane potential (ΔΨ_m). Assessment of ΔΨ_m in hiPSC-CMs and ACMs by TMRE fluorescence showed that despite the ΔΨm suppression that FCCP induces in both cell types, oligomycin addition did not affect ΔΨ_m in ACMs compared to the control, unlike in hiPSC-CMs, suggesting oligomycin insensitivity (Fig. 1M – N). Next, we assessed citrate synthase (CS) activity, a key regulatory enzyme of the TCA cycle and an indirect marker of mitochondrial abundance, and found that ACMs exhibited significantly higher CS activity compared to hiPSC-CMs (Fig. 1O). Measurement of the mitochondrial ATP hydrolytic capacity (reverse ATP synthase activity) (8) in both hiPSC-CMs and ACMs showed significantly higher activity in ACMs, which was completely inhibited following oligomycin injection (Fig. 1P). Finally, we permeabilized both hiPSC-CMs and ACMs to assess substrate-specific oxidation using Pyruvate/Malate (for pyruvate oxidation) and Palmitoyl-CoA/L-Carnitine/Malate (for fatty acid oxidation). In both cases, ACMs demonstrated more robust activity compared to hiPSC-CMs, emphasizing that there were no defects in substrate supply or ETC activities. Interestingly, ACMs exhibited a higher fatty acid oxidization capacity compared to pyruvate oxidation, whereas hiPSC-CMs showed the opposite (Figs. 1Q – T). This finding is consistent with recent literature, which reported that ACMs are more oxidative than hiPSC-CMs (9). Nevertheless, consistent with the bioenergetics of intact cells, permeabilized ACMs displayed higher proton leak compared to ATP-linked respiration, regardless of the substrate used, further suggesting reduced mitochondrial ATP demand (Figs. 1S – T).

In summary, our findings demonstrate that oligomycin insensitivity in ACMs cannot be attributed to defects in substrate supply, ETC, or ATP synthase activity. Instead, our data clearly establish that ACMs possess a fully functional and oligomycin-sensitive mitochondrial ATP synthase, which remains underutilized for coupled respiration due to reduced mitochondrial ATP demand. We anticipate that several factors, including extreme ‘ischemic’ stress or exposure to myosin inhibitors during ACM isolation and/or their adaptation from the in vivo to in vitro system, may contribute to this phenomenon. We also acknowledge that our observations are unique to the isolated mouse ACMs under specified culture conditions. Based on these findings, we propose that ACMs are more suitable for assessing individual ETC activities, whereas hiPSC-CMs are better suited for evaluating coupled respiration. Ultimately, our study underscores the importance of careful interpretation of cellular bioenergetics in heart failure research.

Highlights.

• Mouse adult cardiomyocytes (ACMs) are insensitive to oligomycin treatment

• ACMs rely more on glycolysis than mitochondria for ATP production

• Proton leak in ACMs exceeds ATP-linked respiration across all substrates

• ACMs have functional ATP synthase that is underused due to low ATP demand