Distinct effects of cardiac mitochondrial calcium uniporter inactivation via EMRE deletion in the short and long term

Abstract

Transport of Ca²⁺ into mitochondria is thought to stimulate the production of ATP, a critical process in the heart’s fight or flight response, but excess Ca²⁺ can trigger cell death. The mitochondrial Ca²⁺ uniporter complex is the primary route of Ca²⁺ transport into mitochondria, in which the channel-forming protein MCU and the regulatory protein EMRE are essential for activity. In previous studies, chronic Mcu or Emre deletion differ from acute cardiac Mcu deletion in response to adrenergic stimulation and ischemia/reperfusion (I/R) injury, despite equivalent inactivation of rapid mitochondrial Ca²⁺ uptake. To explore this discrepancy between chronic and acute loss of uniporter activity, we compared short-term and long-term Emre deletion using a novel conditional cardiac-specific, tamoxifen-inducible mouse model. After short-term Emre deletion (3 weeks post-tamoxifen) in adult mice, cardiac mitochondria were unable to take up Ca²⁺, had lower basal mitochondrial Ca²⁺ levels, and displayed attenuated Ca²⁺-induced ATP production and mPTP opening. Moreover, short-term EMRE loss blunted cardiac response to adrenergic stimulation and improved maintenance of cardiac function in an ex vivo I/R model. We then tested whether the long-term absence of EMRE (3 months post-tamoxifen) in adulthood would lead to distinct outcomes. After long-term Emre deletion, mitochondrial Ca²⁺ handling and function, as well as cardiac response to adrenergic stimulation, were similarly impaired as in short-term deletion. Interestingly, however, protection from I/R injury was lost in the long-term. These data suggest that several months without uniporter function are insufficient to restore bioenergetic response but are sufficient to restore susceptibility to I/R.

Graphical Abstract

1. Introduction

The heart, which requires large amounts of energy to continually pump blood, relies primarily on mitochondria to generate ATP [1]. Cardiomyocytes are thus among the most mitochondria-rich cells in the body; approximately 30% or more of the volume of a cardiomyocyte is comprised of mitochondria [2]. Among its many roles in vital cell processes such as gene expression, signal transduction and muscle contraction [3], calcium (Ca²⁺) has been shown to stimulate mitochondrial ATP production [4]. In the mitochondrial matrix, Ca²⁺ activates pyruvate dehydrogenase (PDH) as well as the tricarboxylic acid (TCA) cycle enzymes alpha-ketoglutarate dehydrogenase and isocitrate dehydrogenase [5,6]. This process is thought to be critical for matching energy supply to demand, such as during periods of increased workload [7]. However, overload of mitochondrial Ca²⁺ can induce opening of the mitochondrial permeability transition pore (mPTP) [8]. Mitochondrial Ca²⁺ overload has been implicated in a range of different pathological disorders [9–12], and Ca²⁺ induced mPTP opening has been linked to cell death in ischemia/reperfusion (I/R) injury in the heart [13,14]. Inhibition of mPTP opening or inhibition of the upstream event of mitochondrial Ca²⁺ uptake could therefore be beneficial therapeutically in myocardial infarction. Deeper understanding of the regulation of mitochondrial Ca²⁺ is thus critical for targeting mitochondrial dysfunction in pathology while not compromising physiological bioenergetics.

Mitochondrial Ca²⁺ uptake occurs through a highly selective multiprotein Ca²⁺ channel known as mitochondrial Ca²⁺ uniporter complex [15,16] (hereafter referred to as the “uniporter”), located in the inner mitochondrial membrane. The uniporter is made up of multiple auxiliary subunits, including MCU, the transmembrane protein that oligomerizes to form the Ca²⁺-permeant pore of the complex [17,18] and MCUb, a homologous protein that can act as a negative regulator [19]. Further, the uniporter is regulated by a family of intermembrane space-localized Ca²⁺-sensing proteins comprising three members, MICU1 [20], MICU2 [21], and MICU3 [22], and by a small transmembrane protein named Essential MCU Regulator (EMRE) [23]. EMRE has been shown to stabilize the uniporter serving as an anchor between MCU and MICU1 [23,24], and in a recently solved structure was shown to bind 1:1 with MCU to allow Ca²⁺ conductance through conformational changes in the tetrameric channel [25].

Recent research has continued to provide insight into the contribution of each subunit in regulating the physiological functions of various organ systems. Multiple mouse models of MCU deletion and inactivation have been used to study the role of mitochondrial Ca²⁺ in the heart. The first mouse model used a gene-trap method to globally delete Mcu in the germline in outbred CD-1 mice [26], as germline Mcu deletion was embryonic lethal in inbred C57Bl/6 mice. In these Mcu^−/− mice, mitochondrial Ca²⁺ uptake was impaired across tissues [27] and intramitochondrial Ca²⁺ was decreased in the heart [28]. As expected, given these findings, sensitivity to Ca²⁺-induced mPTP opening was decreased; however, germline Mcu deletion did not confer protection in an ex vivo Langendorff model of I/R injury [26]. This counterintuitive result was supported by similar ex vivo I/R data in a study of mouse model expressing cardiac-specific dominant negative Mcu [29]. Subsequently, a cardiac-specific tamoxifen-inducible mouse model of Mcu deletion (hereafter referred to as Mcu^cKO mice) was developed by crossing floxed Mcu^fl/fl mice to the αMHC-MerCreMer (MCM) transgenic strain, enabling the acute deletion of Mcu in adulthood [30,31]. Though intramitochondrial Ca²⁺ levels were unchanged, upon in vivo I/R via left coronary artery ligation, infarct area was decreased in adult Mcu^cKO mice fed [30] or injected with [31] tamoxifen, showing that acute Mcu deletion did protect against I/R injury (see Table 1 in Discussion). Though protection was found in in vivo I/R and not found in ex vivo I/R, making the type of model used a possible source of this discrepancy, the more likely and interesting possibility would be that mice that have lost MCU acutely have the expected protection against I/R injury, but Mcu^−/− and DN-Mcu mice that experience chronic loss of uniporter function do not retain protection. While activation of unknown compensatory pathways in the chronic absence of MCU could feasibly explain these findings, whether this could occur over long timescales even in adulthood or requires MCU loss during development is unclear.

Table 1.

Genetic mouse models of mitochondrial Ca²⁺ uniporter inactivation.

Author and year	Knock out model	Tamoxifen route and dose	Timing of experiments	Response to β-adrenergic stimulation	Protection after ischemia/reperfusion
Pan et al, 2013 [27]	global MCU deletion	N/A	age 3–5 months	unchanged	not protected, ex vivo
Rasmussen et al, 2015 [29]	cardiac (αMHC) expression of a dominant-negative MCU	N/A	unspecified	diminished	not protected, ex vivo
Kwong et al, 2015 [30]	cardiac (αMHC-MerCreMer) MCU deletion	chow 4 weeks	6 weeks post-tamoxifen	diminished	protected, in vivo
Luongo et al, 2015 [31]	cardiac (αMHC-MerCreMer) MCU deletion	injection, 40 mg/kg/day 5 days	10 days post-tamoxifen (hemodynamics) 3 weeks post-tamoxifen (IR)	diminished	protected, in vivo
Liu et al, 2020 [32]	global EMRE deletion	N/A	age 3–6 months	unchanged	not protected, ex vivo

In metazoan uniporters, EMRE is essential for the transport of Ca²⁺ through MCU [23]. The first mouse model of CRISPR-mediated deletion of Emre in the germline (Emre^−/− mice) in outbred mice turned out to have substantial similarity to the Mcu^−/− model [32]. The absence of mitochondrial Ca²⁺ uptake, decrease in intramitochondrial Ca²⁺ concentration, and decrease in sensitivity to Ca²⁺-induced mPTP opening in mitochondria from Emre^−/− mice [32] recapitulated findings in Mcu^−/− mice. Furthermore, like the Mcu^−/− mice, Emre^−/− mice did not exhibit protection in the Langendorff ex vivo I/R model [32]. Therefore, the potential compensatory response in chronic Mcu deletion models occurs as a result of loss of uniporter Ca²⁺ uptake activity and not loss of MCU per se. Here we sought to determine how short-term and long-term deletion of Emre in the heart would alter mitochondrial and tissue function. We developed a tamoxifeninducible cardiomyocyte-specific model of Emre deletion (Emre^cKO mice) and characterized these mice 3 weeks (short-term) and 3 months (long-term) post-tamoxifen. As expected, we confirmed in the short and long term that EMRE is required for cardiac mitochondrial Ca²⁺ uptake. Like Emre^−/− mice, Emre^cKO mice exhibited a significant reduction in cardiac intramitochondrial Ca²⁺ levels and decreased Ca²⁺-induced mPTP opening, which persisted from short to long-term deletion. Unlike Emre^−/− mice, however, in Emre^cKO mice the short-term deletion of Emre led to blunted responses to adrenergic stimulation and protection from I/R injury in an ex vivo Langendorff model. In the long-term absence of EMRE, mitochondrial Ca²⁺ uptake was still absent in Emre^cKO cardiac mitochondria, and adrenergic response continued to be attenuated. Strikingly, the protection from I/R injury seen upon short-term EMRE loss was no longer observed with long-term EMRE absence, suggesting that protection from I/R is lost in a time-dependent manner with chronic uniporter inactivity.

2. Methods

2.1. Development of cardiac-specific Emre knockout mouse model (Emre^cKO)

Emre ^flox/flox mice (C57BL6/N) from the Mutant Mouse Resource & Research Centers (MMRRC) were crossed with cardiac-specific tamoxifen-inducible Cre mice (αMHC-Mer-Cre-Mer, abbreviated MCM⁺, C57BL6/J) to generate cardiac-specific Emre^cKO mice (C57BL6/N and C57BL6/J hybrids). To induce Emre deletion, 9- to 11-week-old male and female mice were injected intraperitoneally (i.p.) with tamoxifen (30 mg/kg body weight) per day for 5 days, and experiments were conducted 3 weeks or 3 months post-tamoxifen as described in the text. All groups, including Emre^flox/flox and MCM⁺ controls, received tamoxifen. The animal procedures were approved by the University of Minnesota Institutional Animal Care and Use Committee and performed in compliance with all relevant laws and regulations. Data are shown from female mice and from male mice where indicated.

2.2. Mitochondria heart isolation

Heart mitochondrial fractions were obtained as previously described [32]. Briefly, the heart was isolated and placed in an ice-cold isolation buffer, containing (mM): 225 mannitol, 75 sucrose, 5 MOPS, 0.5 EGTA, 2 taurine, pH 7.2. After being minced and homogenized with a T25 Ultra-Turrax disperser, heart tissue was digested for 5 min using trypsin in cold isolation buffer followed by addition of 0.2% BSA to stop digestion. The homogenates were centrifuged at 500 g for 5 min and then the resulting supernatant was centrifuged at 11,000 g for 5 min. The final mitochondria pellet was resuspended in isolation buffer and protein concentration was measured by Bradford assay.

2.3. Ca²⁺ retention capacity (CRC), membrane potential (Δψ), and swelling

The Ca²⁺ retention capacity (CRC) assay was performed by monitoring extramitochondrial Ca²⁺ levels via the fluorescence of the membrane-impermeable Ca²⁺ indicator Calcium Green-5N (Ca Green-5N, 1.0 μM). Isolated mitochondria were suspended (0.1 mg/mL) in respiration buffer containing (mM): 125 KCl, 2 K₂HPO₄, 20 HEPES, 1 MgCl₂, 0.010 EGTA and pH 7.3. The mitochondria were energized with 10 mM succinate and 2 μM of rotenone or where indicated 10 mM glutamate and 5 mM malate. Successive boluses of Ca²⁺ (10 μM) were added to induce mitochondrial Ca²⁺ uptake. Loss of mitochondrial membrane potential (Δψ) upon addition of a large bolus of Ca²⁺ (400 μM) was measured via the fluorescence of the indicator safranin (2 μM) [33]. Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) was used as uncoupling agent. Mitochondrial swelling was measured in isolated mitochondria upon addition of a large bolus of Ca²⁺ (400 μM) as the reduction of absorbance at 540 nm. Fluorescence or absorbance was monitored using a Synergy H1 microplate reader (BioTek Instruments, Winooski, VT, USA).

2.4. Intramitochondrial Ca²⁺ content (_mCa²⁺)

Intramitochondrial free Ca²⁺ was measured as previously described [32]. Briefly, heart mitochondria were isolated in the isolation buffer supplemented with 2 μM Ru360 and 10 μM CGP-37157 to avoid calcium influx and efflux, respectively, during the isolation process. Isolated mitochondria were loaded with 20 μM of Fluo-4 AM at room temperature and incubated for 45 minutes in the dark. Afterwards, the pellet was washed twice with isolation buffer and then was washed again with isolation buffer without EGTA. Finally, the pellet was resuspended in buffer containing (mM): 137 KCl, 20 HEPES, 2 K₂HPO₄ at pH 7.2. After obtaining the fluorescence of the mitochondrial sample F, to calibrate, F_max was obtained after addition of 10 μM of ionomycin and 15 μM of Ca²⁺, and F_min was obtained after addition of 2 mM of EGTA. Fluorescence was monitored using a Synergy H1 microplate reader. Intramitochondrial Ca²⁺ was estimated using the formula: _mCa²⁺ (nM)= K_d*((F – F_min)/(F_max – F)), where K_d is the dissociation constant for Fluo-4, 335 nM. The data were reported normalized to control groups performed in the same experiment.

2.5. ATP production rate

The ATP production rate was measured in fresh heart mitochondria using the ATP bioluminescence assay Kit CLS II (Roche, 11699695001) [34]. Briefly, 10 μg of protein was added to 75 μL of ATP buffer (mM): 125 KCl, 10 HEPES, 5 MgCl₂ and 2 K₂HPO₄, pH: 7.4, along with 75 μL of luciferin/luciferase buffer. Luminescence was monitored using a Synergy H1 microplate reader. The baseline signal was recorded for 1 minute, then 10 mM of succinate or 10 mM glutamate/5 mM malate and 50 μM of ADP were added. The slope was analyzed, and the ATP concentration (nanomoles per minute per mg of protein) was calculated from a fresh ATP standard curve. Where indicated, the ADP buffer was supplemented with 20 μM of Ca²⁺.

2.6. Immunoblotting

Mitochondrial or cardiac tissue protein was resolved on NuPAGE Bis-Tris 4 to 12% gels (Invitrogen) and transferred to nitrocellulose membranes and incubated with following antibodies: MCU (1:2000, Cell Signaling, 14997S), EMRE (1:200, Santa Cruz Biotechnology, sc-86337), MICU1 (1:500, Sigma Aldrich, HPA037480), PDH (1:2000, Santa Cruz Biotechnology, sc-377092), p-PDH (Ser293) (1:2000, Cell Signaling, 31866S), CPT1B (1:1000, Proteintech, 22170-1-AP), p-CAMKII (1:1000, Cell Signaling, 12716S), CAMKIIα/β/γ/δ (1:1000, Santa Cruz Biotechnology, sc-5306), CypD (1:1000, abcam, ab110324), GAPDH (1:5000, Cell Signaling, 97166S, VDAC1 (1:1000, Santa Cruz Biotechnology, sc-390996) and then with LI-COR IRDye secondary antibodies. The membranes were scanned on the Licor Odyssey system.

2.7. Blue native PAGE (BN-PAGE)

Freshly isolated cardiac mitochondria were incubated with NativePAGE sample buffer (Invitrogen) containing 2% ReadyShield protease & phosphatase inhibitor cocktail (Sigma) and 2% digitonin [32]. After 20 minutes on ice the samples were centrifuged for 30 minutes at 18,000 g at 4 °C. Coomassie blue (5%) was added to the supernatants, then 40 μg of protein was loaded. The electrophoresis running buffers were prepared according to the manufacturer’s protocol for the NativePAGE Novex Bis-Tris Gel System (Invitrogen). After electrophoresis gels were transferred onto nitrocellulose membranes for 18 hours at 4 °C. After blocking with 5% milk for 6 hours at 4 °C, membranes were incubated with MCU antibody (1:1000, Sigma Aldrich, HPA016480) overnight at 4 °C and then with HPR-conjugated secondary antibody. The blots were developed with ECL Western Blotting Substrate (Pierce) and quantified using a ChemiDoc MP Imaging System (BioRad).

2.8. Real-Time Quantitative PCR (RT-qPCR)

Heart tissue RNA was isolated by RNA MiniPrep (Zymo Research, R2052). Reverse transcription (10 ng of RNA) and real-time PCR assay were performed by Luna^® Universal One-step RT-qPCR kit (New England BioLabs, E3005L) on a Bio-Rad CFX96 System. The following primer sequences were used: MCU, forward primer: 5’-GTGCCCTCTGATGACGTGACGG-3’, reverse primer: 5’- ATGACAAGCTTAAAGTCATC-3’; EMRE, forward primer: 5’- CATTTTGCCCAAGCCGGTG-3’, reverse primer: 5’-CCTGTGCCCTGTTAATCGTCGT-3’. The following primer sequences were used for reference genes: eukaryotic translation initiation factor EIF35S, forward primer: 5’-CTGAGGATGTGCTGTCTGGGAA-3’, reverse primer: 5’-CCTTTGCCTCCACTTCGGTC-3’; ribosomal protein 36B4, forward primer: 5’-GGCCCTGCACTCTCGCTTTC-3’, reverse primer: 5’-TGCCAGGACGCGCTTGT-3’[35].

2.9. Oxygen consumption rate (OCR)

As previously described [36], mitochondria were loaded in a 96-well Seahorse plate (4 μg of protein per well) in mitochondrial assay solution (MAS) (mM): 220 mannitol, 70 sucrose, 10 KH₂PO₄, 5 MgCl₂, 2 HEPES, 1 EGTA and 0.2% of fatty acid-free BSA, pH: 7.2 supplemented with: 5 mM pyruvate/0.5 mM malate or 40 μM palmitoylcarnitine/0.5 mM malate. Using an Agilent Seahorse XFe96 Analyzer, OCR was analyzed at baseline and following sequential injections of 4mM ADP (state III), 2.5 μg/mL oligomycin (state IV_o), 6 μM FCCP (state III_u), and 4 μM rotenone/4 μM antimycin A. Basal respiration was calculated as the difference between the baseline OCR and post-rotenone/antimycin A OCR (non-mitochondrial oxygen consumption). The respiratory control ratio (RCR) was calculated as the ratio of state III_u to state IV_o. OCR values were normalized to μg of protein.

2.10. Mitochondrial enzyme activity

As described in [37], Complex I (NADH: ubiquinone oxidoreductase) activity was measured by following the absorbance change of NADH at 340 nm (ε: 6.2 mM⁻¹ cm⁻¹ ) using potassium phosphate buffer (50 mM potassium phosphate dibasic with 50 mM of potassium phosphate monobasic, pH: 7.5), 0.100 mM NADH, 3 mg/mL BSA, 0.300 mM KCN and 0.060 mM ubiquinone. Rotenone (10μM) was added to inhibit Complex I to allow subtraction of inhibitor-insensitive activity from the data. Complex II (succinate dehydrogenase) activity was measured spectrophotometrically at 600 nm (ε: 19.1 mM⁻¹ cm⁻¹) in potassium phosphate buffer (25 mM potassium phosphate dibasic with 25 mM potassium phosphate monobasic, pH: 7.5), 20 mM succinate, 0.080 mM 2,6-dichlorophenolindophenol (DCPIP), 1 mg/mL BSA, 0.300 mM KCN and 0.050 mM decylubiquinone (DUB). Malonate (10 mM) was added to inhibit complex II. Complex I and II activity were normalized to citrate synthase activity (CS), measured at 412 nm (ε : 13.6 mM⁻¹ cm⁻¹) in Tris 100 mM (pH: 8.0) buffer containing 0.1% Triton X-100, 0.100 mM 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB), 0.300 mM acetyl CoA, and 0.5 mM oxaloacetic acid. All assays were performed in triplicate on a Synergy H1 microplate reader at 37 °C in volumes of 200 μl/ well at a protein concentration of 0.05 mg/ml. The enzymatic activities were calculated as: nmol min⁻¹ mg⁻¹ = (Δ Absorbance/min × 1,000)/[(extinction coefficient × volume of sample) × (sample protein concentration)].

2.11. H₂O₂ production

Mitochondrial ROS generation was measured spectrofluorometrically (560 nm excitation and 590 nm emission) on a Synergy H1 microplate reader with the H₂O₂-sensitive dye Amplex Red (50 μM) (Invitrogen, A1222) and horseradish peroxidase (5 U/mL). Mitochondria (100 μg protein) were incubated in a solution containing (in mM): 120 KCl, 10 Tris-HCl, 5 MOPS, 5 K₂HPO₄ and 10 μM EGTA, pH 7.4. Baseline fluorescence was monitored for 1 minute, then 10 mM glutamate/5 mM malate was added and monitoring continued for 6 or more minutes. After subtraction of values from a blank well containing no mitochondria, slopes were calculated for the first 5 minutes after addition of glutamate/malate, and converted to units of pmol H₂O₂/(min*mg protein) using a linear fit to a standard curve of Amplex Red fluorescence across H₂O₂ concentrations. As positive and negative controls, antimycin A (500 nM) or 2,4-dinitrophenol (10 nM) respectively were present with a mitochondria sample in separate wells.

2.12. Protein carbonylation assay

The levels of protein carbonyl groups were assessed using the protein carbonyl assay kit (Abcam) according to manufacturer’s instructions. Briefly, freshly isolated mitochondria (100 μg) were diluted with 12% SDS and incubated with DNPH solution for 15 minutes at room temperature, then the reaction was neutralized. Gel electrophoresis (3 μg of protein per sample) and immunoblotting was performed as described under immunoblotting above with the anti-DNP antibody provided in the kit.

2.13. Transthoracic echocardiography

To measure left ventricular (LV) response to beta-adrenergic stimulation, we used serial transthoracic echocardiography to evaluate LV ejection fraction as a surrogate for LV function. We used a Vevo 2100 (Fujifilm VisualSonics, Toronto, CA) ultrasound imaging system to perform echocardiography. Mice were anesthetized with 2% isoflurane and affixed in a supine position to an ECG measuring platform heated to 37C. Ultrasound measurements were performed with a MS400 linear array ultrasound probe operating between 22–55 MHz. The ultrasound probe was placed across the mouse’s thorax to obtain a parasternal long axis view of the heart, and to capture two-dimensional and M-mode cine at mid-LV level. After acquisition of baseline images, at a heart rate between 400 and 550 bpm, mice were injected with isoproterenol (0.4 μg/kg) and serial images were taken every minute for five minutes. Echocardiography images were then analyzed using the Vevo Lab software. LV ejection fraction was calculated using the modified Quinones method from M-mode measurements of at least 3 different cardiac cycles from the parasternal long axis view.

2.14. Langendorff ex vivo ischemia/reperfusion (I/R)

The mice were euthanized using 100U of heparin and 250 mg/kg of sodium pentobarbital. The hearts were then harvested and perfused on the Langendorff apparatus using modified Krebs buffer (in mM): 118 NaCl, 0.5 EDTA, 4.7 KCl, 1.2 MgSO₄, 1.2 Na₂HPO₄, 15 glucose, 2.5 CaCl₂, 25 NaHCO₃, and 0.5 pyruvate. Oxygen probes (Microelectrodes, Inc.) were used throughout the protocol to monitor oxygen, sampling both the cardiac effluent and perfusate. The left ventricular pressure was measured using a balloon filled with water that was coupled to a solid-state pressure transducer (Millar Instruments). After 20 minutes of equilibration on the apparatus, the hearts were subjected to 20 minutes of no-flow global ischemia, followed by 90 minutes of reperfusion. Coronary flow was monitored by an ultrasonic flow probe (Transonic). At baseline the cannula oxygen levels were 524 ± 18 mmHg, while during hypoxia they were 175 ± 11 mmHg. At the start of reperfusion, effluent samples were collected to measure creatine kinase (CK) release, an indicator of heart damage. Heart rate and left ventricular developed pressure were also recorded to calculate the rate pressure product (RPP = HR × DP).

2.15. Data analysis

Data are expressed as mean ± standard deviation (SD). Unpaired Student’s t-test, one-way ANOVA or two-way ANOVA were performed when appropriate to compare experimental groups. A difference was considered statistically significant when p<0.05. Data processing and statistical tests were carried out with GraphPad Prism (v.9.3.1, San Diego, CA, USA).

3. Results

3.1. Emre deletion eliminates mitochondrial Ca²⁺ uptake

Emre^flox/flox (Emre^fl/fl) mice were crossed with the well-characterized αMHC-MerCreMer (MCM⁺) transgenic mouse line to generate a model of acute tamoxifen-inducible cardiomyocyte-specific Emre deletion in adulthood. We confirmed that 3 weeks post-tamoxifen injection, Emre mRNA expression decreased by 92.5% (Fig. S1A) and protein expression was reduced by 90% in Emre^cKO mitochondria relative to MCM⁺ and Emre^fl/fl controls (Fig. 1A). MCU protein expression also decreased but to a lesser extent (60% reduction), but no decrease was observed in Mcu mRNA expression (Fig. S1B), suggesting that in the absence of EMRE downregulation of MCU occurs not at the transcriptional level but potentially due to diminished protein stability. No differences were observed in the levels of other uniporter proteins, for instance MICU1 (Fig. 1A and Fig. S1C), nor proteins involved in mitochondrial Ca²⁺ efflux, including the Na⁺/Ca²⁺/Li⁺ exchanger (NCLX) and Leucine zipper EF-hand transmembrane protein 1 (Letm1) (Fig. S1D). Blue native polyacrylamide gel electrophoresis (BN-PAGE) analysis of cardiac mitochondria revealed that loss of EMRE decreased the molecular weight of the uniporter complex from over 440 kDa [23,38] to approximately 300 kDa, as reported previously [32]. To assess how Ca²⁺ handling is affected by EMRE deletion, we performed Ca²⁺ retention capacity (CRC), mitochondrial swelling, and membrane potential assays. Emre^cKO cardiac mitochondria were unable to take up Ca²⁺ in CRC assays, unlike control MCM⁺ and Emre^fl/fl mitochondria, in which normal mitochondrial Ca²⁺ uptake was observed (Fig. 1C–D, S1E–F). To determine whether Emre^cKO mitochondria can undergo mitochondrial Ca²⁺ overload-induced mitochondrial permeability transition pore (mPTP) opening, we analyzed mitochondrial swelling after adding a large Ca²⁺ bolus. Emre^cKO mitochondria were resistant to swelling relative to MCM⁺ and Emre^fl/fl controls (Fig. 1E, S1G), presumably due to loss of mitochondrial Ca²⁺ uptake. Similarly, in an assay using the dye safranin to monitor depolarization of mitochondrial membrane potential, Emre^cKO mitochondria were able to maintain mitochondrial membrane potential after addition of a large Ca²⁺ bolus, in contrast to MCM⁺ and Emre^fl/fl controls (Fig. 1F).

Figure 1. Short-term Emre deletion impairs mitochondrial Ca²⁺ uptake.

Analyses were performed 3 weeks post-tamoxifen. A) Representative Western blot of EMRE, MCU, and MICU1 in cardiac mitochondria from MCM⁺, Emre^fl/fl and Emre^cKO mice. VDAC was used as a loading control. Below, Western blot quantification shows a 90% reduction in EMRE and a 60% reduction in MCU in Emre^cKO mitochondria compared to control groups. Values represent mean ± SD.**p<0.0001, n = 7 in MCM⁺ and Emre^cKO, n = 5 in Emre^fl/fl. One-way ANOVA test was used for statistical analysis. B) Blue Native gel of freshly isolated mitochondria from Emre^fl/fl and Emre^cKO hearts, immunoblotted with MCU antibody. Gel is representative of 3 mice (combined males and females) per group. C) Representative Ca²⁺ retention capacity (CRC) assay in isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice. The fluorescent Ca²⁺ indicator Calcium Green-5N was used to monitor extramitochondrial Ca²⁺. The arrows represent 10 μM Ca²⁺ additions. Traces are representative of n = 5 independent experiments. D) Ca²⁺ retention capacity calculated from independent traces as shown in (C). The estimated mean Ca²⁺/mg protein was for MCM⁺: 166 ± 30.49, for Emre^fl/fl: 144 ± 28.89, and for Emre^cKO: 10 ± 5 nmol of Ca²⁺/mg protein. Values represent mean ± SD.**p<0.0001, n = 5 per group. One-way ANOVA test was used for statistical analysis. E) Representative mitochondrial swelling assay monitoring absorbance of isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice after addition of 400 μM Ca²⁺. Traces are representative of n = 5 independent experiments. F) Representative traces of safranin fluorescence as an indicator of membrane potential (△Ψ) depolarization in isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice after addition of 400 μM Ca²⁺. As a positive control in Emre^cKO mitochondria, depolarization was induced by addition of 1 μM FCCP. Traces are representative of n = 4 independent experiments. The mean time to Ca²⁺-induced depolarization, calculated as the time from Ca²⁺ addition to 50% of the rise in fluorescence, was for MCM⁺: 9.1 ± 2.8 minutes, for Emre^fl/fl: 5.9 ± 1 minute, and could not be calculated for Emre^cKO as FCCP was required for depolarization.

3.2. Emre deletion leads to lower intramitochondrial Ca²⁺ content and attenuated Ca²⁺-stimulated ATP production

Once we confirmed that cardiac-specific EMRE deletion eliminated mitochondrial Ca²⁺ uptake in heart mitochondria as expected, our next question was whether this deletion affects basal intramitochondrial Ca²⁺ (_mCa²⁺) levels. Relative to MCM⁺ and Emre^fl/fl controls, Emre^cKO mitochondria had lower _mCa²⁺ levels, as might be expected in the absence of mitochondrial Ca²⁺ uptake activity (Fig. 2A). As the dephosphorylation of PDH is _mCa²⁺-dependent [5], we next measured the ratio of phosphorylated pyruvate dehydrogenase (p-PDH) to total PDH in Emre^cKO heart mitochondria and found a ~20% increase compared with control mitochondria (Fig. 2B). These data support that loss of EMRE leads to decreased _mCa²⁺ levels. Because _mCa²⁺ has been shown to activate mitochondrial energy production [39], we sought to assess whether mitochondrial respiration is affected by Emre deletion. Seahorse analysis of isolated mitochondria supplied with either pyruvate and malate or palmitoylcarnitine and malate did not reveal significant differences in oxygen consumption rate (OCR) at baseline, after ADP addition, or after FCCP treatment to decouple mitochondria and promote maximal respiration (Fig. S2A–B). Respiratory control ratios (RCR) likewise were similar between control and Emre^cKO heart mitochondria for either substrate combination (Fig. S2C–D). These data suggest that, unlike Mcu deletion in heart and skeletal muscle [40–42], short-term Emre deletion did not lead to a shift toward for fatty acid oxidation in cardiac mitochondria. Consistently, protein levels of carnitine O-palmitoyl transferase 1b (CPT1) were similar across genotypes (Fig. S2E). Furthermore, though dysfunction of Complex I of the mitochondrial electron transport chain has been associated with increased uniporter stability [43,44], we did not find significant differences in the activity of either respiratory Complex I or Complex II in short-term Emre^cKO and control heart mitochondria (Fig. S2F–G). We then asked whether directly measuring ATP production rate in the presence and absence of Ca²⁺ would reveal Ca²⁺-induced energetic differences in Emre^cKO and control cardiac mitochondria. While the rate of ATP production upon providing ADP was similar in all groups, the addition of Ca²⁺ induced an increase in ATP production rate in control mitochondria but not in Emre^cKO mitochondria (Fig. 2C, S2H). These results suggest that the loss of EMRE prevents stimulation of ATP production by mitochondrial Ca²⁺ uptake.

Figure 2. Short-term Emre deletion leads to lower intramitochondrial Ca²⁺ content and attenuated Ca²⁺-stimulated ATP production.

Analyses were performed 3 weeks post-tamoxifen. A) Quantification of relative free _mCa²⁺. Emre^cKO mitochondria show a reduction in _mCa²⁺ by 56% relative to controls. n=4–8, *p<0.05. One-way ANOVA was used for statistical analysis. B) Representative Western blot of phosphorylated PDH (p-PDH) and total PDH from Emre^fl/fl and Emre^cKO mitochondria. COX IV was used as a loading control. Below, Western blot quantification shows a 25% increase in p-PDH to total PDH ratio in Emre^cKO relative to Emre^fl/fl mitochondria. n=6 per group. Student t-test was used for statistical analysis. C) Analysis of ATP production rate in isolated heart mitochondria stimulated with 50 μM ADP and 10 mM succinate ± 20 μM Ca²⁺. MCM⁺: 5.39 ± 0.43 vs MCM⁺+ 20 μM Ca²⁺: 7.59 ± 0.96, Emre^fl/fl: 6.35 ± 1.09 vs Emre^fl/fl + 20 μM Ca²⁺: 9.01 ± 1.25, and Emre^cKO: 5.46 ± 1.21 vs Emre^cKO + 20 μM Ca²⁺: 6.32 ± 0.78 nmol ATP/(min*mg) protein. Values represent mean ± SD. *p<0.05, n=4–7. One-way ANOVA was used for statistical analysis.

3.3. Short-term Emre deletion leads to reduced response to β-adrenergic stimulation

We next sought to determine whether cardiac-specific Emre deletion in adult mice had physiological consequences in the short term. At 3 weeks post-tamoxifen, Emre^cKO mice showed no overt phenotype and had similar body weights and heart weights as controls (Fig. S3A–B). Because the Ca²⁺-stimulated rate of ATP production was reduced in Emre^cKO cardiac mitochondria, we reasoned that loss of EMRE might attenuate an increase in cardiac energy production upon increased workload. We thus evaluated the ability of Emre^cKO mice to respond to β-adrenergic stimulation. Using echocardiography, we measured left ventricular ejection fraction (EF) and fractional shortening (FS) at baseline and then each minute up to 5 minutes after intraperitoneal injection of the β-adrenergic receptor agonist isoproterenol. In MCM⁺ and Emre^fl/fl mice, EF (Fig. 3A) and FS (Fig. 3B) increased substantially upon stimulation, but in Emre^cKO mice this response was more gradual. Although interplay between CaMKII and MCU in the context of adrenergic stimulation has been documented [45,46], we did not observe a difference in the phosphorylation status of CaMKII with Emre deletion (Fig. S3C). These results suggest that the absence of rapid mitochondrial Ca²⁺ uptake impairs the heart’s ability to respond to adrenergic stimulation, in agreement with previous studies investigating acute cardiac deletion of Mcu.

Figure 3. Short-term Emre deletion leads to reduced response to β-adrenergic stimulation.

Three weeks post-tamoxifen, echocardiography at baseline (time 0), then 1, 2, 3, 4 and 5 minutes after isoproterenol injection (0.4 μg/kg) was used to analyze A) ejection fraction (EF) and B) fractional shortening (FS). A) EF at minute 5: MCM⁺: 76.95% ± 10.38, Emre^fl/fl: 76.55% ± 6.11 and Emre^cKO: 63.70% ± 7.65. B) FS at minute 5: MCM⁺: 45.90% ± 11.39, Emre^fl/fl: 44.38% ± 5.45 and Emre^cKO: 34.01% ± 5.38. Values represent mean ± SD. *p<0.05, n=14–16 per group. Two-way ANOVA test was used for statistical analysis.

3.4. Short-term Emre deletion preserves cardiac function in I/R

So far, we have shown that the short-term deletion of Emre in the adult mouse heart is sufficient to block Ca²⁺ uptake into mitochondria and decrease _mCa²⁺ levels. Consistent with these effects, Emre^cKO mitochondria are resistant to mPTP opening following _mCa²⁺ overload. Cell death due to mPTP opening is thought to contribute to I/R injury; therefore, we evaluated whether Emre^cKO mice would show preserved cardiac function in an ex vivo Langendorff perfusion model of I/R injury. Our results show that during reperfusion Emre^cKO hearts maintained higher rate pressure product (RPP) relative to baseline pre-ischemic levels compared with control hearts, and this protection is even preserved until the end of the reperfusion period (Fig. 4A). Likewise, measures of systolic and diastolic pressure changes (dP/dt_max and dP/dt_min, respectively) were preserved to a greater degree during reperfusion in Emre^cKO mice compared to control mice (Fig. 4B–C). To determine the degree of injury in the hearts, we measured the release of creatine kinase (CK), a marker of cardiomyocyte membrane injury, and found reduced CK release in Emre^cKO mice relative to controls (Fig. 4D). Although Cyclophilin D (CypD) is a known activator of the mPTP, resistance to pore opening does not appear to be mediated by any decrease in CypD protein expression (Fig. S3D). Altogether, the data show that shortly after the loss of EMRE in adulthood, hearts are protected against I/R injury, consistent with the idea that preventing mitochondrial Ca²⁺ uptake during I/R prevents tissue damage.

Figure 4. Short-term Emre deletion preserves cardiac function in I/R.

Three weeks post-tamoxifen, mouse hearts were subjected to 20 minutes of stabilization, 20 minutes of global ischemia and 90 minutes of reperfusion. At 5, 30, 60, and 90 minutes after the onset of reperfusion, relative to baseline pre-ischemia, A) rate pressure product (RPP) = heart rate (HR) × left ventricular pressure (LVP), B) dP/dt_max, C) dP/dt_min, and D) creatine kinase (CK) release were assessed. Values at 90 minutes: A) RPP: MCM⁺: 50.96% ± 17.39, Emre^fl/fl: 44.55% ± 13.05 and Emre^cKO: 87.64% ± 26.68. B) dP/dt_max: MCM⁺: 48.50% ± 13.10, Emre^fl/fl: 47.14% ± 6.74 and Emre^cKO: 72.04% ± 16.62. C) dP/dt_min: MCM⁺: 48.53% ± 11.95, Emre^fl/fl: 47.64% ± 6.45 and Emre^cKO: 67.88% ± 14.23. Values represent mean ± SD. *p<0.05, n=5–6 per group. Two-way ANOVA test was used for statistical analysis.

3.5. Mitochondrial Ca²⁺ handling remains impaired in Emre^cKO mice over time

Thus far, we have confirmed that the short-term loss of EMRE in adulthood is consistent with many aspects of the Mcu^cKO phenotype, further solidifying that like MCU, EMRE is essential for uniporter activity. We next sought to explore why germline Mcu^−/− and Emre^−/− mice do not exhibit the impaired response to adrenergic stimulation and protection from I/R injury seen in the Mcu^cKO mice. This discrepancy has been attributed to remodeling in the chronic absence of uniporter activity, but whether long timescales alone are sufficient for remodeling is unclear. Therefore, we waited 3 months post-tamoxifen, which we will refer to as long-term Emre deletion, to reexamine the phenotype of Emre^cKO mice. We first measured the mRNA and protein expression of EMRE and MCU in isolated cardiac mitochondria to confirm that protein levels relative to controls remained significantly decreased (Fig. 5A), although as before Mcu mRNA is expressed at control levels (Fig. S4A–B). Expression of MICU1 as well as the efflux proteins NCLX and LETM1 were still unchanged between Emre^cKO and control mitochondria (Fig. 5A, S4C–D). Similarly, we verified that at 3 months post-tamoxifen the molecular weights of the uniporter complex with and without EMRE remained the same as at 3 weeks post-tamoxifen (Fig. 5B). Next, we conducted CRC, swelling, and membrane potential assays in Emre^cKO mitochondria 3 months post-tamoxifen and found that the loss of mitochondrial Ca²⁺ uptake and resistance to mPTP opening were maintained, as observed in short-term deletion (Fig. 5C–F, S4E–G).

Figure 5. Mitochondrial Ca²⁺ handling remains impaired with long-term Emre deletion.

Analyses were performed 3 months post-tamoxifen. A) Representative Western blot of EMRE, MCU, and MICU1 in cardiac mitochondria from MCM⁺, Emre^fl/fl and Emre^cKO mice. VDAC was used as a loading control. Below, Western blot quantification of mitochondria from MCM⁺ (n = 3), Emre^fl/fl (n = 6), and Emre^cKO (n = 6) hearts. Values represent mean ± SD. One-way ANOVA was used for statistical analysis. B) Blue Native gel of freshly isolated mitochondria from Emre^fl/fl and Emre^cKO hearts, immunoblotted with MCU antibody. Gel is representative of 3 mice (combined males and females) per group. C) Representative Ca²⁺ retention capacity (CRC) assay in isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice. The fluorescent Ca²⁺ indicator Calcium Green-5N was used to monitor extramitochondrial Ca²⁺. The arrows represent 10 μM Ca²⁺ added. Traces are representative of n = 5 independent experiments. D) Ca²⁺ retention capacity calculated from independent traces as shown in (C). The estimated mean Ca²⁺/mg protein was for MCM⁺ : 154 ± 30.50, Emre^fl/fl : 136 ± 27.01 and Emre^cKO : 6 ± 5.47 nmol of Ca²⁺ / mg protein. Values represent mean ± SD.**p<0.0001, n = 5 in each group. One-way ANOVA test was used for statistical analysis. E) Representative mitochondrial swelling assay monitoring absorbance of isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice after addition of 400 μM Ca²⁺. Traces are representative of n = 5 independent experiments. F) Representative traces of membrane potential (△Ψ) depolarization in isolated heart mitochondria from MCM⁺ (black line), Emre^fl/fl (purple line) and Emre^cKO (green line) mice after the addition of 400 μM Ca²⁺. As a positive control in Emre^cKO mitochondria, depolarization was induced by addition of 1 μM FCCP. Traces are representative of n = 4 independent experiments. The mean time to Ca²⁺-induced depolarization, calculated as the time from Ca²⁺ addition to 50% of the rise in fluorescence, was for MCM⁺: 10.5 ± 2 minutes, for Emre^fl/fl: 12.7 ± 2.3 minutes, and could not be calculated for Emre^cKO as FCCP was required for depolarization.

Similarly, measurements of free _mCa²⁺ levels in Emre^cKO mitochondria 3 months post-tamoxifen suggested that _mCa²⁺ remained decreased (Fig. 6A), consistent with significantly elevated p-PDH to total PDH ratio in Emre^cKO mitochondria (Fig. 6B). As found in the short-term, long-term Emre deletion did not alter OCR or RCR in either pyruvate/malate or palmitoylcarnitine/malate substrate conditions (Fig. S5A–D). Hence, long-term absence of Emre deletion did not appear to lead to a shift toward preferential fatty acid oxidation, and correspondingly CPT1 protein levels were unchanged (Fig. S5E). Complex II activity remained similar between groups 3 months post-tamoxifen; Complex I activity in Emre^cKO mitochondria exhibited a slight downward trend that however did not reach significance (p = 0.0735) (Fig. S5F–G). Additionally, though baseline ATP production rates were similar in Emre^cKO and control mitochondria, the increase in ATP production rate in the presence of Ca²⁺ was still absent in Emre^cKO mitochondria (Fig. 6C, S5H). Taken together, these data suggest that, at the level of isolated mitochondria, long-term absence of EMRE does not lead to substantially different effects compared to those of the short-term.

Figure 6. Lower intramitochondrial Ca²⁺ content and attenuated Ca²⁺-stimulated ATP production are maintained over long-term Emre deletion.

Analyses were performed 3 months post-tamoxifen. A) Quantification of relative free _mCa²⁺ in mitochondria from MCM⁺ (n = 3), Emre^fl/fl (n = 6), and Emre^cKO (n = 6) hearts. Emre^cKO mitochondria show a reduction in _mCa²⁺ by 40% ± 15.01 relative to controls. Values represent mean ± SD. One-way ANOVA was used for statistical analysis. B) Representative Western blot of phosphorylated PDH (p-PDH) and total PDH from Emre^fl/fl and Emre^cKO mitochondria. COX IV was used as a loading control. Below, Western blot quantification of mitochondria from MCM⁺ (n = 3), Emre^fl/fl (n = 6), and Emre^cKO (n = 6) hearts. Values represent mean ± SD. One-way ANOVA was used for statistical analysis. C) ATP production rate in isolated heart mitochondria stimulated with 50 μM ADP and 10 mM succinate ± 20 μM Ca²⁺. MCM⁺: 9.33 ± 2.65 vs MCM⁺+ 20 μM Ca²⁺: 13.41 ± 2.21, Emre^fl/fl: 10.73 ± 1.55 vs Emre^fl/fl + 20 μM Ca²⁺: 15.76 ± 1.90, and Emre^cKO: 6.63 ± 1.59 vs Emre^cKO + 20 μM Ca²⁺: 7.52 ± 1.40 nmol ATP/min*mg protein. Values represent mean ± SD. *p<0.05, n=4–7. One-way ANOVA was used for statistical analysis.

3.6. β-adrenergic response remains diminished in long-term Emre deletion, but protection against I/R injury is lost

After confirming that long-term Emre deletion revealed similar mitochondrial Ca²⁺ handling as in short-term Emre deletion, we sought to assess at 3 months post-tamoxifen how Emre^cKO mice responded to β-adrenergic stimulation. We first verified that long-term Emre deletion did not lead to any significant changes in body weight or heart weight (Fig. S6A–B). We then repeated the echocardiography experiment measuring ejection fraction and fractional shortening at baseline and after isoproterenol injection. Notably, in contrast to germline Emre or Mcu deletion, long-term cardiac-specific Emre deletion in adulthood still resulted in diminished adrenergic response (Fig. 7A–B). The phosphorylation status of CaMKII also remained similar between groups (Fig. S6C). This result suggests that the fight or flight response cannot be restored in a time-dependent manner in the chronic absence of rapid mitochondrial Ca²⁺ uptake ability. After confirming that β-adrenergic response remained blunted in mice with long-term cardiac Emre deletion, we next sought to determine whether protection from I/R injury is similarly preserved. We repeated the Langendorff ex vivo I/R experiment in 3-month post-tamoxifen Emre^cKO and control hearts. Surprisingly, and in contrast to short-term Emre deletion, we observed that long-term Emre deletion did not have a significant effect relative to controls on RPP, dP/dt_max, dP/dt_min, or CK release following reperfusion (Fig. 7C–F). This lack of protection in long-term Emre deletion could not be attributed to increased CypD expression, as CypD protein levels did not vary significantly between Emre^cKO and control mitochondria (Fig. S6D).

Figure 7. β-adrenergic response remains diminished in long-term Emre deletion, but protection against I/R injury is lost.

Three months post-tamoxifen, echocardiography at baseline (time 0), then 1, 2, 3, 4 and 5 minutes after isoproterenol injection (0.4 μg/kg) was used to analyze A) ejection fraction (EF) and B) fractional shortening (FS). A) EF at minute 5: MCM⁺(black line): 80.58% ± 6.19, Emre^fl/fl (purple line): 81.62% ± 5.69 and Emre^cKO (green line): 70.20% ± 11.08. B) FS at minute 5: MCM⁺: 48.42% ± 6.32, Emre^fl/fl: 49.80% ± 5.67 and Emre^cKO: 39.58% ± 9.14. Values represent mean ± SD. *p<0.05, n=10–12 per group. Two-way ANOVA was used for statistical analysis. Three months post-tamoxifen, mouse hearts were subjected to 20 minutes of stabilization, 20 minutes of global ischemia and 90 minutes of reperfusion. At 5, 30, 60, and 90 minutes after the onset of reperfusion, relative to baseline pre-ischemia, C) rate pressure product (RPP) = heart rate (HR) × left ventricular pressure (LVP), D) dP/dt_max, E) dP/dt_min, and F) creatine kinase (CK) release were assessed. Values represent mean ± SD. *p<0.05, n=5–6 per group. Two-way ANOVA was used for statistical analysis.

We then considered that apart from excessive mitochondrial Ca²⁺, reactive oxygen species (ROS) can also act as a trigger for mPTP opening [47]. Despite equivalent rates of ROS production in isolated Emre^cKO and control mitochondria in the short and long term as measured by Amplex Red (Fig. S7A–C), assessment of protein carbonylation revealed that Emre^cKO mitochondria have significantly less oxidative stress in the short-term than in the long-term (Fig. 8A–B). Furthermore, although oxidative stress levels between Emre^cKO and control mitochondria in the long-term are similar, in the short-term Emre^cKO mitochondria show a trend toward decreased oxidative stress relative to control mitochondria that approaches significance (p = 0.0548). These data suggest that a transient decline in oxidative stress in short-term but not long-term Emre deletion is associated with protection against I/R injury; over time, oxidative stress returns to control levels, and concurrently protection against I/R injury conferred by acute loss of mitochondrial Ca²⁺ uniporter activity is lost.

Figure 8. Short-term Emre deletion leads to lower oxidative stress levels than long-term Emre deletion.

A) Representative Western blot using anti-DNP antibody to detect carbonyl groups modified by DNPH as a marker of oxidized proteins in cardiac mitochondria from short-term and long-term Emre^fl/fl and Emre^cKO mice. VDAC was used as a loading control. B) Western blot quantification of mitochondria from short-term Emre^fl/fl: 1.00 ± 0.14 (n = 9), short-term Emre^cKO: 0.72 ± 0.19 (n = 9), long-term Emre^fl/fl: 1.05 ± 0.21 (n = 9), and long-term Emre^cKO: 1.21 ± 0.30 (n = 9). Values represent mean ± SD, *p<0.05. One-way ANOVA was used for statistical analysis.

4. Discussion

Here we present the novel finding that cardiac-specific Emre deletion in adult mice has distinct effects in the short and long term. While a blunted response to isoproterenol is preserved from 3 weeks to 3 months after Emre deletion, an initial protection against I/R injury is no longer observed after 3 months. This dichotomy exists despite our data showing that in isolated mitochondria at both time points Emre deletion eliminates mitochondrial Ca²⁺ uptake and reduces sensitivity to Ca²⁺-induced mPTP opening.

Lack of rapid mitochondrial Ca²⁺ uptake was also seen in mitochondria from mice with germline Mcu or Emre deletion as well as mice undergoing cardiac-specific Mcu deletion in adulthood [30,31]. Our results in a new cardiac-specific Emre knockout mouse model support the multiple studies showing that EMRE is essential to uniporter function [23,24,32]. Furthermore, like Mcu^−/− and Emre^−/− mice [28,32] but unlike Mcu^cKO mice [30,31], Emre^cKO mice have reduced cardiac intramitochondrial Ca²⁺ concentration, with lasting effects even 3 months after tamoxifeninduced deletion. In Emre^cKO mice, we also observe a large decrease in MCU expression, to a greater extent than observed in the germline Emre^−/− mice in which MCU decreased only slightly [32]. Our findings in these mice are thus not attributable solely to the absence of EMRE alone; rather, Emre^cKO mice represent another model in which the uniporter is nonfunctional. Nonetheless, in both Emre^cKO and germline Emre^−/− mice, and in the germline Mcu^−/− mice where EMRE was significantly decreased [28], the data suggest that a coordinated downregulation of MCU in the absence of EMRE and vice versa leads to lower intramitochondrial Ca²⁺ levels. As EMRE expression was not reported in Mcu^cKO mice, it is possible that lesser EMRE downregulation with MCU in these circumstances might explain why intramitochondrial Ca²⁺ was unchanged in Mcu^cKO mice [30,31].

Despite reduced basal intramitochondrial Ca²⁺ levels after short-term Emre deletion, Emre^cKO cardiac mitochondria produced ATP at the same rate as control mitochondria when supplied with ADP without Ca²⁺. This result echoes previous measurements of oxygen consumption or ATP production in Mcu^−/−, Emre^−/−, or Mcu^cKO cardiac mitochondria [28,30,32]. However, Ca²⁺-induced stimulation of ATP production was observed in control but not Emre^cKO and other uniporter-deficient cardiac mitochondria [28,30], consistent with the uniporter’s purported role in delivering Ca²⁺ to the matrix to provide a boost in ATP. Thus, in the absence of a functional uniporter, matching of energy supply to demand, for instance in the fight or flight response, might be expected to be attenuated. Consistent with this hypothesis, here we show with echocardiography that short-term deletion of Emre in adulthood, as was shown for Mcu deletion in in vivo hemodynamics experiments [30,31], resulted in impaired cardiac contractile response to adrenergic stimulation (see Table 1). In contrast, the hearts of global Mcu^−/− and Emre^−/− mice responded indistinguishably from wild-type mice to isoproterenol, measured in the former by cardiac catheterization [28] and in the latter by echocardiography [32]. Recent work has shown that Langendorff-perfused global Mcu^−/− hearts show no increase in mitochondrial Ca²⁺ following isoproterenol addition, yet contractility increases similarly in Mcu^−/− and wild-type hearts [48]. These results suggest that in the Mcu^−/− and Emre^−/− models, the chronic absence of uniporter activity beginning in the germline may have led to compensatory mechanisms that act to meet increased energy demand. However, what form this compensation takes or when it occurs are unclear. To address the latter question, we assessed Emre^cKO mice at a long time interval (3 months) after tamoxifen administration in adulthood and found that impaired adrenergic response persisted even after long-term EMRE deletion. Although it cannot be excluded that recovery of the rapid response to adrenergic stimulation could gradually take place at timescales longer than 3 months after the loss of uniporter activity starting in adulthood, one possibility is that restoration of the adrenergic response in global Mcu^−/− and Emre^−/− mice does not occur merely in a time-dependent manner and instead may require the absence of uniporter function during embryogenesis. Notably, the expression of DN-Mcu driven by the αMHC promoter, which is expressed in the myocardium shortly after birth [49], also results in reduced response to isoproterenol [29]. Taken together, one might speculate that the onset of uniporter deficiency at any time after birth stably attenuates the ability of the mitochondria to respond to a sudden increase in energy demand in the fight or flight response. It remains to be tested whether compensation for uniporter loss in adulthood can occur gradually over timescales longer than 3 months or whether compensation is only possible when uniporter loss is experienced early in development in conditions of germline Mcu or Emre deletion.

The inactivation of mitochondrial Ca²⁺ uptake due to Emre deletion would be predicted to protect against I/R injury, and in line with this reasoning, at 3 weeks post-tamoxifen, Emre^cKO mice were able to better maintain cardiac function compared with controls in an ex vivo Langendorff I/R model. These results are consistent with studies in Mcu^cKO mice undergoing in vivo I/R [30,31], demonstrating that protection from I/R can manifest in both in vivo and ex vivo models (see Table 1). Hence, this experimental distinction is not the cause of divergent results in Mcu^cKO mice (shown in vivo to be protected from I/R) compared with Mcu^−/−, Emre^−/−, and DN-Mcu mice (shown ex vivo to be unprotected). Moreover, it was previously unclear whether the lack of protection against I/R injury in Mcu^−/−, Emre^−/−, and DN-Mcu mice arose from compensatory cellular pathways in response to chronic uniporter loss, and if so whether these pathways are only active when the uniporter is absent in development or are activated as function of time. Our data support the latter hypothesis, as by 3 months post-tamoxifen, Emre^cKO hearts exhibited a similar decrease in cardiac function during I/R as control hearts, suggesting that the protection from I/R injury evident at 3 weeks post-tamoxifen was lost in the intervening time. We conclude that although mice gain protection from I/R injury after acute deletion of Mcu or Emre in adulthood, this effect appears to be transient. Within a few months, Emre^cKO hearts, from which isolated mitochondria are still unable to rapidly take up Ca²⁺, experience I/R-induced damage to roughly the same extent as control hearts. Furthermore, concurrently with I/R protection in short-term Emre deletion, protein carbonylation potentially decreases but later recovers to control levels in the long term. This interesting phenomenon argues that time-dependent compensation for uniporter loss, potentially due to a return to baseline oxidative stress levels, is sufficient to restore susceptibility to I/R, but the nature of this compensation remains incompletely resolved.

We examined multiple cellular pathways known to be linked to MCU activity to determine whether or not in the short- or long-term absence of EMRE these pathways would be altered. Loss of MCU has been shown to alter metabolic substrate preference in the heart and skeletal muscle, in which glucose oxidation is impaired or unchanged while fatty acid oxidation is favored [40–42]. Our data suggest that OCR in the presence of substrates for either glucose oxidation or fatty acid oxidation is similar between control and Emre^cKO heart mitochondria in both the short-and long-term. However, it was noted within one study that although OCR measured in Langendorff-perfused and paced hearts increased in DN-Mcu mice, Seahorse analysis of control and DN-Mcu isolated mitochondria showed no differences [29]. Hence, though we were unable to detect intrinsic mitochondrial differences, further study is needed to uncover metabolic alterations, including potentially fuel preference, in animals with short- and long-term Emre deletion.

Additionally, it has been shown that during chronic β-adrenergic stimulation, CaMKIIδB activation upregulates Mcu expression at the transcriptional level [45]. However, in the context of both short- and long-term EMRE loss our data show that while MCU protein decreases, CaMKII phosphorylation as well as Mcu mRNA levels remain unchanged. Indeed, the modulation of uniporter activity was shown to be downstream of the CaMKII pathway, and not necessarily vice versa. Similarly, decreased Complex I activity in multiple contexts was shown to lead to compensatory increase in MCU levels and activity [43,44]; whether alterations in MCU activity can reciprocally impact Complex I is less clear. Mild decreases in Complex I and II protein levels were previously observed following acute isoproterenol challenge in Mcu^cKO hearts 3 months post-tamoxifen [40]. In our study, though no statistically significant decrease in Complex II activity occurred in short- or long-term Emre deletion, and short-term Emre^cKO and control Complex I activity were also similar, at the long-term timepoint Complex I activity trended downwards in Emre^cKO mitochondria relative to controls, nearly achieving significance. Since it is thought that Complex I and MCU interact such that mild ROS leak from Complex I oxidizes MCU [44], one might speculate that gradually over time, a prolonged absence of this interaction in Emre^cKO mice could result in even slightly increased oxidative damage to Complex I itself that may decrease its activity. The interaction of Complex I and the uniporter complex, and how perturbation thereof might directly and indirectly affect sensitivity to I/R injury, is amply deserving of future study.

Furthermore, we observed that protein carbonylation is transiently decreased in the short-term absence of EMRE relative to long-term EMRE absence. These data, which temporally correlate with the transient protection from I/R injury, are notable given the well-documented involvement of ROS in opening the mPTP [47]. A gradual decline in Complex I activity and potential uptick in ROS leakage may play a part in the resurgence of oxidative stress in the long-term absence of EMRE, yet what mediates the transient decrease in oxidative damage is still an open question. Which proteins undergo less or more oxidation in the short or long term is also important: for instance, oxidation of CypD is reported to target it to the mitochondrial inner membrane where it can activate mPTP [50]. The germline loss of MCU has been suggested to lead to alterations in the post-translational modification status of CypD that prime mitochondria toward mPTP opening [51]. More detailed study in Emre^cKO mice of time-dependent CypD expression and modifications, particularly oxidation, is warranted. Finally, cardiomyocytes that chronically lack a functional uniporter may undergo a cell death pathway in I/R distinct from mPTP-triggered necroptosis, given evidence that the CypD inhibitor Cyclosporine A is protective in I/R in wild-type hearts but not Mcu^−/− or Emre^−/− hearts [26,32]. What this cell death pathway or pathways might be or what might lead to its activation during the chronic absence of uniporter activity is still unclear.

In summary, we have reported the characterization of a novel tamoxifen-inducible cardiac-specific mouse model of Emre deletion, with which we have examined mitochondrial Ca²⁺ handling, response to adrenergic stimulation, and susceptibility to I/R injury at short (3 week) and long (3 month) post-tamoxifen time points. In most respects long-term Emre deletion closely resembled short-term Emre deletion. In isolated mitochondria, the differences between Emre^cKO and controls at 3 weeks post-tamoxifen – lack of mitochondrial Ca²⁺ uptake, increased p-PDH to total PDH ratio, reduced Ca²⁺ levels, resistance to Ca²⁺-induced swelling, and attenuated Ca²⁺-stimulated ATP production – were still present at 3 months post-tamoxifen. We found that Emre deletion in both the short- and long-term impaired isoproterenol-induced increase in contractility; however, only short-term Emre deletion was protective against reduced cardiac function during I/R, potentially due to reduced oxidative stress that appears restricted to the short-term time point. This discrepancy underscores that the compensatory mechanism that allows only mice lacking a functional uniporter starting in the germline to retain their adrenergic response is fundamentally different from the mechanism that in the span of a few months following uniporter loss in adult mice restores sensitivity to I/R. The nature of these uniporter deficiency-induced pathways and the cellular changes they elicit remain a complex problem to be addressed in future research.

Highlights.

• A tamoxifen-inducible, cardiac-specific mouse model of Emre deletion was generated

• Mitochondrial Ca²⁺ uptake is eliminated after short and long-term EMRE loss

• Response to adrenergic stimulation is blunted after short and long-term EMRE loss

• Short-term EMRE loss is protective against ischemia/reperfusion (I/R) injury

• Protection against I/R injury is lost in the long-term absence of EMRE