Inhibition of the mPTP and Lipid Peroxidation Is Additively Protective Against I/R Injury

Abstract

Background:

During myocardial ischemia/reperfusion (I/R) injury, high levels of matrix Ca²⁺ and reactive oxygen species (ROS) induce the opening of the mitochondrial permeability transition pore (mPTP), which causes mitochondrial dysfunction and ultimately necrotic death. However, the mechanisms of how these triggers individually or cooperatively open the pore has yet to be determined.

Methods:

Here, we use a combination of isolated mitochondrial assays and in vivo ischemia reperfusion surgery in mice. We challenged isolated liver and heart mitochondria with Ca²⁺, ROS and Fe²⁺ to induce mitochondrial swelling. Using inhibitors of the mPTP (cyclosporine A (CsA) or ADP) lipid peroxidation (Fer-1, MitoQ) we determined how the triggers elicit mitochondrial damage. Additionally, we used the combination of inhibitors during I/R injury in mice to determine if dual inhibition of these pathways is additivity protective.

Results:

In the absence Ca²⁺, we determined that ROS fails to trigger mPTP opening. Instead, high levels of ROS induce mitochondrial dysfunction and rupture independently of the mPTP through lipid peroxidation (LIPOX). As expected, Ca²⁺ in the absence of ROS induces mPTP-dependent mitochondrial swelling. Sub-toxic levels of ROS and Ca²⁺ synergize to induce mPTP-opening. Furthermore, this synergistic form of Ca²⁺- and ROS-induced mPTP opening persists in the absence of cyclophilin D (CypD), suggesting the existence of a CypD-independent mechanism for ROS sensitization of the mPTP. These ex vivo findings suggest that mitochondrial dysfunction may be achieved by multiple means during I/R injury. We determined that dual inhibition of the mPTP and LIPOX is significantly more protective against I/R injury than individually targeting either pathway alone.

Conclusions:

In the present study, we have investigated the relationship between Ca²⁺ and ROS and how they individually or synergistically induce mitochondrial swelling. Our findings suggest that Ca²⁺ mediates mitochondrial damage through the opening of the mPTP, While ROS mediates its damaging effects through LIPOX. However, sub-toxic levels both Ca²⁺ and ROS can induce mPTP-mediated mitochondrial damage. Targeting both of these triggers to preserve mitochondria viability unveils a highly effective therapeutic approach for mitigating I/R injury.

Graphical Abstract

INTRODUCTION

Over 800,000 people in the United States have a heart attack each year.¹ During myocardial ischemia, an obstructed artery in the myocardium results in reduced blood supply and subsequent oxygen deprivation, which causes dysregulation of Ca²⁺ homeostasis as ATP-dependent ion channels become dysfunctional.² This results in toxic levels of Ca²⁺ ions entering into the mitochondrial matrix.³ In order to reduce the amount of ischemic damage, blood flow is restored to the affected myocardium by reperfusion therapy. Although this treatment is overall beneficial, this sudden increase in oxygenated blood generates a paradoxical form of cardiac injury caused by oxidative stress which contributes to the overall infarct size.^{4, 5} Many forms of cell death have been implicated in ischemia/reperfusion (I/R) injury, however, the prominent death-inducing agents are toxic levels of Ca²⁺ and reactive oxygen species (ROS), both of which are well-established triggers of mitochondrial permeability transition pore (mPTP).⁶

The mPTP is defined as a non-selective, voltage-dependent pore within the inner-membrane of mitochondria that allows for the passive diffusion of solutes up to 1.5 kDa in size.⁷ Following the opening of the pore: mitochondria membrane potential dissipates lose cristae density and swell, and mitochondrial-dependent ATP production is lost.⁷ Unfortunately, the molecular identity of the mPTP remains to be fully elucidated.^{8, 9} Currently, the known mPTP regulators are the cyclophilin D (CypD) within the matrix of the mitochondria, the adenine nucleotide translocator (ANT) family within the inner mitochondrial membrane, and the Bcl-2 family, on the outer mitochondrial membrane.¹⁰ Indeed, targeting CypD during I/R either genetically (Ppif^−/−) or pharmacologically using a CypD inhibitor known as cyclosporine A (CsA) results in reduced infarct size in mice.^11-13 The ANT family of proteins exist within the inner mitochondrial membrane to shuttle ADP and ATP into and out of the organelle.^{11, 14} Similarly, mitochondria lacking the ANT family or treated with ADP are desensitized to mPTP opening.^{11, 15, 16} Recently it has been shown that mice lacking the predominant form of the ANT family in skeletal muscle are protected from necrotic cell death in muscular dystrophy.¹⁷ Bax and Bak are responsible for increasing outer mitochondrial membrane permeability during apoptosis.¹⁸ Furthermore, deletion of Bax and Bak desensitizes the mPTP to Ca²⁺ and reduces infarct size following I/R injury.¹⁸ Despite the promising results of targeting the mPTP to reduce infarct size in animal models, CsA was unable to significantly reduce I/R injury following myocardial infarction (MI) in humans in various clinical trials.^19-21 This lack of translatable protection may be explained by potential compensatory forms of cell death during I/R, as many forms of cell death have been hypothesized to occur during this injury.²² Therefore, investigating possible alternative cell death mechanisms that may compensate during I/R injury must be explored in order to identify more efficacious and translatable therapeutic strategies.

Even though Ca²⁺ and ROS are known triggers of the mPTP, the mechanisms of how each of these triggers open the mPTP is not understood.²³ Elucidating the relationship between each trigger and their ability to mediate mitochondrial damage would provide valuable insight to the progression of I/R injury, since it is accepted that reducing either Ca²⁺ and ROS overload reduces infarct size following I/R injury.²⁴ Alternatively to mPTP opening, ROS has been implicated in mediating mitochondria damage through lipid peroxidation (LIPOX). LIPOX is the downstream effector of the ferroptosis pathway. Ferroptosis is a form of regulated necrosis that proceeds in an iron-dependent manner by overwhelming the endogenous glutathione-dependent hydroperoxidase that converts lipid peroxides into non-toxic lipid alcohols. Within the mitochondrial matrix, Fe²⁺ can readily convert respiratory chain by-products such as H₂O₂ into ROS such as hydroxyl radical via the Fenton reaction.^{25, 26} This reaction positively contributes to ferroptosis by leading to the oxidation of polyunsaturated fatty acids within lipid bilayers. This LIPOX renders membranes permeable as radical lipids attack neighboring lipids eventually leading to loss of membrane integrity.^{27, 28} Indeed, synthetic antioxidant pharmaceutical agents such as Fer-1 and MitoQ can inhibit LIPOX propagation and block ferroptosis.^29-32 Inhibition of ferroptosis by Fer-1 or MitoQ has been shown to significantly reduce infarct size in murine hearts.^32,33-36

In this manuscript, we explored the relationship between the triggers of I/R injury, Ca²⁺ and ROS, and how they independently or synergistically regulate mitochondrial damage through mPTP opening or LIPOX. We found that ROS, in the absence of Ca²⁺, does not trigger mPTP opening, however, high levels of ROS induce mitochondrial rupture through LIPOX. Furthermore, sub-toxic levels of ROS lower the threshold of Ca²⁺ required to trigger the mPTP. Additionally, we elucidated the contributions of mPTP-dependent necrosis and ferroptosis to I/R injury. We conclude that dual inhibition of the mPTP and LIPOX through CsA and MitoQ respectively, leads to a significant reduction in infarct size compared to single inhibition. Our study substantiates a model where both Ca²⁺ and ROS can contribute to mitochondrial dysfunction through two distinct pathways or through synergistic means via desensitization of the mPTP. Inhibiting both factors could hold clinical significance, potentially mitigating cardiac damage in the aftermath of an ischemic event.

MATERIALS AND METHODS

Data Availability.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Animal Models.

Wild type (WT) C57BL/6J and Ppif^−/−, the gene that encodes CypD, 3 month old male and female mice were utilized for the mitochondrial swelling and capacity assays. Ppif^−/− mice were previously generated as described.³⁷ Mice were kept on a 12 hour light/dark cycle and were allowed free access to food and water. Baylor College of Medicine (BCM) follows all guidelines and regulations regarding the use of animals. All studies performed were approved by the Institutional Animal Care and Use Committee (IACUC) which oversees all institutional animal care with the goal of conserving animal well-being, and enforces that investigators comply with federal law as well as institutional policies. WT mice were euthanized as described in the approved protocol and equal numbers of male and female mice were used for all experiments.

Heart and Liver Mitochondrial Isolation.

Whole heart and liver was isolated from WT and Ppif^−/− mice and washed in cold isolation buffer (225 mM mannitol, 75 mM sucrose, 5 mM HEPES, 1 mM EDTA, pH 7.4). Heart or liver was minced into 1 mm pieces in 7 mL isolation buffer. Atria were removed from the heart and ventricles were minced into 1mm pieces in 2 mL isolation buffer. Heart and liver tissues were then homogenized using a Teflon/glass homogenizer (approximately 10 strokes) on ice. Liver and heart homogenates were then centrifuged at 800 rcf for 5 min at 4°C. The supernatant was collected and centrifuged at 10,000 rcf for 10 min at 4°C. Then the supernatant was aspirated and the pellet was suspended in 7 mL isolation buffer and centrifuged again at 10,000 rcf for 10 min at 4°C. The final pellets are suspended in 0.5-1 mL KCL buffer (125 mM KCL, 20 mM HEPES, 2 mM KH₂PO₄, 40 μM EGTA, pH 7.2). Mitochondrial sample concentration was determined by protein concentration using a NanoDrop.

Mitochondrial Swelling, Ca²⁺ Retention Capacity Assays, or mitochondrial LIPOX assays.

The mitochondrial swelling assay and the Ca²⁺ retention capacity assay has been previously described in detail.³⁸ For analysis of mitochondrial swelling, isolated mitochondria is loaded into a quartz cuvette containing a final volume of 1 mL physiological containing KCL buffer 1 mM malate (Sigma-Aldrich), 7 mM Pyruvate (Sigma-Aldrich), and 10 μM BODIPY^™ 581/591 C11 (Invitrogen). Several assays also included inhibitors such as 300 μM ADP (Sigma-Aldrich, A2754), 2 μM CsA (Sigma-Aldrich, 30024), 10 μM Fer-1 (Sigma-Aldrich, SML0583), 10-400 μM MitoQ (Selleckchem, S8978), 5 mM EDTA (Sigma-Aldrich), and 5 mM NAC (Sigma-Aldrich). Two mg of mitochondria were used per fluorometric assay. Loaded cuvettes equilibrate for approximately 5 min prior to beginning the assay using Felix Software. The settings of light scattering on the Felix software were edited depending on the dye parameters (Ex/Em 488/520 for BODIPY-C11). A fluorimeter (PTI QuantaMaster 800, Horiba Scientific) was utilized to measure transmittance and fluorescence simultaneously. Sequential additions of 20 μM - 40 μM calcium chloride (CaCl₂) (Sigma-Adrich, C4901), 40 μM iron (II) chloride (FeCl₂) (Sigma-Aldrich, 372870), and 500 μM - 200 mM 70% tBHP (Acros Organics, 180340050) were added to the cuvette. Delta absorbance was determined by subtracting the end of each run from the baseline absorbance (the absorbance prior to the addition of mitochondrial swelling inducers).

Ischemia/Reperfusion Injury drug administration and surgery.

Mice were injected intraperitoneally with 10 mg/kg CsA (2.5% 100mg/mL CsA in DMSO, 29.5% PEG 300, 5% Tween-80, 63% saline) or vehicle (2.5% DMSO, 29.5% PEG 300, 5% Tween-80, 63% saline) for four consecutive days (24 hr apart per injection). On the fourth day of injections, mice were subjected to ischemia reperfusion surgery. The I/R injury surgical model in mice has been described previously.³⁹ Mice were anesthetized with isofluorane and prior to ventral incision, a local anesthetic was applied (Lidocaine/bupivacaine). A 1 cm incision was performed on the ventral side of the neck, in which an intubation tube is inserted into the trachea. Left lateral thoracotomy (1 cm incision) was performed and a slip knot 8-0 nylon suture was tied around the anterior surface of the left ventricle to ligate the proximal left anterior descending (LAD) coronary artery. Following an ischemic period of 60 min, the knot is untied. Then, mice undergo a 24 hour reperfusion period prior to heart collection. The ischemic region and infarct were subsequently identified by post-mortem staining with Evan’s blue and 2,3,5-triphenyltetrazolium chloride, respectively. Animals that did not survive the procedure were not included in the study.

Transmission Electron Microscopy.

Mitochondria were isolated as previously described in this section. In a cuvette, 2 mg mitochondria were loaded into the KCL-assay buffer. After treating the sample with CaCl₂, FeCl_2, tBHP until swelling is observed, the contents of the cuvette were collected and mitochondria were pelleted at 5000 rcf for 5 min at 4°C. Pelleted mitochondria were then submerged in cold Karnovsky’s fixative (3% glutaraldehyde and 3% paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.4) for overnight incubation at 4°C. Further sample preparation, cross sectioning, and electron microscope imaging was performed by the Biological Electron Microscopy Lab at Rice University.

Statistical Analysis.

Statistical analyses were conducted using GraphPad Prism. Results are represented as the average with standard error of the mean (SEM) for the indicated number of samples. All replicates presented are biological replicates. The difference of the mean between multiple groups was compared using one-way ANOVA followed by a Dunnett’s test when comparing to one control group or a Bonferroni post-test analysis when all groups were compared. P-values of <0.05 or less were considered statistically significant as indicated within the figure legends. Sample size was predetermined by first establishing normality using previously generated large data sets then performing power analyses through Power AX. The analyses showed 80% power for detection of differences in fluorometric assays with 3 animals per group and differences in infarct size with 12 animals per group. No experiment-wide/across-test multiple test correction was applied.

RESULTS

Ca²⁺, ROS, and Fe²⁺ all are able to induce mitochondrial swelling

In order to analyze the mitochondrial dysfunction induced by Ca²⁺, ROS, or Fe²⁺ we first determined the appropriate concentrations necessary to cause mitochondrial swelling using heart and liver mitochondria. WT mitochondria were challenged with Ca²⁺, ROS (tert-butyl hydroperoxide (tBHP)), or Fe²⁺ in a dosage-dependent manner to determine the adequate level of each trigger necessary to induce mitochondrial swelling (Figure 1A through 1F) and the results were quantified (Figure S1A through S1F). We determined that mitochondria isolated from hearts were more resistant to swelling induced by tBHP than liver mitochondria; however, at higher concentrations, tBHP was able to reproducibly cause heart mitochondrial swelling.

Figure 1. Heart mitochondria are more resistant to tBHP- but not Ca²⁺- or Fe²⁺-induced mitochondrial swelling compared to liver mitochondria.

A and B, Representative mitochondrial swelling assays (n=3) using 2 mg of mitochondria isolated from WT heart (A) and liver (B) treated with a range of total Ca²⁺ concentrations as labeled. Each arrow indicates a bolus of 1/3 of the total Ca²⁺ C and D, Representative mitochondrial swelling assays (n=3) using 2 mg of mitochondria isolated from WT heart (C) and liver (D) treated with a range of tBHP concentrations as labeled. E and F, Representative mitochondrial swelling assays (n=3) using 2 mg of mitochondria isolated from WT heart (E) and liver (F) treated with a range of total Fe²⁺ concentrations as labeled. Each arrow indicates a bolus of 1/4 of the total Fe²⁺.

Mitochondrial swelling induced by ROS and Fe²⁺ leads to irreversible mitochondrial rupture

Mitochondrial swelling induced by Ca²⁺ is less severe compared to tBHP or Fe²⁺, as demonstrated visually by transmission electron microscopy and quantitatively by the decrease in relative absorbance (Figure 2A through 2E). Additionally, Ca²⁺-dependent mPTP opening is reversible upon the addition of Ca²⁺ chelating agent EDTA, as shown by a positive increase in absorbance following swelling, however, tBHP-induced mitochondrial swelling was not reversible with the addition of ROS scavenger, N-acetyl-l-cysteine (NAC) (Figure 2F through 2H). Importantly, pretreatment with EDTA or NAC was sufficient to block all swelling induced by Ca²⁺ or tBHP, respectively (Figure 2F and 2G). Both the TEM results and the irreversible nature of ROS induced swelling implicate mitochondrial rupture.

Figure 2. Ca²⁺-mediated mitochondrial dysfunction is distinct from ROS-mediated dysfunction.

A, Representative images of transmission electron microscopy images of isolated WT heart mitochondria treated with 120 μM Ca²⁺ incubated for 10 minutes, 160 μM Fe²⁺ incubated for 50 minutes, 200 mM tBHP incubated for 50 minutes, or untreated (control). Images were selected based on most consistent display of mitochondrial morphology B, Representative mitochondrial swelling assays (n=3) using heart mitochondria treated with 120 μM Ca²⁺, 160 μM Fe²⁺, or 200 mM tBHP. C, Quantification of change in absorbance from (A). D, Representative mitochondrial swelling assays (n=3) using liver mitochondria treated with 120 μM Ca²⁺, 160 μM Fe²⁺, or 60 mM tBHP. E, Quantification of change in absorbance from (D). F, Representative mitochondrial swelling assays (n=3) using WT liver mitochondria treated with or without EDTA (5 mM) before (pre-treatment) and after (post-treatment) the treatment of Ca²⁺ (120 μM). G, Representative mitochondrial swelling assays (n=3) using WT liver mitochondria treated with or without NAC (5 mM) before (pre-treatment) and after (post-treatment) the treatment of tBHP (60 mM). H, Quantification of mitochondrial shrinkage were calculated by subtracting the absorbance value at the end of the assay from the absorbance value at the time of the post-treatment of EDTA or NAC from (F and G).

ROS is not a trigger of the mPTP.

Ca²⁺ and ROS have been previously established as the main triggers of mPTP opening.^{8, 24, 40} To determine if mitochondrial dysfunction induced by either Ca²⁺ or ROS was indeed through mPTP opening, we performed mitochondrial swelling assays with established inhibitors of the mPTP. To inhibit the mPTP we utilized CsA, which inhibits CypD, a well-established regulator of the mPTP, and ADP, which desensitizes the pore through the ANT family. WT heart mitochondria were incubated with either CsA or ADP prior to challenging with consecutive boluses of Ca²⁺ or tBHP, and subsequent swelling was quantified (Figure 3A through 3D). As expected, Ca²⁺ induced mitochondrial swelling was mitigated by the addition of either CsA or ADP (Figure 3A and 3B), however, ROS-mediated mitochondrial swelling was refractory to both mPTP inhibitors (Figure 3C and 3D). To establish generality, similar experiments and analyses were performed using WT liver mitochondria. Indeed, tBHP induced liver mitochondrial swelling was not reduced by the addition of CsA or ADP unlike Ca²⁺-induced swelling (Figure 3E through 3F). To determine if these results were specific to this singular concentration of tBHP (60 mM) we performed a similar dose response of tBHP as in Figure 1D in the presence of both CsA and ADP and observed no change in mitochondrial swelling at any dose of tBHP tested (Figure S2A and S2B). In addition, the classical mPTP inhibitors did not affect Fe²⁺-mediated swelling in either heart or liver mitochondria (Figure S2A through S2D). Together, these data indicate that Ca²⁺ induces mitochondrial dysfunction via the mPTP, whereas tBHP and Fe²⁺ mediate mitochondrial dysfunction in an mPTP-independent manner.

Figure 3. tBHP-induced mitochondrial swelling is not prevented by mPTP inhibition.

A, Mitochondrial swelling assays of 120 μM total Ca²⁺ treated WT heart mitochondria with or without mPTP desensitizers 2 μM CsA or 300 μM ADP. B, Quantification of maximum swelling (delta absorbance) from (A). C, Mitochondrial swelling assays of 200 mM tBHP treated WT heart mitochondria with or without mPTP desensitizers 2 μM CsA or 300 μM ADP. D, Quantification of maximum swelling from (C). E, Mitochondrial swelling assays of 120 μM total Ca²⁺ treated WT liver mitochondria with or without mPTP desensitizers 2 μM CsA or 300 μM ADP. F, Quantification of maximum swelling from (E). G, Mitochondrial swelling assays of 60 mM tBHP treated WT liver mitochondria with or without mPTP desensitizers 2 μM CsA or 300 μM ADP. H, Quantification of maximum swelling from (G).

tBHP- and Fe²⁺-dependent mitochondrial swelling is dependent on LIPOX.

To determine how tBHP and Fe²⁺ mediate mitochondrial swelling, we examined the contribution of LIPOX. We utilized a membrane-localized LIPOX indicator, BODIPY-C11, which when excited fluorescently emits a detectable wavelength which signals the active oxidation of membrane.⁴¹ In addition to mPTP inhibitors, we also utilized established antioxidant drugs that are known to inhibit ferroptosis, ferrostatin-1 (Fer-1) and Mitoquinone (MitoQ), by preventing the continuous propagation of LIPOX. We incubated WT heart mitochondria with BODIPY-C11, with or without inhibitors of the mPTP (CsA, ADP), or inhibitors of LIPOX (Fer-1, MitoQ) prior to challenging the mitochondria to boluses of Ca²⁺, tBHP, or Fe²⁺ to induce swelling (Figure 4A and 4B). Both tBHP and Fe²⁺, but not Ca²⁺, led to significant increases in BODIPY-C11 fluorescence (Figure 4A and 4B). Additionally, BODIPY-C11 excitement was inhibited by the addition of Fer-1 and MitoQ, but not by CsA or ADP (Figure 4A and 4B). Furthermore, these LIPOX inhibitors had no effect on Ca²⁺-dependent mitochondrial swelling, but significantly blocked mitochondrial swelling induced by tBHP and Fe²⁺ in samples isolated from WT hearts (Figure 4C through 4H). These experiments were consistently recapitulated using WT liver mitochondria (Figure S3A through S3H).

Figure 4. LIPOX inhibitors are effective against ROS but not Ca²⁺-induced mitochondrial swelling.

A, Representative trace of BODIPY-C11 excitation assays (n=3) of 200 mM tBHP treated WT heart mitochondria with or without mPTP desensitizers or LIPOX inhibitors (2 μM CsA and 300 μM ADP, or 10 μM Fer-1 and 400 μM MitoQ). B, Representative trace of BODIPY-C11 excitation assays (n=3) of 160 μM total Fe²⁺ or 120 μM total Ca²⁺ treated WT heart mitochondria, with or without mPTP desensitizers or LIPOX inhibitors (2 μM CsA and 300 μM ADP, or 10 μM Fer-1 and 10 μM MitoQ). C, Representative mitochondrial swelling assays (n=3) of 120 μM total Ca²⁺ treated WT heart mitochondria with or without LIPOX inhibitors (10 μM Fer-1 and 10 μM MitoQ). D, Quantification of maximum swelling (delta absorbance) of (C). E, Representative mitochondrial swelling assays (n=3) of 200 mM tBHP treated WT heart mitochondria with or without LIPOX inhibitors 10 μM Fer-1 and 400 μM MitoQ. F, Quantification of maximum swelling of (E). G, Representative mitochondrial swelling assays (n=3) of 160 μM total Fe²⁺ treated WT heart mitochondria with or without LIPOX inhibitors 10 μM Fer-1 and 10 μM MitoQ. H, Quantification of maximum swelling of (G).

Sub-toxic levels of ROS and Ca²⁺ synergize to induce mPTP-dependent mitochondrial swelling.

Previous studies have shown that ROS and Ca²⁺ together are able to sensitize mPTP opening to induce mitochondrial dysfunction.^{6, 42-45} To determine if these triggers can indeed synergize in combination, we treated mitochondria isolated from WT mouse liver with sub-toxic concentrations of Ca²⁺ (45 μM) and/or tBHP (10 mM). Independently, these low concentrations of Ca²⁺ or tBHP were not sufficient to elicit mitochondrial swelling, however, in combination they induce mitochondrial swelling (Figure 5A and 5B). Additionally, we determined that this synergistic effect was dose-dependent, as varying levels of tBHP (250 μM to 20 mM) were able to synergize with a constant level of sub-toxic Ca²⁺ (45 μM) to induce mitochondrial swelling in a concentration-dependent fashion (Figure 5C). Conversely, when keeping the level of tBHP constant (10 mM) and utilizing varying dosages of Ca²⁺ (10 - 80 μM) in combination, we found that Ca²⁺ also dose-dependently affected synergistic swelling (Figure 5D). We performed a similar set of experiments using heart mitochondria. We determined that a higher level of tBHP is required to induce a synergistic swelling response in hearts with a constant level of Ca²⁺ (45 μM), likely due to the resistance of heart mitochondria to tBHP-dependent swelling as observed in initial experiments (Figure 5E). Furthermore, we observed synergistic swelling using this constant sub-toxic level of tBHP (50 mM) while increasing the concentration of Ca²⁺ (20–80 μM) (Figure 5F).

Figure 5. Sub-toxic levels of ROS and Ca²⁺ synergize to induce mitochondrial swelling.

A, Representative mitochondrial swelling assays (n=3) of WT liver mitochondria treated with 10 mM tBHP and/or 45 μM Ca²⁺. B, Quantification of maximum swelling of (A). C, Dosage response of mitochondrial swelling assays using varying concentrations of tBHP (ranging from 250 μM- 20 mM) in combination with a constant level of 45 μM Ca²⁺ on WT liver mitochondria. D, Representative dosage response (n=3) of mitochondrial swelling assays using varying concentrations of Ca²⁺ (ranging from 10 μM- 80 μM) in combination with a constant level of 10 mM tBHP on WT liver mitochondria. E, Representative dosage response (n=3) of varying concentrations of tBHP (ranging from 10 mM- 200 mM) in combination with a constant concentration of 45 μM Ca²⁺ on WT heart mitochondria. F, Representative dosage response (n=3) of varying concentrations of Ca²⁺ (ranging from 20 μM-80 μM) in combination with a constant concentration of 50 mM tBHP on WT heart mitochondria.

To determine if the synergistic effect by Ca²⁺ and ROS was dependent on the mPTP or LIPOX, we incubated WT liver mitochondria with mPTP inhibitors CsA, ADP, and CsA + ADP, or with LIPOX inhibitors Fer-1, or MitoQ prior to challenging them with previously determined sub-toxic levels of Ca²⁺ and tBHP. Notably, CsA, CsA + ADP, and MitoQ significantly reduced the synergistic swelling to the greatest extent (Figure 6A and 6B). Additionally, Fer-1 significantly reduced swelling to a lesser degree probably due to the nature of the mechanism of action between the LIPOX inhibitors. Surprisingly, ADP has no significant effect on this type of swelling (Figure 6A and 6B). To determine if the mPTP or LIPOX contributed to the synergistic swelling observed in heart mitochondria, we performed similar experiments with CsA and MitoQ (Figure 6C and 6D). Levels of Ca²⁺ and ROS were chosen based on previous experiments which established sub-toxic levels that produced a substantial swelling response in combination (Figure 5F). Similar to liver mitochondria, CsA and MitoQ were both able to mitigate this form of swelling, with CsA exhibiting a stronger inhibitory effect.

Figure 6. ROS and Ca²⁺ synergistic swelling is dependent upon mPTP opening.

A, Representative mitochondria swelling assays (n=3) of WT liver mitochondria treated with sub-toxic levels of ROS and/or Ca²⁺ (10 mM tBHP and 45 μM Ca²⁺), with or without mPTP desensitizers (2 μM CsA or 300 μM ADP), or LIPOX inhibitors (10 μM Fer-1 or 200 μM MitoQ). B, Quantification of maximal swelling from (A). C, Representative mitochondria swelling assays (n=3) of WT heart mitochondria treated with sub-toxic levels of ROS and/or Ca²⁺ (50 mM tBHP and 60 μM Ca²⁺), with or without mPTP desensitizer (2 μM CsA), or LIPOX inhibitors (200 μM MitoQ). D, Quantification of maximal swelling from (C). E, Representative mitochondria swelling assays (n=3) using CypD null liver mitochondria treated with sub-toxic levels ROS and/or Ca²⁺ (500 μM tBHP and 240 μM Ca²⁺), with or without 300 μM ADP, 10 μM Fer-1 or 200 μM MitoQ. F, Quantification of maximal swelling from (E).

Cysteine modifications on CypD have been previously reported to alter mPTP sensitivity.⁴⁰ Specifically, Ca²⁺-dependent mitochondrial swelling is desensitized when cysteine 203 of CypD is mutated to serine (C203S), suggesting that ROS may be acting through CypD to synergize with Ca²⁺ to trigger the mPTP.⁴⁶ To determine if CypD is required for ROS/Ca²⁺ synergistic swelling, we isolated liver mitochondria from Ppif^−/− mice and subjected them to low Ca²⁺ and tBHP (Figure 6E). Since these mitochondria are resistant to Ca²⁺-mediated mPTP opening compared to WT, we performed a dose response to identify a sub-toxic level of Ca²⁺ that was unable to induce mitochondrial swelling alone (240 μM Ca²⁺) (Figure S4A and S4B). Notably, we also performed a dose response of tBHP using Ppif^−/− liver and heart mitochondria and found their sensitivity to tBHP-dependent swelling and LIPOX to be similar to WT mitochondria (Figure S4C through S4F). Indeed, sub-toxic tBHP and Ca²⁺ induced synergistic mitochondrial swelling in the absences of CypD (Figure 6E and 6F). Notably, this swelling was completely blocked by the addition of ADP, but was unaffected by the LIPOX inhibitors suggesting that this synergistic swelling is mPTP-dependent (Figure 6E and 6F). Together, these findings suggest that ROS-dependent mPTP sensitization to Ca²⁺ may involve additional protein modifications beyond CypD.

Dual inhibition of mPTP-dependent necrosis and LIPOX additively protects against cardiac I/R injury.

Our ex vivo data suggest that ROS or Ca²⁺ can mediate mitochondrial dysfunction through LIPOX or mPTP opening, respectively. Here we translated these results to in vivo I/R injury. Individually, inhibition of mPTP-dependent necrosis by CsA or ferroptosis by vitamin E, Fer-1, and MitoQ have all been reported to reduce cardiac infarct size in mice subjected to I/R injury.^{13, 32, 36, 47} To determine whether the therapeutic effects of CsA and MitoQ are additively protective, we performed myocardial I/R injury (1 hr ischemia, 24 hr reperfusion) on mice treated with vehicle, CsA (10 mg/kg), MitoQ (5 mg/kg), or in combination (10 mg/kg CsA + 5 mg/kg MitoQ) at various time points prior to surgery (Figure 7A). Quantification of area at risk demonstrated a similarity in the affected area of ischemia between the different cohorts (Figure 7B). Indeed, individual treatment of CsA or MitoQ significantly reduced infarct size, however, mice treated with both inhibitors exhibited significantly greater protection from I/R injury (Figure 7C and 7D). These findings indicate that both forms of cell death play distinct roles in the development of the infarct. Moreover, they suggest that a more potent therapeutic approach involves targeting both ROS-LIPOX-dependent and Ca²⁺-mPTP-dependent mitochondrial dysfunction, as opposed to focusing on a single inhibition method.

Figure 7. Dual inhibition of mPTP-dependent necrosis and LIPOX additively protects against cardiac I/R injury.

A, Schematic of I/R injury and processing with pre-treatment intraperitoneal (IP) injection timing of Veh, 10 mg/kg CsA, 5 mg/kg MitoQ, or in combination. B, Quantitation of the area at risk (AAR). C, Representative cross sectional images of TCC and Evan’s blue dye stained hearts from each labeled cohort. Infarcts are circled by white dotted lines. Images are representative of the mean infarct sizes. D, Quantification of infarct region (IR) over AAR (n= 12 per cohort).

DISCUSSION

ROS and Ca²⁺ have been implicated as interchangeable triggers of the mPTP, however the mechanisms of how ROS or Ca²⁺ induce mPTP opening remain elusive. The fundamental questions on whether or not ROS contributes to pore opening independently of Ca²⁺, or if these triggers can synergize to induce mitochondrial dysfunction remain controversial.⁴⁰ Previous work has demonstrated that Ca²⁺ is sufficient to induce mPTP opening independently of ROS, which can be blunted by CsA and/or ADP.^{48, 49} Other studies have suggested that ROS can independently induce mPTP opening, which may be prevented by mPTP inhibitors (CsA, bongkrekic acid, trifluoperazine), or antioxidants (trolox, deferoxamine).^50-55 Alternatively, previous work has shown that ROS can synergistically sensitize mPTP opening in the presence of Ca²⁺ or other stimuli such as FCCP, inorganic phosphate, or oxidized pyridine nucleotides.^{6, 42-45} Additionally, ROS has been determined to sensitize Ca²⁺-dependent mPTP opening by S-nitrosylation of cysteine 203 within CypD, which is thought to be a potential trigger site for mPTP activation.⁴⁶ Our data demonstrates that ROS, in the absence of Ca²⁺, is not a trigger of the mPTP. Instead, we found that ROS mediates mitochondrial damage via LIPOX. Furthermore, sub-toxic levels of ROS are able to sensitize the mPTP to low Ca²⁺ levels. Notably, this synergistic effect was present in mitochondria lacking CypD, suggesting that ROS-dependent mPTP sensitization involves protein modifications beyond CypD. These data support a model where high levels of ROS, devoid of Ca²⁺ or in the context of mPTP inhibition, can lead to LIPOX and subsequently ferroptosis; contrarily, low levels of ROS can sensitize the mPTP to Ca²⁺, thus leading to mPTP-dependent necrosis at lower Ca²⁺ levels than would be normally required (Figure 8).

Figure 8. The roles of Ca²⁺ and ROS in the promotion of mitochondrial dysfunction during I/R injury.

Schematic representation of our hypothesis on the concentration-dependent contribution of Ca²⁺ and/or ROS in mPTP- or LIPOX-dependent mitochondrial dysfunction. High levels of Ca²⁺ in the absence of ROS, induces mPTP opening regulated by both CypD and ANTs. High ROS levels in the absence of Ca²⁺, induces LIPOX- dependent mitochondrial rupture. Low levels of ROS can sensitize the mPTP to Ca²⁺-dependent mPTP opening. These diverse forms of mitochondrial stress each contribute to cardiomyocyte death during I/R injury and dual inhibition of these pathways is additively protective.

We observed a tissue specific difference between liver and heart mitochondria treated with tBHP. We found that mitochondria isolated from heart require over 3-fold more tBHP to induce swelling compared to liver mitochondria. Additionally, the swelling traces in the heart seemed to be biphasic and the LIPOX inhibitors, which were able to block tBHP-mediated liver mitochondrial swelling were only able to inhibit the second phase of swelling that tBHP induced in heart mitochondria. Thus, suggesting that the first phase of heart mitochondrial swelling is independent of LIPOX. Perhaps the extremely high levels of tBHP required to induce heart mitochondria swelling may cause unregulated mitochondrial damage to a subset of mitochondria within a micro-domain adjacent to the point of injection in the cuvette explaining the first uninhibited phase of swelling. Another difference we observed was between Ca²⁺ and ROS-dependent mitochondrial swelling. The extent and reversibility of the swelling observed by TEM and spectrometry was vastly different between the two triggers. We find that LIPOX-dependent mitochondrial damage results in the destruction of mitochondria whereas mPTP-dependent swelling seems to plateau before the point of rupture making it a reversible event. This may give credence to the physiological nature of the mPTP, as depolarization through the mPTP results in the inhibition of both ATP generation and ROS production. Perhaps in some contexts the mPTP plays a protective role by extending the potential recovery time of the mitochondrion if the stress subsides.

Throughout the manuscript, both CsA (CypD inhibitor) and ADP (ANT substrate) are used as desensitizers of the mPTP. Recently, the combined addition of both desensitizers have been shown to completely block Ca²⁺-dependent mPTP opening.¹¹ This and other genetic proof led to the hypothesis that there may be multiple pore forming components of the mPTP. One being composed of the ANT family, which may work independently of CypD and the other being an unknown entity that is dependent on CypD to open. Interestingly, only CsA and not ADP was able to inhibit Ca²⁺ and ROS synergistic swelling in WT liver mitochondria, suggesting that this form of mPTP opening relies on the undefined CypD-dependent pore. Correspondingly, treatment with CsA was also able to mitigate synergistic swelling in heart mitochondria. The previously described cysteine modification of CypD that sensitizes the mPTP may explain this phenomenon. However, in the absence of CypD, ROS and Ca²⁺ synergistic swelling persists and is now mitigated by ADP. These results suggest that when CypD is absent, ROS-mediated mPTP sensitization may function through the ANT family. Further examination of cysteine residues within the ANT family need to be explored in order to elucidate this mechanism. Conserved cysteines (C₅₇, C₁₆₀, and C₂₅₇) within the matrix loops of the ANTs have been suggested to be sites targeted by oxidative stress that facilitate a pore open state.⁵⁶ These may be the sites that are responsible for ROS sensitization to Ca²⁺-dependent mPTP opening in the absence of CypD. In addition, MitoQ was able to significantly reduce synergistic swelling in WT heart and liver mitochondria. Perhaps mPTP sensitization by cysteine modification on CypD is prevented through the general ROS scavenging properties of MitoQ, or a low level of LIPOX is sufficient to lower the Ca²⁺ needed to trigger the pore.⁵⁷ Interestingly, Fer-1 was able to reduce synergistic swelling to a lesser extent than MitoQ, which suggests that inhibition of LIPOX is slightly protective against this stimuli. Previous studies have shown a potential connection between mPTP opening and lipid peroxidation.^58,59 Further investigation is essential to clarify the intricate relationship between these necrotic insults, particularly within context-dependent nuances.

During I/R injury Ca²⁺ accumulates in the mitochondrial matrix, which results in mPTP opening and mPTP-dependent necrotic cell death.⁵⁸ Indeed inhibitors of CypD are effective at reducing infarct size in animal models of I/R injury.¹¹ However, CypD inhibition by CsA failed to translate in clinical trials.^{19, 20} Therefore, more effective therapeutic approaches to reduce cardiomyocyte cell death are needed to potentially overcome the clinical hurdles of this spontaneous and variable injury. During reperfusion, the sudden reoxygenation of the O₂ deprived tissue leads to the production of ROS.⁵ Previous studies have shown that mitochondrial depolarization occurs during reperfusion, which suggests that ROS is triggering factor of mitochondrial dysfunction during this injury model.^{52, 59} Another source of ROS during I/R is from myocardial iron (Fe²⁺) levels that increase within the area of myocardium effected by I/R injury due to the lysis of hemoglobin-rich red blood cells caused by reperfusion treatment.⁶⁰ Our data suggests that mPTP-dependent necrosis and ferroptosis additively contribute to infarct size following I/R injury. Indeed, the inhibition of mPTP-dependent necrosis by CsA and ferroptosis by MitoQ additively reduces infarct size in mouse hearts exposed to I/R injury. Our ex vivo data suggest that ROS can lead to LIPOX or synergize with Ca²⁺ to induce mitochondrial dysfunction, therefore, inhibition of mPTP opening may not completely block all of the regulated forms of necrosis in the infarcted heart. Alternatively, targeting ROS alone, does not account for ROS-independent Ca²⁺-dependent mitochondrial dysfunction via mPTP opening. Thus, inhibition of both ROS and Ca²⁺-dependent mitochondrial dysfunction is an ideal therapeutic approach to preserve mitochondrial integrity and function in the face of I/R injury, which leads to enhanced cardiomyocyte survival and reduced infarct size (Figure 8).

CONCLUSION

We have found that the balance between Ca²⁺ and ROS concentrations can dictate which necrotic cell death pathway (mPTP vs. ferroptosis) becomes initiated at the level of the mitochondria during I/R injury (Figure 8). When isolated heart and liver mitochondria are treated with Ca²⁺, mitochondrial dysfunction occurs through mPTP opening. However, when mitochondria isolated from the hearts and livers were challenged with ROS (tBHP) or Fe²⁺ in the absence of Ca²⁺, mitochondria dysfunction was LIPOX-dependent and mPTP-independent. Additionally, sub-toxic levels of ROS and Ca²⁺ can synergize and lead to mitochondrial dysfunction in an mPTP-dependent manner. Further mechanistic insight into how Ca²⁺ triggers mPTP opening and how ROS sensitizes this triggering event is needed. In vivo, dual inhibition of mPTP-dependent necrosis and ferroptosis is additively protective against I/R injury. It is clear that multiple forms of regulated necrotic cell death are engaged during I/R injury. Determining the most efficacious therapeutic inhibitory combination to mitigate the forms of cell death that can independently occur or that may compensate for the one another is required in order for this treatment approach to translate to human disease given the complexities and spontaneity of myocardial infarctions.

The Novelty and Significance Section.

What is known?

• Calcium (Ca²⁺) and reactive oxygen species (ROS) are two major contributing factors that lead to mitochondrial dysfunction and cell death and may contribute to heart disease.

• Ca²⁺ and ROS are considered to be triggers of the mitochondrial permeability transition pore (mPTP), which prolonged opening leads to mitochondrial swelling and abolishes ATP production.

• How Ca2+ and ROS trigger mPTP opening independently or synergistically is unknown.

What new information does this article contribute?

• ROS is not an actual trigger of the mPTP and mPTP opening is absolutely dependent on Ca²⁺.

• ROS, in the absence of Ca²⁺ leads to mPTP-independent lipid peroxidation (LIPOX)-dependent mitochondrial swelling/dysfunction.

• ROS is able to reduce the Ca²⁺ concentration that is required to activate the mPTP in the heart.

• ROS/Ca²⁺ synergistic swelling occurs in the absence of known mPTP regulator Cyclophilin D (CypD).

• Dual inhibition of mPTP-dependent mitochondrial damage and LIPOX-dependent mitochondrial damage is additively protective against cardiac ischemia reperfusion injury in mice.

In this study we investigate the capability and mechanisms of ROS and Ca²⁺ to cause mitochondrial damage in the heart. Toxic levels of ROS and Ca²⁺ are the two major contributing factors that lead to heart cell death during and after cardiac ischemia. It was thought that ROS and Ca²⁺ are both triggers of the mPTP, however, our data show that ROS in the absence of Ca²⁺, is unable to trigger mPTP opening. Instead, ROS is able to elicit mitochondrial damage through increasing LIPOX. Additionally, we show that iron facilitates mitochondrial damage similarly to ROS and that sub-toxic levels of ROS and Ca²⁺ are able to synergize and trigger mPTP opening. CypD is a known regulator of the mPTP and it is thought that cysteine modifications by ROS within CypD can sensitize the mPTP to Ca²⁺. We determined that this sub-toxic synergistic effect of Ca²⁺ and ROS to trigger mPTP opening is present in the CypD null background. Finally, we show that inhibition of mPTP through CypD inhibitor cyclosporine A and inhibition of mitochondrial ROS through mitoquinone, a mitochondrial-targeted antioxidant, is additively protective at reducing infarct size following ischemia reperfusion injury in mice.

Non-standard Abbreviations and Acronyms

ANT

adenine nucleotide translocator

CsA

cyclosporine A

CypD

cyclophilin D

Fer-1

ferrostatin-1

I/R

ischemia reperfusion

LIPOX

lipid peroxidation

MI

myocardial infarction

MitoQ

mitoquinone

mPTP

mitochondrial permeability transition pore

NAC

N-acetyl-l-cysteine

ROS

reactive oxygen species

tBHP

tert-butyl hydroperoxide