Assessment of Open Probability of the Mitochondrial Permeability Transition Pore in the Setting of Coenzyme Q Excess

The mPTP mediates the permeability transition (PT), whereby the IMM becomes abruptly permeable to small molecules and solutes^1,2. This striking phenomenon is a distinct departure from the characteristic impermeability of the IMM, which is fundamental for establishing the electrochemical gradient necessary for oxidative phosphorylation³. PT, unlike other mitochondrial transport mechanisms, is a high-conductance, nonspecific, and nonselective process, allowing the passage of a range of molecules up to 1.5 kDa^4,5. The mPTP is a voltage-gated channel within the IMM whose opening alters ΔΨ, ATP production, calcium homeostasis, reactive oxygen species (ROS) production, and cell viability⁴.

At the pathologic extreme, uncontrolled and prolonged high-conductance opening of mPTP leads to the collapse of the electrochemical gradient, matrix swelling, depletion of matrix pyridine nucleotides, outer membrane rupture, release of intermembrane proteins (including cytochrome c), and ultimately, cell death^4,6. Such pathological mPTP opening has been implicated in cardiac ischemia-reperfusion injury, heart failure, traumatic brain injury, various neurodegenerative diseases, and diabetes^1,7. However, low-conductance mPTP opening is physiological in nature and, in contrast to high-conductance opening, does not lead to profound depolarization or mitochondrial swelling⁴.

Low-conductance opening of the pore restricts permeability to ~300 Da, allows the passage of protons independent of ATP synthesis, and is a potential source of physiological proton leak⁵. Physiologic mPTP opening causes a controlled decline in ΔΨ, increases electron flux through the respiratory transport chain, and results in a short burst or flash of superoxide, contributing to ROS signaling⁸. Regulation of such transient mPTP opening is important for calcium homeostasis and normal cellular development and maturation^4,9,10,11. Transient pore opening in developing neurons, for example, triggers differentiation, while the closure of the mPTP induces maturation in immature cardiomyocytes^4,5.

Although the functional significance of the mPTP in health and disease is well established, its precise molecular identity remains debated. Progress on the molecular structure and function of the mPTP has been comprehensively reviewed elsewhere¹². Briefly, currently, high- and low- conductance states of the mPTP have been hypothesized to be mediated by distinct entities¹². The leading candidates are the F1/F0 ATP synthase (ATP synthase) and adenine nucleotide transporter (ANT) for high- and low-conductance modes, respectively¹².

Despite the lack of consensus regarding the exact identity of the pore-forming component of the mPTP, certain key characteristics have been detailed. A well-established feature of the mPTP is that it is regulated by the electrochemical gradient such that depolarization of the IMM leads to pore opening¹³. Prior work has shown that the redox state of vicinal thiol groups alters the voltage gating of the mPTP, such that oxidation opens the pore at relatively higher ΔΨs, and thiol group reduction results in closed mPTP probability¹⁴. However, the identity of the proteinaceous voltage sensor is unknown.

Various small molecules that modulate the open probability of the pore have been identified. For example, the mPTP can be stimulated to open with calcium, inorganic phosphate, fatty acids, and ROS and can be inhibited by adenine nucleotides (particularly ADP), magnesium, protons, and CsA^5,12. The mechanisms of action of some of these regulators have been elucidated. Mitochondrial calcium triggers mPTP opening at least in part by binding to the β-subunit of the ATP synthase¹⁵. ROS can activate the mPTP by decreasing its affinity for ADP and enhancing its affinity for cyclophilin D (CypD), the best-studied proteinaceous mPTP activator¹⁶. The mechanism of activation of the mPTP by inorganic phosphate and fatty acids is less clear. As for endogenous inhibitors, ADP is thought to inhibit the mPTP by binding at the ANT or ATP synthase, while magnesium exerts its inhibitory effect by displacing calcium from its binding site^15,17,18,19.

Low pH inhibits mPTP opening by protonating histidine 112 of the regulatory oligomycin sensitivity-conferring protein (OSCP) subunit of the ATP synthase^12,20,21. The prototypical pharmacologic inhibitor of the mPTP, CsA, acts by binding CypD and preventing its association with OSCP^22,23. Previous work has also shown that a variety of CoQ analogs interact with the mPTP, inhibiting it or activating it²⁴. In recent work, we found evidence of a pathologically open mPTP, excessive proton leak, and inefficient oxidative phosphorylation due to a CoQ deficiency in forebrain mitochondria of newborn FXS mouse pups²⁵.

Closure of the pore with exogenous CoQ blocked the pathologic proton leak and induced morphologic maturity of dendritic spines²⁵. Interestingly, in the same animals, FXS cardiomyocytes had excessive CoQ levels and closed mPTP probability compared to wildtype controls²⁶. Although the cause of these tissue-specific differences in CoQ levels is unknown, the findings underscore the concept that endogenous CoQ is likely a key regulator of the mPTP. However, there is a major gap in our knowledge because the mechanism of CoQ-mediated inhibition of the mPTP remains unknown.

Regulation of the mPTP is a critical determinant of cell signaling and survival⁴. Thus, detecting mPTP opening within mitochondria is key when considering specific pathophysiological mechanisms. Typically, the threshold for high-conductance pore opening is determined using calcium to trigger the permeability transition. Such calcium loading leads to the collapse of the membrane potential, rapid uncoupling of oxidative phosphorylation, and mitochondrial swelling^27,28. We sought to develop a method to detect low-conductance mPTP opening in situ, without inducing it per se.

The approach exploits the role of the mPTP as a proton leak channel. To do so, Clark-Type and TPP⁺ ion-selective electrodes were employed to simultaneously measure oxygen consumption and membrane potential, respectively, in isolated mitochondria during leak respiration²⁹. The threshold for mPTP opening was determined by the onset of CsA-mediated inhibition of proton leak at specific membrane potentials. Using this approach, differences in voltage gating of the mPTP in the context of CoQ excess were precisely defined.

Institutional Animal Care and Use Committee of Columbia University Medical Center approval was obtained for all methods described. FXS (Fmr1 KO) (FVB.129P2-Pde6b⁺ Tyr^c-ch Fmr1^tm1Cgr/J) and control (FVB) (FVB.129P2-Pde6b⁺ Tyr^c-ch/AntJ) mice used as the model systems for this study were commercially acquired (see the Table of Materials). Five to eleven animals were used in each experimental group. Postnatal day 10 (P10) mice were used to model for a time point in human infancy.

1. Mitochondrial isolation from mouse heart

1. Prepare buffers for mitochondrial isolation and respiration experiments as described in Table 1. Store at 4 °C if made in advance.
   1. Prepare 15% density gradient medium: dilute commercial density gradient medium to 80% volume/volume (v/v) with density gradient diluent. Further dilute 80% density gradient medium 15% v/v by adding mitochondrial isolation buffer (MI)/bovine serum albumin (BSA).
2. Isolate mitochondria from fresh, not frozen tissue for these experiments and use on the day of preparation, typically within five hours²⁶. Carry out all isolation steps on ice.
   1. Decapitate the mouse and excise the heart. Immediately wash 1-2 times in a Petri dish with ice-cold MI/BSA. Cut off the atrium and mince the tissue.
   2. Transfer the minced heart tissue to a glass-glass homogenizer containing 1 mL of MI/BSA. Homogenize with a looser (A) pestle (10 strokes), then a tighter (B) pestle (10 strokes).
   3. Centrifuge the homogenate at 1,100 × g for 2 min at 4 °C to remove nuclear and cellular debris.
   4. Take the supernatant gently without touching any fluffy pellet, and carefully layer it on top of 700 μL of 15% density gradient medium in a centrifuge tube. Centrifuge at 18,500 × g for 15 min at 4 °C.
   5. Resuspend the pellet in 1 mL of sucrose buffer (SB)/BSA and centrifuged at 10,000 × g for 10 min at 4 °C.
   6. Completely remove and discard the supernatant. Resuspend the pellet in SB/BSA to a final volume of 55 μL.
   7. Quantify mitochondrial protein content using a standard assay such as the bicinchoninic acid (BCA) protein assay.

2. Mitochondrial oxygen (O₂) consumption and ΔΨ

1. Oxygen electrode assembly and calibration
   1. Set up and calibrate the oxygen electrode (see the Table of Materials) according to the manufacturer's instructions.
      1. Place a drop of 50% potassium chloride (KCl) electrolyte solution on top of the dome of the electrode disc.
      2. Place a small piece ~2 cm² cigarette paper spacer covered with a slightly larger piece of the provided polytetrafluoroethylene (PTFE) membrane over the electrolyte drop.
      3. Using the applicator tool provided, push the small electrode disc O-ring over the dome of the electrode.
         NOTE: Ensure that there are no air bubbles and that the membrane is smooth.
   2. Top up the reservoir well with the electrolyte solution.
   3. Place the larger O-ring in the recess around the electrode disc well.
   4. Install the disc into the electrode chamber and connect it to the control unit.
   5. Add 2 mL of air saturated deionized water to the reaction chamber and the PTFE-coated magnet to the chamber.
   6. Connect the chamber to the rear of the control unit.
   7. Set the temperature to 37 °C and the stirring speed to 100.
   8. Allow 10 min for the system temperature to equilibrate before commencing calibration.
   9. Under the Calibration tab, select Liquid Phase Calibration to perform liquid phase calibration. Confirm that the temperature is set to 37 °C , the stirring speed is 100, and the pressure is set to atmospheric pressure (101.32 kPa).
   10. Press OK and wait for the signal to plateau.
   11. When a plateau is reached, press OK.
   12. Establish zero O₂ in the chamber by adding a small amount (~20 mg) of sodium dithionite.
   13. Press OK and wait for the signal to plateau.
   14. When a plateau is reached, press Save to accept the calibration.
2. TPP⁺-selective electrode assembly
   1. Fill the TPP⁺- selective electrode tip with 10 mM TPP solution using the provided syringe and flexible needle, taking care to avoid air bubbles.
   2. Assemble the TPP⁺-selective electrode apparatus, including the reference electrode and electrode holder, (see Table of Materials) according to the manufacturer's instructions.
   3. Briefly, loosen the electrode holder cap and insert the internal reference electrode into the TPP⁺ tip. Tighten the cap to secure the tip. Connect the provided cable to the electrode holder and the auxiliary port of the control box.
   4. Insert the TPP⁺-selective electrode and reference electrodes into the adapted plunger assembly for ion-selective electrodes. Connect the reference electrode to the reference port of the control box.
   5. Insert the TPP⁺-selective and reference electrodes into the adapted plunger assembly for ion-selective electrodes.
3. Reaction chamber preparation
   1. Add the reaction mixture to the reaction chamber: 10 mM succinate (complex II substrate), 5 μM rotenone (complex I inhibitor), 80 ng mL⁻¹ nigericin (to collapse pH gradient across IMM), 2.5 μg mL^-1 oligomycin (to induce state 4 respiration), and respiration buffer (RB)/BSA reaction buffer to a final volume of 1 mL. Take care to avoid introducing air bubbles into the chamber.
   2. Close the chamber with the adapted plunger assembly with the TPP⁺-selective and the reference electrode in place. Select GO to start recording. Once the chamber is closed, introduce additional reagents directly into the reaction solution using separate microsyringes modified with plastic tubing to adjust the needle length.
4. TPP⁺ calibration
   1. Once a stable voltage signal is obtained, calibrate the TPP⁺-selective electrode by adding 1 μM increments of a 0.1 mM TPP solution to a final concentration of 3 μM. Observe that there is a logarithmic decline in the TPP⁺ voltage signal with each addition.
      NOTE: Calibrate the TPP⁺-selective electrode at the start of each experiment.
5. Data acquisition
   1. Allow the O₂ and TPP⁺ traces to stabilize, and add 100 μg of freshly prepared cardiomyocyte mitochondria to the reaction chamber to a final concentration of 0.1 mg mL^-1 via the reagent addition port in the plunger assembly. Observe the decrease in the O₂ levels in the chamber as mitochondria become energized and consume O₂ and the abrupt increase in the TPP⁺ voltage signal as the mitochondria generate a membrane potential and take up TPP⁺ from the solution.
   2. Add 2.5 μg mL^-1 oligomycin to induce state 4 respiration.
      NOTE: O₂ consumption rates will now represent the rate of proton leak respiration. Oligomycin can be added to the reaction chamber prior to TPP⁺ as an alternative.
6. Assessing the open probability of the mPTP
   1. Observe that ΔΨ declines over time during leak respiration. Once the desired ΔΨ is reached, add 1 μM CsA (mPTP inhibitor) to the reaction chamber to assess open probability of the mPTP at that specific ΔΨ.
   2. Measure the effect of CsA on O₂ consumption and ΔΨ before and after CsA addition
   3. Determine the voltage dependence of mPTP opening by varying the ΔΨ at which CsA is added in successive experiments.

Typical O₂ consumption and ΔΨ curves generated in these experiments are shown (Figure 1A,B). The logarithmic decline in the voltage signal with TPP⁺ calibration is shown at the start of each experiment. The absence of this logarithmic pattern may suggest a problem with the TPP⁺ selective electrode. Mitochondria typically generate ΔΨ immediately upon addition to respiratory buffer. ΔΨ can be interpreted from changes in TPP⁺ voltage based on the Nernst equation (logarithmic ratio of [TPP⁺ inside mitochondrial matrix] to [external TPP⁺])³⁰. Note that physiological ΔΨ approximates the 2 μM TPP⁺ level of the standard curve. Voltage thresholds were thus defined relative to the 2 μM TPP⁺ standard (denoted as Δ mV TPP⁺ voltage signal).

For example, "high ΔΨ" was defined as ~Δ0 mV (at the 2 μM TPP⁺ level), an "intermediate ΔΨ" was set at ~Δ5 mV (below 2 μM TPP⁺), and "low ΔΨ" was set at ~Δ10 mV (below 2 μM TPP⁺) (Figure 1A,B). In pilot experiments, mitochondria at Δ0 mV exhibited 100% mPTP closed probability and those at Δ10 mV exhibited 100% open probability; representative graphs are shown in Figure 1A,B. The Δ5 mV threshold was thus arbitrarily chosen as an intermediate ΔΨ. Note that in these experiments, upon addition of mitochondria to the reaction chamber, mitochondria are in state 2 respiration where they are exposed to excess substrate. Mitochondria are not stimulated to enter state 3 with ADP; as a result, once oligomycin is added, they transition directly to state 4 oligomycin (leak respiration). As a result, an appreciable difference in oxygen consumption is typically not observed after oligomycin addition, suggesting that the ATP synthase contributes minimally to respiration during state 2 (Figure 1A,B). To evaluate the open or closed status of the mPTP, the O₂ consumption rate and membrane potential just before and just after CsA addition are compared (Figure 1C). A decrease in the O₂ consumption rate and an increase and stabilization of ΔΨ in response to CsA indicate closure of an open mPTP (blockade of pore-mediated proton leak) (Figure 1C, black curves).

When the mPTP is closed, there is no decrease in O₂ consumption rate, and ΔΨ does not increase and stabilize but continues to fall (Figure 1C, red curves). The experimental results demonstrate similar closed and open mPTP probabilities at high and low ΔΨs, respectively, in FVB control and Fmr1 KO cardiac mitochondria (Figure 1D). These findings are consistent with known voltage-sensitive properties of the mPTP, where the pore tends to be closed at high ΔΨs and open at low ΔΨs¹³. Thus, using this method, we reliably demonstrated the usual voltage dependence of the mPTP in both strains. Interestingly, at the intermediate ΔΨ (Δ5 mV threshold), Fmr1 KO cardiac mitochondria demonstrated increased closed mPTP probability compared to FVB controls (Figure 1D). Thus, Fmr1 KO cardiac mitochondria demonstrated a shift in gating such that the pore required a lower ΔΨ for opening to be triggered. This is consistent with the previous finding that Fmr1 KOs have elevated levels of CoQ and a relatively closed mPTP²⁶. Thus, the method identified an optimal ΔΨ threshold to resolve differences in mPTP opening. Importantly, defining this ΔΨ threshold will enable further investigation to better understand how CoQ regulates the mPTP and how CoQ deficiency leads to pathological pore opening in FXS.

Figure 1: Detection of low-conductance mPTP open probability. (A, B) Representative curves of simultaneous O₂ consumption (red, numbers are rates in nmol of O₂ mL^-1 min^-1 mg protein^-1) with ΔΨ (black, [TPP⁺]). Arrows indicate the addition of TPP⁺, mitochondria, oligomycin, and CsA. High, intermediate, and low voltage thresholds relative to the 2 μM TPP⁺ calibration level are shown. Addition of CsA at high (A) and low (B) voltage thresholds is shown. (C) Magnified section of O₂ consumption (upper) and ΔΨ (lower) curves in (A) and (B) at the point of CsA addition, illustrating closed mPTP (red) at high ΔΨ (insensitive to CsA) and open mPTP (black) at low ΔΨ (sensitive to CsA). (D) Summary graphs of mPTP open and closed status at high, low, and intermediate ΔΨs are shown for FVB controls (black) and Fmr1 KOs (gray). Five to 11 animals were evaluated at each ΔΨ threshold; p values were calculated using a chi-squared test, *p < 0.05. Abbreviations: ΔΨ = mitochondrial membrane potential; mPTP = mitochondrial permeability transition pore; TPP⁺ = tetraphenylphosphonium; mito = mitochondria; oligo = oligomycin; CsA = cyclosporine A; ΔΨ = ; KO = knockout. Please click here to view a larger version of this figure.

Table 1: Buffers and solutions.
* Filter through 0.2 μm filters
**Adjust pH with 3 M potassium hydroxide as needed.
***Denotes buffer components that are added just prior to mitochondrial isolation.
Abbreviations: MI = mitochondrial isolation buffer; SB = sucrose buffer; RB = Respiration buffer; BSA = fatty acid-free bovine serum albumin; HEPES = 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; EGTA = ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid; KCl = potassium chloride; KH₂PO₄ = potassium dihydrogen phosphate; MgCl₂ = magnesium chloride; AP5A = P1,P5-diadenosine-5' pentaphosphate pentasodium.

This paper describes a method to assess the open probability of the mPTP. Specifically, the voltage threshold for low-conductance mPTP opening was determined by assessing the effect of CsA inhibition on proton leak over a range of ΔΨs. Using this technique, we could identify differences in voltage gating of the mPTP between FXS mice and FVB controls consistent with their differences in tissue-specific CoQ content. Critical to the success of this methodology is that mitochondria are freshly isolated prior to use and are of good quality. Mitochondria that are either damaged during isolation or too old will be unable to generate appropriate ΔΨ and will typically be uncoupled. Note that while matrix pH typically contributes to the proton motive force, nigericin is used to collapse ΔpH across the IMM in these experiments, as previously described in standard protocols assessing proton leak²⁹. Therefore, in this case, the proton motive force is equivalent to ΔΨ, which simplifies the experimental approach.

Another important consideration is that some of the inhibitors used in the respiration mixture (such as rotenone, oligomycin, and CsA) are difficult to rinse away between experiments because they adhere to plastic surfaces of the reaction chamber, including the ion-selective electrodes. The chamber must be washed routinely with ethanol and a suspension of crudely prepared freeze-thawed mitochondria to "mop up" these inhibitors from the chamber and reduce the likelihood of inhibitor carry-over that can affect the reproducibility of experiments.

As mitochondria must be freshly prepared to maintain their integrity, this method is not amenable to high-throughput analyses. In addition, relatively large amounts of mitochondria are needed for each experiment. Typically, for P10 animals, enough mitochondria can be isolated from one organ to conduct a single experiment. Therefore, testing different conditions on the same batch of mitochondria is not usually possible. If experiments are performed with older animals, this limitation is less of an issue, given the relatively larger size of tissues and organs.

A number of well-described methods exist to study the open probability of the mPTP in isolated intact mitochondria. Techniques, such as mitochondrial matrix swelling and calcium loading or retention capacity assays, necessarily trigger high-conductance PT and thus do not permit the study of physiologic low-conductance opening that is of interest in these experiments²⁸. The patch-clamp technique provides a powerful and direct method to study the voltage dependence of high- or low-conductance states of the mPTP^31,32. However, it is a laborious technique, where only a single mitochondrion and a single experimental condition can be studied at a time, which makes comparative studies more difficult^32,33.

In addition, the outer mitochondrial membrane is typically lysed, and the inner mitochondrial membrane is typically ruptured in the patch-clamp methods used for isolated mitochondria, which may result in the loss of mPTP regulatory factors that are not intimately associated with the IMM³¹. Low-conductance opening of the mPTP has also been studied using a variety of fluorescence techniques typically employing a tetramethylrhodamine ester (to assess changes in ΔΨ) either alone or in combination with calcein, whose intramitochondrial fluorescence is susceptible to cobalt quenching when the mPTP is open^8,34. However, in fluorescence-based approaches, as changes in ΔΨ are expressed as relative changes in fluorescence, the approach typically does not allow for precise measurement of ΔΨ^8,34.

This method provides a new alternative approach to assess the low-conductance opening of the mPTP with respect to ΔΨ. Using this approach, investigators can study the regulation of physiologic and pathologic mPTP voltage gating. This technique also holds promise to help understand the developmental role of the mPTP in organ maturation as it can be applied to various tissues and different stages of development. This approach can be readily applied to study voltage gating of the mPTP in FXS mouse forebrain tissue, which previously exhibited decreased CoQ levels²⁵. Further, we propose that this technique can also be used to study how other agents, endogenous or exogenous, impact voltage gating of the mPTP. Thus, it will prove to be a useful tool to expand the understanding of the contribution of the mPTP to health and disease.