Simultaneous Quantification of Mitochondrial ATP and ROS Production Using ATP Energy Clamp Methodology

Abstract

Several methods are available to measure ATP production by isolated mitochondria or permeabilized cells but have several limitations, depending upon the particular assay employed. These limitations may include poor sensitivity or specificity, complexity of the method, poor throughput, changes in mitochondrial inner membrane potential as ATP is consumed, and/or inability to simultaneously assess other mitochondrial functional parameters. Here we describe a novel nuclear magnetic resonance (NMR)-based assay that can be carried out with high efficiency in a manner that alleviates the above problems.

1. Introduction

Here we describe a novel, highly sensitive and specific nuclear magnetic resonance (NMR)-based ATP assay that can be carried out with reasonably high throughput using small amounts of mitochondrial isolates or permeabilized cells (see Note 1). There are several advantages of this method. First, the assay allows for simultaneous fluorescent measurements. For example, it is possible to quantify the production of reactive oxygen species (ROS) simultaneously with ATP production measurement, as we discuss below. Another major advantage of the assay is that it avoids the problem of changing mitochondrial membrane potential (ΔΨ) while ADP is concerted to ATP, as occurs in conventional assays. In contrast to conventional assays, ΔΨ in our assay is clamped at fixed levels determined by the amount of ADP added.

1.1. 2DOG ATP Energy Clamp

To assess mitochondrial functional parameters at fixed ΔΨ, we use excess 2-deoxyglucose (2DOG) (see Note 2) and hexokinase (HK) to generate an “ATP energy clamp” (Fig. 1). The conversion of 2DOG to 2DOG phosphate (2DOGP) occurs rapidly and irreversibly, thereby effectively clamping ADP concentrations and ΔΨ dependent on the amount of exogenous ADP added (see Note 3). This enables titration of membrane potential at different fixed values with mitochondria in respiratory states ranging from state 4 (no ADP, maximal potential) to state 3 (high levels of ADP, reduced potential) (Fig. 2).

Fig. 1.

The 2-deoxyglucose (2DOG) energy clamp. Saturating amounts of 2-deoxyglucose (2DOG) and hexokinase (HK) recycle ATP back to ADP by rapidly and irreversibly converting 2DOG into 2-deoxyglucose phosphate (2DOGP). The resulting ADP availability is clamped at levels determined by the amount of ADP added. IMM inner mitochondrial membrane

Fig. 2.

Computer tracings of inner membrane potential vs. time obtained by incubating normal mouse liver mitochondria, 0.5 mg/mL (panel a) or heart mitochondria, 0.25 mg/mL (panel b), fueled by the combined substrates of 5 mM succinate + 5 mM glutamate + 1 mM malate. ADP was added in incremental amounts to generate the final total recycling nucleotide phosphate concentrations (indicated by arrows). Potential was determined using a tetraphenylphosphonium (TPP) electrode [2]. After each addition, a plateau potential was reached, consistent with recycling at a steady ADP concentration and generation of a stepwise transition from state 4 to state 3 respiration. Note that the potential shown on the y-axis represents electrode potential (not mitochondrial potential). The actual ΔΨ follows a similar pattern after calculation using the Nernst equation based on the distribution of tetraphenylphosphonium (TPP) external and internal to mitochondria

1.2. Use of the 2DOG ATP Energy Clamp to Quantify ATP Production in Isolated Mitochondria and Simultaneous Assessment of H₂O₂ Production

Mitochondria are added to individual wells of 96-well plates in a total volume of typically 60 μL and incubated at 37 °C in respiratory buffer in the presence of HK, 2DOG, and the desired concentration of ADP. [6-¹³C]2DOG is used in two-dimensional (2D) ¹³C/¹H-HSQC NMR-based assays, while unlabeled 2DOG is used in one-dimensional (1D) ¹H NMR-based assays (see below). After incubation for the desired time, the contents of the microplate wells are removed to tubes on ice containing oligomycin to inhibit ATP synthase. The tubes are then centrifuged, and the supernatants are held at −20°C until NMR analysis. To prepare the NMR sample for measurement of ATP production by mitochondria, the assay supernatants are added to 5 mm (OD) standard NMR tubes along with appropriate NMR buffer (see below).

ATP production rates are calculated based on the percent conversion of 2DOG to 2DOGP as determined by NMR, the initial 2DOG concentration, incubation volume, and incubation time. The assay clearly has several advantages over existing methods (see Note 4).

In order to simultaneously assess H₂O₂ production, mitochondrial incubations can be carried out in the presence of DHPA. Moreover, it is possible to use other fluorescent probes to assess mitochondrial parameters such as membrane potential and calcium uptake as long as the probes do not interfere with the NMR detection.

1.3. Utilization in Recent Studies

We have used this assay in several recent studies designed to assess muscle, heart, liver, brain, and brown adipose tissue mitochondrial function under normal and perturbed metabolic conditions. By clamping ΔΨ, we found that ATP production by muscle mitochondria isolated from insulin deficient rodents was limited not only by reduced respiration but also by an inability to harness membrane potential for ATP production [1]. We also showed that ROS production per unit of ATP generated was greater in these diabetic muscle mitochondria [1]. In further work, we observed that muscle mitochondria isolated from high-fat fed obesity-prone mice was associated with a greater oxidative cost of ATP production [2]. In a further study of a rodent model of type 2 diabetes, we observed that ATP production by succinate-energized muscle mitochondria was impaired due to decreased utilization of membrane potential and that more ADP was required for peak respiration [3].

In more recent work, we found that membrane potential primarily determines the relationship of complex II energized-respiration to ADP concentrations by isolated muscle mitochondria [4]. We further observed that oxygen flux in succinate-energized mitochondria peaked at lower levels of added ADP (higher potential) but actually decreased at higher ADP (lower potential). Further work revealed that this biphasic pattern of oxygen flux versus ADP concentrations was due to membrane potential dependent oxaloacetate inhibition of succinate dehydrogenase [5, 6]. Of note, the abovementioned biphasic respiratory pattern was dependent on tissue type from which mitochondria were isolated, being barely evident in liver mitochondria but prominent for brain, muscle, and heart mitochondria [4].

In another study, we examined the effect of calcium on enhancing respiration by mitochondria energized by complex I substrates in the presence of varying concentrations of ADP. We found that the stimulatory effect of calcium on mitochondrial function was substrate dependent and most prominent over intermediate respiratory status [7]. On the other hand, ROS production was positively and continuously associated with increasing calcium concentrations [7]. In that work, we further observed that the inhibition of respiration by calcium at higher concentrations can occur independent of opening of the mitochondrial permeability transition pore.

Finally, we have used this 2DOG ATP energy clamp methodology to study liver and heart mitochondrial function in mice fed with high dietary fat [2] and the effect of uncoupling protein I on respiration in brown adipose tissue [5].

2. Materials

2.1. Isolation of Mitochondria

1. Isolation medium: 0.25 M sucrose, 5 mM HEPES (pH 7.2), 0.1 mM EDTA, 0.1% BSA (fatty acid-free).

2. Purification medium: 30% v/v Percoll^®. Dilute three parts Percoll^® with seven parts of isolation medium. 2.4 mL of Percoll^® + 5.6 mL of isolation medium = 8 mL, sufficient for two centrifuge tubes. Keep on ice.

3. Beckman XL-80 ultracentrifuge or similar instrument, precooled to 4 °C, SW60 swinging-bucket rotor with caps and greased O-ring seals, polyallomer centrifuge tubes ~4.2 mL max capacity per tube.

2.2. Assay Incubation

1. Respiration medium: 0.3% fatty acid-free BSA, 120 mM KCl, 1 mM EGTA, 5 mM KH₂PO₄, 2 mM MgCl₂, 10 mM HEPES, pH 7.2.

2. Microplate with 96-wells (e.g., a Costar #3792 black round-bottom plate).

3. 10 U/mL hexokinase (HK), 5 mM 2-deoxyglucose (2DOG) or [6-¹³C]2DOG, 10-acetyl-3,7-dihydroxyphenoxazine (DHPA or Amplex Red, Invitrogen), horseradish peroxidase (HRP).

2.3. Sample Preparation for NMR Spectroscopy

1. Sample dilution buffer: 120 mM KCl, 5 mM KH₂PO₄, 2 mM MgCl₂, pH 7.2.

2. Deuterium oxide (D₂O).

3. Standard 7-in. (length) × 5 mm (outer diameter) NMR tubes.

2.4. NMR Spectroscopy

1. NMR spectrometer equipped with a dual or triple resonance probe and capable of acquiring ¹H and ¹H/¹³C HSQC NMR spectra.

2. NMR spectrometer equipped with an auto sample changer, thus capable of continuous data acquisition of multiple samples, for example, 60 samples.

3. Methods

3.1. Isolation of Mitochondria from Tissue

1. Carry out all procedures on ice or at 4 °C.

2. Harvest the tissue and rinse in isolation medium.

3. Homogenize up to 1 g of tissue in 10–15 mL of isolation medium using a Potter-Elvehjem-type tissue grinder in an ice bucket. Fibrous tissues should be minced with scissors to aid tissue disruption. The Teflon pestle is mounted on a drill set to approximately 300 rpm. Four to six passes are typically required. Optionally, a subsequent pass of the homogenate through a ground-glass-style homogenizer can increase the mitochondrial yield from fibrous tissues.

4. Centrifuge homogenate at 500 × g for 10 min (low speed spin) (see Note 5).

5. Transfer the supernatant to Sorvall-type tube (Oak Ridge screw cap). Discard the pellet and centrifuge at 10,000 × g for 10 min (high-speed spin). Discard supernatant.

6. Wash the mitochondrial pellet with isolation medium without BSA.

7. Resuspend the final pellet at ~50% v/v in isolation medium without BSA.

3.2. Further Purification of Isolated Mitochondria (See Note 6)

1. Add 3.8 mL of 30% Percoll^® solution to each SW60 polyallomer tube on ice.

2. Resuspend mitochondrial crude prep pellets in 0.1 mL of isolation medium containing BSA.

3. Lay the mitochondria on top of the Percoll^® solution and insert the tubes into the buckets.

4. Use a balance to precisely equalize the mass of the buckets + tubes + lids, adding isolation medium containing BSA to adjust the mass. Ensure that the contents of the tubes are within 3 mm of the top of the tubes.

5. Securely attach the bucket caps. A liquid sample within a sealed bucket will be protected from the vacuum applied to the centrifuge chamber.

6. Hang the buckets on the precooled SW61 rotor at 4 °C and spin for 30 min at 30,000 rpm (~90,000 × g).

7. The pure mitochondria band appears near the bottom of the tube, just above a clear and dense mass of Percoll^®. Remove all contaminating fractions above the mitochondria band with a pipette.

8. Transfer the mitochondria band to a 1.5 mL centrifuge tube.

9. Add 1 mL of isolation medium without BSA. Spin in a microfuge at 8000 × g for 5 min at 4 °C. Remove the supernatant. If subjecting the mitochondria to endogenous Ca²⁺ depletion, then stop here and proceed with a calcium depletion protocol (see Note 7). Otherwise, resuspend the pellet in 1 mL of isolation medium without BSA and spin in a microfuge at 8000 × g for 5 min at 4 °C.

10. Resuspend the final washed pellet in BSA-free isolation medium and keep on ice.

3.3. Assay Incubation

1. Warm a 96-well microplate to 37 °C for 10 min. If carrying out simultaneous fluorescent detection (for example, for reactive oxygen species), warm the plate in the plate reader and set the plate reader gain as desired.

2. Preload all reagents (before adding mitochondria) to wells with 1.2× respiration medium (upon subsequent addition of mitochondria the medium will be 1×) in a total volume of 50 μL. For rat hind limb muscle mitochondria [1], we used the following final or 1× assay concentrations: 5 mM succinate + 5 mM glutamate + 1 mM malate (or other mitochondrial fuel selection and concentrations as desired), 10 U/mL hexokinase (HK), 5 mM 2-deoxyglucose (2DOG) (using [6-¹³C]2DOG for 2D NMR detection or unlabeled 2DOG for 1D NMR detection, see below), and ADP at desired concentration (up to 100 μM). For simultaneous assay of reactive oxygen species, add 20 μM 10-acetyl-3,7-dihydroxyphenoxazine (DHPA or Amplex Red, Invitrogen) plus 5 U/mL of horseradish peroxidase (HRP).

3. Wells containing no added substrate should be present on the plate to serve as background control. Reserve some wells for inclusion of an H₂O₂ standard curve, typically containing a range of H₂O₂ concentrations up to a maximum of 5–10 μM. These standard curve wells do not have added mitochondria.

4. Two to three wells should be included in the plate to serve as positive controls and/or standards for NMR quantification of the amount of 2DOGP. These wells have the same final total volume of 60 μL as the others, but only contain 5 mM 2DOG or [6-¹³C]2DOG (depending on whether 1D or 2D NMR methods are used for measurement, respectively, see below), 10 U/mL hexokinase, and 10 mM ATP in 1× respiration medium. The use of twofold molar excess of ATP with respect to 2DOG is to ensure full conversion of 2DOG to 2DOGP.

5. To start the assay, add 10 μL of 6× concentrated mitochondria suspended in 1× respiration medium. The well has a total volume of 60 μL now. The plate reader program is started and fluorescence determined at a frequency of once per minute or greater. Final mitochondrial concentrations are typically 1–0.5 mg/mL.

6. After incubation, typically for 5–30 min, harvest the wells by transferring the well contents to 500 μL tubes containing 1 μL of 120 μM oligomycin. Then immediately centrifuge the tubes at 10,000 × g for 4 min at 4 °C.

7. Transfer the supernatants to new labeled tubes and hold them at −20 °C until NMR sample assembly.

3.4. Fluorescent Assessment of H₂O₂ Production

1. As indicated above, the assay wells contain DHPA (20 μM) and HRP (5 U/mL).

2. Monitor DHPA fluorescence over the duration of the assay at 544 nm excitation and 590 nM emission.

3. Determine the average fluorescence over the duration of the assay for all points in the standard curve. Subtract the average background fluorescence from all values for the standard curve. The background value is defined as fluorescence in the absence of H₂O₂ (the zero H₂O₂ point in the standard curve).

4. For all wells containing mitochondria, determine the slope of fluorescence as a function of time (see Note 8). It is not necessary to subtract a background fluorescence value in order to determine slope for these wells since the background is constant.

5. Use curve fitting software to assess background-subtracted standard curve fluorescence as a function of the molar amount of H₂O₂ present in the well. Then use the fitted curve equation to convert slope values for the wells containing mitochondria to molar values (e.g., pmol per min).

3.5. Processing the Well Contents for NMR-Based ATP Assay

1. Add 0.39 mL of sample dilution buffer, 50 μL of deuterium oxide (D₂O), and 40 μL of assay well supernatant to a standard 7-in. (length) × 5 mm (outer diameter) NMR tube.

2. Deliver the prepared NMR samples (kept at 4 °C) to the NMR facility.

3.6. NMR Spectroscopy for Quantifying ATP Production

1. In our studies, we use a Bruker Avance II 500 MHz NMR spectrometer equipped with a 5 mm TXI triple resonance non-cryoprobe operating at 37°C. The spectrometer is also equipped with an automatic sample changer capable of holding a maximum of 60 samples. The amount of ATP produced by the mitochondria is quantified by measuring the amount of 2DOGP produced from 2DOG in the presence of hexokinase as described above.

2. For precise measurement of the amount of 2DOGP formed, duplicate or triplet control samples are prepared during the assay (see Subheading 3.3). The control samples have the same total volume of 60 μL as the other samples, but only contain 5 mM 2DOG or [6-¹³C]2DOG (depending on whether 1D or 2D NMR methods are used for measurement, respectively. See Note 9), 10 U/mL hexokinase, and 10 mM ATP in 1× respiration medium. These control samples are then subjected to the same protocol for NMR sample preparation (see Subheading 3.5). Due to the presence of excess ATP, these control samples have the 2DOG being fully converted into 2DOGP. Therefore, these control samples serve as standards for quantification of 2DOGP formation within the mitochondrial samples and are also used to check for assay reproducibility since duplicate or triplet control samples are used.

3. Load the control samples and mitochondrial samples onto the sample changer. We can run 60 samples continuously without interruption at this spectrometer.

4. Use a control sample or mitochondrial sample to lock, tune, and shim. Save the optimized shimming parameters which serve as the starting shimming setting for the subsequent automatic robot run. Also, use this sample to calibrate ¹H and ¹³C channel pulse widths (¹³C pulse calibration is needed only when [6-¹³C]2DOG is used).

5. For Bruker Topspin software, start the automation program ICONNMR Set up the automatic robot run by choosing 1D ¹H NMR experiment if unlabeled 2DOG is used or choosing 2D ¹³C/¹H HSQC NMR experiment if[6-¹³C]2DOG is used in the samples. Enter appropriate pulse program parameters such as pulse power level, pulse width, relaxation delay, number of scans, water presaturation parameters, number of t1 increments. Also, set to tune ¹H channel and shim on every sample. It takes about 25 min to collect either a 1D ¹H spectrum or a 2D ¹³C/¹H HSQC spectrum for each sample. Start the robot run.

6. Transfer the collected NMR data to another Linux computer where the acquired data are processed by using NMRPipe package [8] and analyzed using NMRView [9].

7. Using the assigned ¹H and ¹³C NMR resonances of 2DOG and 2DOGP [1], measure the peak intensities in NMRView of 2DOGP in the processed control and mitochondrial samples.

8. Quantify the amount of 2DOGP present in the mitochondrial samples by comparing the peak intensity of 2DOGP of the mitochondrial sample with that of the control/standard samples. Since a fixed amount of 2DOG (5 mM) is used in the assay incubation (see Subheading 3.3), the amount of 2DOGP formed corresponds to the amount of reduction of 2DOG and thus can be expressed as percentage of conversion from 2DOG to 2DOGP. The amount of 2DOGP determined for the mitochondrial samples corresponds to the amount of ATP produced by the mitochondria. Representative data derived from the analysis by 1D and 2D NMR methods are shown in Fig. 3.

Fig. 3.

Quantification of ATP generated by mitochondria via measurement of 2DOGP converted from 2DOG by 1D ¹H NMR method (a) and 2D ¹³C/¹H HSQC NMR method (b). In panel a, representative data are derived from the incubation of isolated gastrocnemius muscle mitochondria from a control rat. Mitochondria (0.1 mg/mL) are incubated for 20 min in respiration medium containing 5 mM 2DOG, excess hexokinase, variable concentrations of ADP, and fueled with 5 mM succinate, 5 mM glutamate, and 1 mM malate. It clearly shows that the percent 2DOGP formed by the mitochondrial incubation increases with increasing amount of ADP used in the incubation. No subs mean that substrates of succinate, glutamate, and malate are not added. In panel b, overlay of representative 2D ¹H/¹³C HSQC spectra is shown for the H6/C6 region of 2DOG and 2DOGP of the samples that contain mitochondria from a control (black) and diabetic (red) rat. Mitochondria (0.1 mg/mL) are incubated for 20 min in respiration medium containing 5 mM [6-¹³C]2DOG, excess hexokinase, and 1 mM ADP, and fueled with 5 mM succinate, 5 mM glutamate, and 1 mM malate. The cross peaks are labeled, and 1D slices through the cross peaks of 2DOGP are included and clearly show that the control mitochondria produce more 2DOGp or ATP than the diabetic mitochondria

3.7. Calculation of ATP Production Rates

1. Determine the percent conversion of 2DOG to 2DOGP. From the known initial 2DOG concentration and percent conversion, determine the molar amount of 2DOGP formed which equals to the molar amount of ATP generated. Calculate ATP production rates in the microplate assay wells based on the volume in microplate, the amount used in NMR, and incubation time.

4. Notes

1. The cytoplasm of permeabilized cells is replaced by the respiratory media enabling assessment of mitochondrial function independent of cytoplasmic events. The methods described herein are for isolated mitochondria. Permeabilized cells can also be used by adapting the methodology beginning in Subheading 3.3.

2. Cytoplasmic hexokinase is competitively inhibited by 2-deoxyglucose, but this is not an issue in isolated mitochondria or permeabilized cells wherein the cytoplasm is replaced by the incubation medium.

3. The 2DOG clamp has been used in the past to assess mitochondrial-bound HK activity at constant ADP [10], but not to assess mitochondrial physiology or ATP production as described herein. Mitochondrial HK is not an issue in our assay since the amount added HK is in excess.

4. There are several advantages to our ATP assay. First, ΔΨ is clamped, allowing assessment of ATP as a function of its direct driving force (i.e., ΔΨ). Second, the assay is sensitive enough to measure ATP production in small numbers of mitochondria. We found that the 1D ¹H NMR method is 34-fold more sensitive and the 2D ¹H/¹³C HSQC NMR method is 41-fold more sensitive when compared to 1D ³¹P NMR for ATP detection [1]. The higher sensitivity of the 2D NMR method is due to the fact that the chemical shifts of the two H6 protons of the β-anomeric form of 2DOGP are degenerate, resulting in detection of one single C6/H6 HSQC cross peak with high intensity. Third, both the 1D and 2D NMR spectra are highly specific. Fourth, throughput is quite good since we add mitochondria to multiple wells of a 96-well plate, incubate, spin off the mitochondria, and then save the samples for NMR analysis. Fifth, a powerful aspect is that we can assess mitochondrial ROS simultaneously with ATP quantification by NMR since the ROS probe does not interfere with mitochondrial ATP production or with NMR detection of 2DOGP for ATP quantification [1]. Therefore, the ATP assay described herein has clear advantages over conventional methods. Fluorescent and bioluminescent measurements are sensitive, but lack specificity and may be confounded by background interference or variations in light emission [11]. Quantifying ATP by phosphorous NMR is not sensitive enough and requires long acquisition time unless large amounts of mitochondria are used. High-pressure liquid chromatography is precise, but cumbersome. The ATP:O ratio is often considered representative of ATP production. However, ATP itself is not measured and the ratio can be altered by any condition that affects uncoupling, ATP synthase, or respiration.

5. To increase the mitochondrial yield with fibrous tissues, save the first low-speed pellet for resuspension and regrinding using the ground-glass-type homogenizer. Spin the homogenate at low speed. Combine the supernatants from both low-speed spins prior to the high-speed spin.

6. Crude mitochondria pellets obtained by standard methods can be further purified using a self-generating Percoll^® gradient. We use a published method [12] described for liver mitochondria (which we adapted for heart and skeletal muscle mitochondria) by using a centrifugal force of 95,000 × g to establish the gradient. Others recommend only 30,000 × g. In our initial attempts by using 30,000 × g, we did not get acceptable results as evidenced by loose, fluffy, and diffuse bands or mitochondria that stayed only at the top of the tube. When we increased the centrifugal force to 95,000 × g, we obtained excellent separation of mitochondria from contaminants. In this way, the mitochondrial band (near the bottom of tube) is clearly separated from less-dense contaminants and broken mitochondria (upper and middle bands, respectively).

7. Calcium depletion can be carried out as follows: Percoll-purified mitochondria (0.8–1.2 mg) are incubated for 6 min at 37 °C in 1 mL of ionic respiratory buffer (105 mM KCl, 10 mM NaCl, 5 mM Na₂HPO₄, 2 mM MgCl₂, 10 mM HEPES pH 7.2, 1 mM EGTA, 0.1 mM malate, 3.2 mM ATP, 0.2% defatted BSA). The mitochondria are then pelleted at 10,000 × g and the pellet washed twice with ice-cold isolation medium (0.25 M sucrose, 0.1 mM EDTA, 3 mM HEPES, pH 7.25).

8. The observed increase in Amplex Red fluorescence over time is usually quite linear. The slope for each well can be converted to molar H₂O₂ per unit time per mg of mitochondrial protein with the aid of the standard curve.

9. The 2D NMR method is preferred since it is more sensitive and has almost no background interference.