Assessment of Mitochondrial Oxygen Consumption Using a Plate Reader-based Fluorescent Assay

Victoria Del Gaizo Moore; Gregory Haenel; Aiofe Judge

doi:10.3791/65760

JoVE Journal > Biochemistry

Bioquímica

Assessment of Mitochondrial Oxygen Consumption Using a Plate Reader-based Fluorescent Assay

Published: April 12, 2024

doi:

10.3791/65760

Victoria Del Gaizo Moore¹, Gregory Haenel², Aiofe Judge¹

¹Chemistry Department,Elon University, ²Biology Department,Elon University

Summary

Assessment of oxygen consumption provides integral information about mitochondrial function. Using a phosphorescent probe with a fluorescent plate reader, accurate and reproducible data can be obtained easily without specialized equipment. This assay enables any lab to measure the oxygen consumption of isolated mitochondria and calculate respiratory control ratios.

Abstract

Mitochondria serve many important functions, including cellular respiration, ATP production, controlling apoptosis, and acting as a central hub of metabolic pathways. Therefore, experimentally assessing mitochondrial functionality can provide insight into variations among different populations or disease states. Additionally, it is valuable to assess whether isolated mitochondria are healthy enough to proceed with experiments. One characteristic often used to compare mitochondrial function in different samples is the rate of oxygen consumption. Oxygen consumption and subsequent calculation of the respiratory control ratio in either intact cells or mitochondria isolated from tissue can serve all three purposes. Using mitochondria isolated from the livers of brush lizards in conjunction with a phosphorescent probe that is sensitive to the fluctuations in oxygen concentration of a solution, we measured oxygen consumption using a fluorescent plate reader. This method is not only quick and efficient but also can be conducted with a small amount of mitochondria and without the need for specialized equipment. The step-by-step protocol described here increases the accessibility of mitochondrial functional assessment to researchers.

Introduction

Mitochondria are organelles, approximately the size of bacteria, found in eukaryotic cells. They are unique organelles because they contain DNA and have two membranes, an outer and inner one. Mitochondria's outer and inner membranes are separated by an intermembrane space, and the inner membrane folds into structures called cristae around the innermost compartment, called the matrix. These cristae increase the inner membrane's surface area so that multiple processes that use the cristae can occur simultaneously. While mitochondria are involved in many cellular functions such as controlling apoptosis and housing multiple metabolic pathways, their vital role in the production of ATP is essential for cell survival. In fact, 90% of a cell's energy is derived from mitochondria¹. ATP production involves the generation of an electrochemical difference across the outer and inner membranes, termed the mitochondrial membrane potential (Δψ), which arises as H⁺ ions are pumped from the matrix into the intermembrane space. ATP production is ultimately harnessed during the oxidation of reducing equivalents via electron movement through the mitochondrial respiratory chain (ETC). The final electron acceptor is molecular oxygen (O₂). As oxygen is consumed, the H⁺ concentration differential builds up to its maximum, at which point H⁺ ions move down their concentration gradient from the intermembrane space to the matrix by passing through the ATP synthase complex. The movement of H⁺ ions causes a conformational change in ATP synthase, and ADP is brought into proximity with inorganic phosphate to react and generate ATP. Finally, ATP is translocated out of the mitochondrial matrix into the cytosol and can either be stored or used to facilitate reactions due to the large amount of free energy released during the hydrolysis of its phosphates. This whole process is termed oxidative phosphorylation, and since oxygen is consumed, mitochondria are said to respire².

The buildup and strength of Δψ, the amount of O₂ reduced (termed oxygen consumption), as well as the generation of ATP can all be used as indications of cell health. Mitochondrial functional studies, such as measurement of Δψ, total ATP content and production, and oxygen consumption can be quantified either by traditional biochemical methods or fluorescence and luminescence in plate-based assays. For example, mitochondrial membrane potential can be compared among different samples using fluorescent dyes such as tetramethylrhodamine ethyl ester, which binds specifically to mitochondria. ATP generation can be monitored by adding a luminescent protein to a reaction whose changes correlate to ATP concentration. Quantification of oxygen consumption rates, or absolute rates of respiration, during OXPHOS, can help elucidate the causes of disparities in mitochondrial function and energy metabolism. Oxygen consumption assessment can be used to calculate respiratory control ratios (RCRs). RCR values describe the ability of mitochondria to make ATP in response to the influx of ADP, which is the main function of mitochondria. RCR values signify the overall condition of isolated mitochondria and allow for the comparison of responses to different experimental treatments. Differences in RCR values may represent mitochondrial dysfunction or indicate a biological difference between different mitochondria isolated from two or more sources. Another important measure of function in isolated mitochondria is mitochondrial efficiency defined as moles of ATP synthesized per moles of O₂, or the P/O ratio³.

Given the amount of information that can be gathered from measuring mitochondrial parameters and various instances in which this information can be utilized, the ability to efficiently gather functional data can be useful in many different research areas. Mitochondrial oxygen consumption measurements have been performed for decades with very specific instrumentation—using a Clark electrode, which can be limited by the sample size necessary to carry out measurements, and more recently, sophisticated instruments that can measure mitochondrial respiration and multiple other parameters but can be cost-prohibitive. This protocol is an adapted alternative approach using an oxygen-sensitive phosphorescent probe (MitoXpress)⁴^,⁵. The probe signal is detected with a plate reader in time-resolved fluorescence mode for continuous measurements over time. Phosphorescence has a larger energy difference between the absorbed and emitted photon compared to fluorescence and therefore, is better suited for continuously monitoring changes in signal. This enables almost any lab to perform these measurements, not just those that focus on mitochondrial metabolism or who can afford highly specialized equipment. The model system we utilize is isolated mitochondria from three tree lizards, two parental species and one introgressed (containing nuclear DNA from one parental species and mitochondria from the other—hybrids). These lizards were chosen because we hypothesized there are metabolic and energy consequences for hybrids having different nuclear and mitochondrial DNA sources. We utilized a commercially available assay kit with a multi-mode plate reader that can increase access to this type of assay to more researchers and research fields.

Protocol

Lizards were euthanized by CO₂ asphyxiation followed by immediate decapitation in accordance with policies outlined by The Office of Animal Laboratory Welfare and Elon's Institutional Animal Care and Use Committee guidelines.

1. Isolation of mitochondria⁶

NOTE: Keep all solutions cold (Table 1) and samples on ice throughout these steps.

Remove the liver, weigh it, and then rinse it with ~3 mL of ice-cold phosphate-buffered saline (1x, -/-).
Mince the liver in 1 mL of L-MIB with a fresh razor blade.
Bring the total volume of tissue plus L-MIB to 2 mL and mechanically disrupt the liver cells for four passes with a Dounce homogenizer.
Spin the homogenate at 300 × g (37 °C, 10 min).
Transfer the supernatant into a fresh tube and place it on ice.
Resuspend the pellet in 2 mL of L-MIB, rehomogenize, and spin again at 300 × g (37 °C, 10 minutes).
Combine the supernatants from both spins and centrifuge at 10,000 × g (37 °C, 10 min).
Resuspend the pellet in ~0.350 mL of L-MIB.
Determine the amount of total protein (mg of mitochondria/mL of L-MIB) to use as an approximation for mitochondrial content⁴^,⁵^,⁶.

2. Oxygen consumption

Prewarm all the solutions used (Table 1) in the following steps to 30 °C in a water bath.
Dilute the mitochondria in LEB to 6 mg/mL based on the results of the protein concentration assay in step 1.9.
Set up the assay in a sterile 96-well black-wall, clear bottom plate (Table 2).
1. Add 50 µL of the sample (L-EB buffer or mitochondria sample) to the appropriate wells.
2. Add 50 µL of treatment (L-EB, Glutamate/Malate, or Glutamate/Malate w/ADP).
3. Dilute the probe stock 1:10 in L-EB and then add 100 µL of the fluorescent probe to every well.
4. Gently add 50 µL of heavy mineral oil to every well to exclude ambient oxygen.
Read fluorescent measurements from the bottom of the plate at 380/650 nm excitation/emission every 1.5 min for 45 min. Use kinetic mode with a time delay of 30,000 µs and a measurement window of 100 µs.

3. Data analysis

Export the raw data file from the plate reader computer as a .xls file.
Open the file, then copy and paste the raw data into a new tab, and label the rows and columns appropriately.
Plot the L-EB buffer-only to the control sample and L-EB + G/M as relative fluorescent units (RFU) versus time.
NOTE: These lines should be relatively flat as there are no mitochondria present, and therefore, no change in oxygen concentration should be detected.
Plot the L-EB + G/M sample values as RFU versus time.
Determine where the readings become more consistent and flatten out for the controls.
NOTE: This is where data analysis for experimental samples should commence (marked by arrows in Figure 1A). Reactions can take 10-15 min to flatten out and reach their maximum.
Make a new plot only containing the data after the time point established in step 3.5 (called "trimmed" data) and add the best-fit line to visualize the raw data to be analyzed.
Calculate the average RFU value of the trimmed control buffer-only data (highlighted in yellow). This value is used as the minimum RFU value as the baseline in the oxygen calculation below.
Calculation of oxygen consumption rate (µM O₂/min)
1. Copy and paste the linear region of the raw data from step 3.6 (i.e., the trimmed data) to determine the concentration of oxygen in the mitochondrial samples in G/M + ADP treatment at each time point⁷^,⁸^,⁹^,¹⁰.
2. Use equation (1)⁵ to determine the oxygen concentration at each time point. [O₂]_a is the oxygen concentration in air-saturated buffer (235 mM at 30 °C). I(t), Ia, and Io are the fluorescent signal of the sample + probe at time t (e.g., sample + G/M + ADP), the average signal of the probe in air-saturated buffer calculated in step 3.7, and the maximal signal in deoxygenated buffer (set at the maximum signal achievable), respectively.
  (1)
3. Plot the [O₂]_t values calculated in step 3.8.2 versus time and add a linear trendline along with the equation of the best-fit line.
  NOTE: Use the slope of each line as the oxygen consumption rate. Duplicates or triplicates for any sample can be aggregated and used for the comparison of the rate of consumption.
Calculate the RCR by dividing mitochondrial respiration with and without ADP, which represent states 3 and 2, respectively.

Representative Results

Oxygen consumption rate and mitochondrial RCR were determined from the mitochondria of three different lizards using an assay kit with a phosphorescent oxygen-sensing probe and a standard fluorescence plate reader. Previous research established that the probe in this kit directly correlates to oxygen consumption, where phosphorescence is quenched by molecular oxygen and the fluorescent signal increases as oxygen levels decrease due to mitochondrial respiration⁷^,¹¹. RFU values are recorded over time for both control and experimental samples and then analyzed. The RFU readings for control wells (buffer only, buffer treated with glutamate/malate, and buffer treated with glutamate/malate + ADP) were plotted against time to establish that any changes in fluorescence in experimental sample wells are due to the presence of mitochondria. Furthermore, the glutamate/malate + ADP control allows for the baseline amount of oxygen during data collection. Experimental samples containing isolated mitochondria were also plotted on the same graph as the controls.

Figure 1A demonstrates that RFUs increase over time from a sample of isolated mitochondria, as opposed to the relatively flat RFU values of the controls. Since the reactions take time to equilibrate, the first 10 or so readings, until the control sample readings are not as noisy and level out, are removed from the data set (Supplemental Table S1 and Supplemental Figure S1). Next, the remaining values from the mitochondria sample treated with glutamate/malate + ADP are transformed into the concentration of oxygen at each timepoint using equation (1) (see protocol step 3.2) and then plotted versus time. Figure 1B shows the outcome of the experimental sample seen in Figure 1A (crosses) where the R² value is very close to 1 and the slope of the line can be calculated. The slope of the line is negative because as oxygen is consumed, the total oxygen concentration in that sample decreases. This same process is repeated for the same mitochondrial samples treated with glutamate/malate without ADP to obtain oxygen consumption rates (Supplemental Table S2 and Supplemental Figure S2). Then, mitochondrial respiration with and without ADP, which represent states 3 and 2, respectively, can be used to calculate the RCR for each sample type (Supplemental Figure S2). When compared side-by-side and, the average RCR among the three different types of tree lizards indicates that the RCR of the hybrid with introgressed mitochondria is significantly lower than that of either parental type (Figure 1C). Hybrid lizard mitochondria had oxygen consumption rates intermediate to the two parental types (Figure 2).

Figure 1: Example oxygen consumption assay results. (A) Raw data from buffer-only control reactions and reactions including isolated mitochondria that were treated with glutamate and malate (G/M), with or without the addition of ADP. Fluorescence was monitored over time. As oxygen is consumed, fluorescence increases. Buffer-only reactions (G/M, indicated by circles, and G/M + ADP, indicated by triangles) and mitochondria with G/M treatment (indicated by squares) showed no change. Mitochondria treated with G/M + ADP (indicated by plus signs) consumed oxygen. (B) Changes in oxygen concentration over time calculated from the raw data for mitochondria treated with G/M + ADP presented in A. (C) RCR values for mitochondria from three different types of lizards measured at room temperature. This figure is reproduced from Haenel and Del Gaizo Moore¹². Statistical analysis was performed using ANOVAs in R. Abbreviations: G = glutamate; M = malate; RCR = respiratory control ratio. Please click here to view a larger version of this figure.

Figure 2: Oxygen consumption rates of mitochondria from three types of lizards. Boxplots show median oxygen consumption rates as a bar and means as a dot. The parental types, Urosaurus graciosus and Urosaurus ornatus, were significantly different from each other. The introgressed form (Hybrid) was intermediate to the parental types and not significantly different from either. This figure is reproduced from Haenel and Del Gaizo Moore¹². Abbreviations: Ug= Urosaurus graciosus; Uo = Urosaurus ornatus. Please click here to view a larger version of this figure.

Table 1: Composition of solutions used in the protocol. Please click here to download this Table.

Table 2: Plate map for oxygen consumption assay. For each sample, set aside at least 6 wells (3 control wells: L-EB only, L-EB with G/M treatment, L-EB with G/M + ADP treatment; 3 experimental well/sample: mitochondria with L-EB treatment, L-EB + G/M, and L-EB G/M +ADP treatment). When possible, run each experimental sample in duplicate or triplicate. Abbreviations: G = glutamate; M = malate; L-EB = lizard-specific experimental buffer. Please click here to download this Table.

Supplemental Figure S1: Data plots. (A) Plot of all the raw data for comparison of samples and controls. (B) Plot of the rimmed raw data from highlighted columns in Supplemental Table S1, demonstrating that only the wells that contained samples of mitochondria had an appreciable increase in RFUs. Please click here to download this File.

Supplemental Figure S2: Oxygen consumption and respiratory control ratios. (A) Plot of the oxygen consumption data from Supplemental Table S1 to determine the State 2 and State 3 values by finding the equations of the best-fit lines. (B) RCR ratio calculation using the slopes of the lines in (A), which represent State 2 and State 3. Please click here to download this File.

Supplemental Table S1: Representative example of raw data. Highlighted columns are data that are used for figures and calculations (i.e., trimmed data). Please click here to download this File.

Supplemental Table S2: Representative example of trimmed data and subsequent oxygen consumption calculations. Please click here to download this File.

Discussion

Measuring mitochondrial function is useful when comparing different samples, such as disease versus non-disease states, different tissue types from the same animal, or between different sample types. We used the later comparison to test our hypothesis that there is a metabolic consequence to hybrid tree lizards that have introgressed mitochondria. There are a variety of ways to ascertain mitochondria function experimentally, including quantification of Δψ, total ATP content, ATP production, and respiratory control and rate of oxygen consumption. Oxygen consumption measurements of isolated mitochondria can be carried out with specialized equipment such as a Clark electrode or the more sophisticated Seahorse XFs. However, these methods can be either limiting by the sample size necessary to carry out measurements or cost-prohibitive. Multi-mode plate readers that can read fluorescence have become a standard piece of equipment in cell biology and biochemistry labs. Thus, an alternative approach to quantitatively analyze mitochondrial oxygen consumption is to use a commercially available kit with a fluorescent plate reader. This step-by-step protocol describes how to measure oxygen consumption as well as carry out the calculations. and economically in isolated mitochondria. Precisely measuring the ability of mitochondria to make ATP when supplied with ADP, referred to as State 3 respiration, allows direct comparison of OXPHOS between samples and differences can signify functional disparities. Furthermore, State 3 respiration also indicates the quality of mitochondria that were isolated, which can determine whether an isolation procedure has worked and/or the sample is good enough to be used in other assays. RCR values, which are calculated using State 2 (glutamate and malate only, to build Δψ) and State 3 values (glutamate and malate plus ADP, to allow Δψ to be used for ATP production), give yet another piece of information for comparison and about the health of the isolated mitochondria. The data presented demonstrate that the method offered in this paper can help decipher variations between different sample types and measure oxygen consumption rates and RCR values without the need for highly specialized equipment, thereby making these types of measurement more accessible.

In this protocol, we freshly isolated mitochondria from lizards, and oxygen consumption was determined using a MitoXpress Xtra Hs assay following the manufacturer's protocol and modified according to Hynes et al.⁴ and Will et al.⁵ Lizards ranged from 3.3 g to 5.1 g in total body mass with an average of 0.170 g of liver tissue/lizard, which yielded ~0.350 mL of concentrated, crude mitochondria. The isolated mitochondria were used in the oxygen consumption assay at 6 mg/mL protein. If the mitochondrial protein concentration, as determined in protocol step 1.7, is low then the sample can be re-spun at 10,000 × g and the pellet resuspended in a smaller volume of L-MIB. It should also be noted that we have successfully used as little as 4 mg/mL for small liver samples or when the entirety of the liver was not available.

To ensure gas and temperature equilibration of samples at the start of the assay, all solutions used in protocol section 2 should be prewarmed to 30 °C in a water bath. In protocol section 3, the rates of change of dissolved oxygen (mM/min) are extrapolated from the initial slopes of the decreasing oxygen concentration profiles for each sample (protocol step 3.3; Supplemental Table S1 and Supplemental Figure S1). The sample + G/M treatment rates represent State 2 mitochondrial respiration, whereas the sample with G/M + ADP treatment rates represent State 3 mitochondrial respiration. Thus, RCRs are calculated for each sample by dividing state 3 values by state 2 values for each sample (Supplemental Table S2).

Oxygen consumption can either be compared between unique samples as demonstrated here or among different treatments in one sample. In the latter case, additional controls may be necessary, such as vehicle-only control with and without mitochondria, to rule out the possibility that the chemical treatment itself does not interact with the phosphorescent probe. Furthermore, oxygen consumption and RCR data collected can be used in conjunction with data collected from the same samples with other assays. Once the mitochondria are isolated, they can be utilized in multiple assays to provide a more complete picture of their function¹². Measurement of Δψ, total ATP content, and ATP turnover rate are some of the assays that have been combined. Furthermore, the use of ETC, OXPHOS, and ATP synthase inhibitors with the assays gives an even fuller understanding of different mitochondrial sample types, strengthening any conclusions drawn from the data generated.

Another parameter that can be measured in isolated mitochondria is mitochondrial efficiency, which is defined as the P/O ratio of moles of ATP synthesized per mole of O₂³. P/O ratios could be calculated in an extension of this protocol by allowing State 3 mitochondria to exhaust the ADP supplied in the reaction and then adding an aliquot of ADP substrate followed by inhibition of the ATP synthase by the addition of oligomycin. These additional steps allow for the calculation of State 4, which occurs when mitochondria continue to consume oxygen and maintain a strong Δψ to make maximal ATP, provides information about the leakiness of mitochondria (if H⁺ ions are crossing back into the matrix without the use of the ATP synthase and therefore dissipating Δψ), ultimately indicating mitochondria dysfunction. Moreover, State 4 measurements can also reveal if the ADP/ATP translocator, which moves newly formed ATP from the matrix while at the same time bringing ADP substrate in, is functional and if any leakage of ADP and ATP across the membranes is occurring. The addition of oligomycin, which inhibits ATP synthase, is a control to determine if the mitochondria are leaky due to actual dysfunction. While measuring State 4 so that the P/O ratio could be calculated was beyond the scope of the information we sought for comparison among the three mitochondrial sample types, it is another piece of data about mitochondrial function and could be done as an extension of this protocol. Finally, it is worth noting that mitochondria can be isolated from cells in culture and different tissues from a mouse or rat (e.g., liver, heart, spinal cord) among others. Adjustments to the MIB and EB formulations as well as the isolation protocols may be necessary⁴^,¹³^,¹⁴^,¹⁵^,¹⁶. Furthermore, according to the manufacturer's instructions, this assay could be performed with whole, intact cells. Therefore, the method presented is not limited to lizards or liver mitochondria, allowing it to be suitable to a broad array of experimental models across many different scientific subdisciplines.

Declarações

The authors have nothing to disclose.

Acknowledgements

This research was funded by NSF CHE- 1229562 (VDGM) and grants from Elon University's Faculty Research and Development Committee (VDGM and GH) and the Undergraduate Research Program (AJ).

Materials

96-well black/optical bottom plates	Thermo Fisher	265301	Untreated black-wall plates with clear bottoms.
ADP	Sigma	A2754	Dilute 100 µM stock with EB immediately before use.
BSA	Thermo Fisher	BP1600-100	Make 2 mg/mL stock in water for protein assay.
Dulbeccos 1x PBS (-/-)	Sigma	D8537	Make sure the PBS is without Mg²⁺ or Ca²⁺ ions.
EGTA	Sigma	E3889
K₂HPO4	Sigma	P3786
KH₂PO4	Sigma	P0662
L-glutamic acid	Sigma	G1251
L-glutamic acid potassium salt	Sigma	S372226
L-malic acid	Sigma	M8304
L-malic acid mono-potassium salt	Sigma	49601
MitoXpress oxygen consumption kit	Agilent	MX-200-4	Kit contains probe stock and HS mineral oil.
MOPS	Sigma	M3183
Protien Assay Dye (5x)	BioRad	500-0006	Any protein assay can substitute.
R version 3.3	R Core Development Team 2016
Thermomax microplate reader EnSpire Multi-mode Plate reader and software	PerkinElmer		Standard fluorescent plate-reader
Trisma base	Sigma	T6066	Any version of Tris base can be utilized.

Referências

Drew, B., Leeuwenburgh, C. Method for measuring ATP production in isolated mitochondria: ATP production in brain and liver mitochondria of Fischer-344 rats with age and caloric restriction. American Journal of Physiology-Regulatory, Integrative andComparative Physiology. 285 (5), R1259-R1267 (2003).
Ballard, J. W. O., Pichaud, N. Mitochondrial DNA: more than an evolutionary bystander. Functional Ecology. 28 (1), 218-231 (2014).
Brand, M. D., Nicholls, D. G. Assessing mitochondrial dysfunction in cells. Biochemical Journal. 437 (3), 575-575 (2011).
Will, Y., Hynes, J., Ogurtsov, V. I., Papkovsky, D. B. Analysis of mitochondrial function using phosphorescent oxygen-sensitive probes. Nature Protocols. 1 (6), 2563-2572 (2006).
Hynes, J., et al. Investigation of drug-induced mitochondrial toxicity using fluorescence-based oxygen-sensitive probes. Toxicological Sciences. 92 (1), 186-200 (2006).
Lampl, T., Crum, J. A., Davis, T. A., Milligan, C., Del Gaizo Moore, V. Isolation and functional analysis of mitochondria from cultured cells and mouse tissue. Journal of Visualized Experiments. (97), 52076 (2015).
Frezza, C., Cipolat, S., Scorrano, L. Organelle isolation: functional mitochondria from mouse liver, muscle, and cultured fibroblasts. Nature Protocols. 2 (2), 287-295 (2007).
Dunn, D. A., Pinkert, C. A. Nuclear expression of a mitochondrial DNA gene: mitochondrial targeting of allotopically expressed mutant ATP6 in transgenic mice. Journal of Biomedicine and Biotechnology. 2012, 1-7 (2012).
Irwin, R. W., et al. Medroxyprogesterone acetate antagonizes estrogen up-regulation of brain mitochondrial function. Endocrinology. 152 (2), 556-567 (2011).
Ma, D., et al. Cancer cell mitochondria targeting by pancratistatin analogs is dependent on functional complex II and III. Scientific Reports. 7 (1), 42957 (2017).
Lakowicz, J. R. . Principles of fluorescence spectroscopy. , (2006).
Haenel, G. J., Del Gaizo Moore, V. Functional divergence of mitochondria and coevolution of genomes: cool mitochondria in hot Llzards. Physiological and Biochemical Zoology. 91 (5), 1068-1081 (2018).
Vinsant, S., et al. Characterization of early pathogenesis in the SOD1 G93A mouse model of ALS: part I, background and methods. Brain and Behavior. 3 (4), 335-350 (2013).
Del Gaizo Moore, V., et al. Chronic lymphocytic leukemia requires BCL2 to sequester prodeath BIM, explaining sensitivity to BCL2 antagonist ABT-737. Journal of Clinical Investigation. 117 (1), 112-121 (2007).
Chonghaile, T. N., et al. Pretreatment mitochondrial priming correlates with clinical response to cytotoxic chemotherapy. Science. 334 (6059), 1129-1133 (2011).
Del Gaizo, V., MacKenzie, J. A., Payne, R. M. Targeting proteins to mitochondria using TAT. Molecular Genetics and Metabolism. 80 (1-2), 170-180 (2003).