- Split View
-
Views
-
Cite
Cite
James Hynes, Lisa D. Marroquin, Vladimir I. Ogurtsov, Katerina N. Christiansen, Gregory J. Stevens, Dmitri B. Papkovsky, Yvonne Will, Investigation of Drug-Induced Mitochondrial Toxicity Using Fluorescence-Based Oxygen-Sensitive Probes, Toxicological Sciences, Volume 92, Issue 1, July 2006, Pages 186–200, https://doi.org/10.1093/toxsci/kfj208
- Share Icon Share
Abstract
Mitochondrial dysfunction is a common mechanism of drug-induced toxicity. Early identification of new chemical entities (NCEs) that perturb mitochondrial function is of significant importance to avoid attrition in later stages of drug development. One of the most informative ways of assessing mitochondrial dysfunction is by measuring mitochondrial oxygen consumption. However, the conventional polarographic method of measuring oxygen consumption is not amenable to high sample throughput or automation. We present an alternative, low-bulk, high-throughput approach to the analysis of isolated-mitochondrial oxygen consumption using luminescent oxygen-sensitive probes. These probes are dispensable and are analyzed in standard microtitre plates on a fluorescence plate reader. Respiratory substrate and adenosine diphosphate (ADP) dependencies of mitochondrial oxygen consumption were assessed using the fluorescence-based method, and results compared favourably to conventional polarographic analysis. To assess assay performance, the method was then applied to the analysis of a panel of classical modulators of oxidative phosphorylation. The effect of uncoupler concentration was analyzed in detail to identify factors which would be important in applying this method to large scale NCE screening and mechanistic investigations. Results demonstrate that the 96-well format can accommodate up to ∼ 200 compounds/day at a single concentration or alternatively IC50 values can be generated for ∼ 25 compounds. Throughput may be increased by moving to a 384-well plate format.
Over recent years, there has been a resurgence in interest in mitochondrial biology, due mainly to a growing appreciation of the complexity and importance of the cellular mitochondrial network. This network plays a central role in various cellular functions, producing 95% of cellular ATP requirements and participating in a variety of physiological processes. It is central to the progression and regulation of apoptosis (Waterhouse et al., 2002), and mitochondrial dysfunction has been implicated in numerous disease states, including neurological and cardiovascular disorders, as well as drug-induced toxicities and even aging (Amacher, 2005; Mattson and Kroemer, 2003; Moyle, 2005; Sullivan and Brown, 2005). As a result of this importance and complexity, any disturbance of mitochondrial function can have detrimental and wide-reaching implications.
Xenobiotics have long been suspected of disrupting mitochondrial function leading to injury. Recent advances in the understanding of the molecular mechanisms involved (Boelsterli, 2004) have revealed that these mechanisms are much more complex than originally envisioned. These advances are directly relevant to the area of drug development, as drugs, such as certain antiviral agents, lipid-lowering drugs, cholesterol-lowering drugs, antibiotics, analgesics, and cancer chemotherapeutics have all been implicated in inducing mitochondrial dysfunction. Not surprisingly, some of the most important targets of this drug-induced mitochondrial toxicity are involved in oxidative phosphorylation, including complexes I–IV of the electron transport chain (ETC), the F1/FO ATPase and the adenine nucleotide transporter (ANT). Mitochondrial DNA has also been shown to be a target, with the demonstration that the antiretroviral agent AZT (3′-Azido-3′-Deoxythymidine) induces depletion of mitochondrial DNA. This depletion decreases the levels and activity of encoded proteins including certain components of oxidative phosphorylation (Lewis et al., 2003). Other examples include certain fibrates used to treat hyperlipidemia which have been shown to inhibit ETC complex I (NADH-cytochrome c reductase) in vitro (Brunmair et al., 2004), while doxorubicin is known to cause mitochondrial dysfunction by selectively abstracting electrons from complex I through a redox cycling mechanism (Wallace, 2003). In vivo this dysfunction usually manifests itself as liver, skeletal muscle, cardiac muscle, or CNS toxicity.
Due to the serious implications of drug-induced mitochondrial toxicity, there is a need within drug development for earlier identification and elimination of new chemical entities (NCEs) which exhibit the potential to perturb mitochondrial function. Various analytical techniques have been employed to investigate mitochondrial function, including the use of absorbance spectroscopy to assess the redox state of mitochondrial cytochromes, fluorescent calcium dyes, measurement of ATP levels, reactive oxygen species (ROS) production, mitochondrial membrane potential (ΔΨm), and oxygen consumption. Whereas all these assays can provide information on a particular mechanism, with the exception of ROS production and ΔΨm, none are useful for high-throughput screening (HTS) approaches. Cellular ROS production is a useful parameter of mitochondrial function as increased ROS production has been implicated in various disease states (Li et al., 2003) and drug-induced toxicity (Jaeschke et al., 2002). The main sites of ROS production are complexes I and III of the ETC, where unpaired electrons “leak” and interact with oxygen to form highly reactive radical species such as superoxide ions and hydroxyl radicals (Brand et al., 2004). The degree of ROS production may be assessed by loading mitochondria with dyes, which, on reaction with these radical species, produce fluorescent derivatives (e.g., reduced xanthene dyes such as dichlorohydrofluorescein) (Haugland, 2002). This approach facilitates analysis of ROS providing an insight into the activity of the ETC, but can be limited due to the nonspecific nature of product formation.
While knowledge of altered ΔΨm is highly informative, in isolation it gives limited information on the source of any observed alteration and is insensitive to disruption of certain mitochondrial functions. Uncoupling and inhibition of ETC activity both cause a decrease in ΔΨm, while inhibition of the ATPase or the ANT has little or no effect. In addition, it has been noticed that fluorescent potential–sensitive probes themselves may interfere with mitochondrial function (Scaduto and Grotyohann, 1999).
One of the most informative ways of assessing mitochondrial function is by analyzing mitochondrial oxygen consumption. This parameter can be affected by the inhibition of either the citric acid cycle or the β-oxidation of fatty acids. The supply of reducing equivalents can also be limited by altered outer mitochondrial membrane permeability. Oxygen consumption can be modulated by the uncoupling of oxidative phosphorylation by xenobiotics, uncoupling proteins, or the opening of the mitochondrial permeability transition pore and by the inhibition of individual components of the ETC or the F1/FO ATP synthase. Other factors such as calcium-mediated alterations of regulatory mechanisms, cardiolipin damage, or inhibition of specific transporters can also contribute to altered oxygen consumption. These different pathways are illustrated schematically in Figure 1.
Polarography using the Clark-type oxygen electrode (Clark, 1959) has long been the main technique for measuring pO2 and oxygen consumption, and is still widely used on a laboratory scale. Although this measurement approach has proven very useful, the methodology is associated with a number of inherent limitations. These include the invasive nature of measurement, oxygen consumption by the electrode, sensitivity to mass exchange (stirring requirements), sterility and reuse issues, electrode poisoning, and signal drift (Papkovsky, 2004). However, the primary limitations in the context of toxicity assessment within drug discovery are the low sample throughput and the lack of flexibility associated with electrode-based systems.
Quenched-fluorescence oxygen sensing is an emerging technology which provides an attractive alternative to oxygen electrodes (Papkovsky, 2004). Using solid-state oxygen sensors, this methodology has been successfully applied to the analysis of biological samples (Wodnicka et al., 2000). These systems can however suffer from certain limitations, including unsatisfactory sensor-response times or measurement format inflexibility. More recently, the development of dispensable water-soluble oxygen probes (Hynes et al., 2003; O'Donovan et al., 2005) has increased the convenience, flexibility, and performance of this fluorescence-based oxygen-sensing methodology. The approach relies on the use of dispensable, long-decay phosphorescent oxygen probes, which confer increased measurement flexibility in comparison to standard solid-state sensors and may be applied to a broad range of applications (Papkovsky, 2004). The long emission decay of these probes also allows the use of time-resolved fluorescence detection; thereby improving both probe performance and resistance to optical interferences, particularly when working with complex biological samples.
In this study, we have explored the feasibility of using oxygen-sensitive probes and standard microtitre plate format to assess the function of isolated mitochondria. The main performance characteristics of the newly developed method of quantitative assessment of biological oxygen consumption are evaluated. Assay response to respiratory substrates and ADP activation is investigated, and the effect of classical inhibitors and uncouplers is assessed and compared to results obtained using polarographic analysis. The aim of the work outlined was to examine the utility of such a high-throughput analytical approach in the early screening of NCEs.
MATERIALS AND METHODS
Materials.
All chemicals were purchased from Sigma-Aldrich (St. Louis, MO) and were of highest purity available. Phosphorescent oxygen-sensitive probe, type A65N-1, was from Luxcel Biosciences (Cork, Ireland). The Bicinchoninic Acid (BCA) kit for protein determination was from Pierce (Rockford, IL). Black body clear bottom 96-well plates (Costar 3631) were purchased through VWR (Westchester, PA).
Animals.
Care and maintenance were in accordance with the principles described in the Guide for Care and Use of laboratory Animals (NIH Publication 85-23, 1985). Male Sprague-Dawley rats (150–180 g) were purchased from Charles River (Wilmington, MA). Animals were housed in pairs in a controlled environment with constant temperature (21 ± 2°C) and a 12-h light/dark cycle. Food and water were available ad libitum. Animals were euthanized with an overdose of carbon dioxide. Organs were rapidly excised and placed into ice-cold mitochondrial isolation buffers (see below).
Isolation of liver mitochondria.
Liver mitochondria were isolated as previously outlined (Lapidus and Sokolove, 1993) with minor modifications. All steps were carried out at 4°C. Briefly, 3 g of liver was minced and washed in solution I (210mM mannitol, 70mM sucrose, 5mM Hepes, 1mM EGTA, and 0.5% fatty acid free BSA [Sigma A0281], pH 7.4) until the homogenate was blood free. Five volumes of solution I were added and the tissue homogenized using a smooth glass grinder with Teflon pestle driven by a power drill. The homogenate was then adjusted to eight volumes with solution I and centrifuged at 700 × g for 10 min at 4°C. The supernatant was filtered through two layers of cheesecloth and recentrifuged for 10 min at 14,000 × g to precipitate the mitochondrial fraction. The supernatant was then discarded, mitochondria were washed by resuspending in 20 ml of solution I and spinning at 10,000 × g for 10 min at 4°C. This washing step was repeated in solution II (210mM mannitol, 70mM sucrose, 10mM MgCl2, 5mM K2HPO4, 10mM 3-(N-Morpholino) propanesulfonic Acid (MOPS), and 1mM EGTA, pH 7.4) (Heisler, 1991). Finally, the mitochondria were resuspended in 0.7 ml of solution II, and protein concentration was determined using the BCA method according to the manufacturer's protocol. Respiration buffer consisted of 250mM sucrose, 15mM KCl, 1mM EGTA, 5mM MgCl2, and 30mM K2HPO4, pH 7.4.
Isolation of muscle and heart mitochondria.
Heart and muscle mitochondria were isolated as described (Messer et al., 2004). Two rat hearts or approximately 3 g of mixed skeletal muscle (EDL, soleus, gastrocnemicus) were freed of blood and connective tissue and placed in nine volumes of solution I (100mM KCl, 40mM Tris-HCl, 10mM Tris-base, 5mM MgCl2, 1mM EDTA, and 1mM ATP, pH 7.4). Tissue was finely minced with Type XXIV protease (Sigma-P8038) added at 5 mg/g wet tissue, and then incubated for 7 min with mixing and additional mincing. Adding an equal volume of solution I terminated the protease digestion. The mixture was then homogenized for 30 s with an Ultra-Turrax tissue homogeniser (IKA, T25) at 11,000 rpm (setting I). The homogenate was centrifuged at 4°C for 10 min at 700 × g; the supernatant was rapidly filtered through two layers of cheesecloth and recentrifuged at 14,000 × g for 10 min at 4°C. The resulting supernatant was discarded, and the mitochondrial pellet was resuspended in solution II (100mM KCl, 40mM Tris-HCl, 10mM Tris-base, 1mM MgSO4, 0.1mM EDTA, 0.2mM ATP, and 2% BSA, pH 7.4) and centrifuged at 7000 × g for 10 min at 4°C. The supernatant was discarded and the mitochondria were subjected to washing with 20 ml of solution III (100mM KCl, 40mM Tris-HCl, 10mM Tris-base, 1mM MgSO4, 0.1mM EDTA, and 0.2mM ATP, pH 7.4) and centrifugation at 3500 × g. Finally, the mitochondrial pellet was resuspended in a minimal volume of solution IV (220mM Mannitol, 70mM Sucrose, 10mM Tris-HCl, and 1mM EGTA, pH 7.4) for further use. Protein concentration was determined using the BCA kit according to the manufacturer's protocol. Respiration buffer consisted of 100mM KCl, 50mM MOPS, 10mM K2HPO4, 10mM MgCl2, and 1mM EGTA, pH 7.4.
Polarographic measurements of mitochondrial respiration.
Measurements on a standard Oxytherm system (Hansatech Instruments, Norfolk England) equipped with a Clark-type electrode disc and temperature-controlled sample chamber were carried out at 30°C, using 2 ml of tissue specific respiration media and continuous stirring to ensure uniform distribution of dissolved oxygen throughout the experiment. Basal respiration (state 2) was initiated with either succinate (10mM) or glutamate/malate (both at 5.0mM) and recorded for 3–5 min. State 3 (maximal respiration) was initiated by adding a bolus of ADP giving a final concentration of 250μM. Respiration rates were determined according to Estabrook, 1967. All additions were made through a small diameter plunger with stopper using Hamilton syringes. After all ADP had been consumed, mitochondria returned to basal respiration (state 4).
Fluorescence-based assay of mitochondrial respiration.
A65N-1 oxygen probe supplied as dry reagent in a vial, 1 nmol, was reconstituted in 1 ml of mitochondrial incubation buffer and diluted to a concentration of 100nM. One hundred microliters of this solution was pipetted into each well of a 96-well plate (10 pmol of probe per well). Fifty microliters of mitochondria stock solutions were added to each well giving the desired final concentration of mitochondria, followed by 50 μl of substrate (25mM and 12.5/12.5mM for succinate and glutamate/malate final concentration, respectively) without or with ADP (1.65mM final concentration) in incubation buffer. For drug treatments, compound stock solutions were prepared in DMSO and added to the wells to give the indicated final concentrations (the final DMSO content was not more than 0.5% vol/vol). All drug concentrations are presented as nmol/mg of mitochondrial protein. Finally, 100 μl of heavy mineral oil was added to each well to seal the samples from ambient oxygen, and the plate was placed in a fluorescence plate reader Safire2 (Tecan) equilibrated at 30°C and monitored over a period of 20–120 min (depending on the experiment) measuring probe fluorescence signal in each well every 1.5 min in kinetic mode. Instrument settings were 380/650 nm excitation/emission, filters, a delay time of 30 μs and a measurement window of 100 μs, and active temperature control of the microplate compartment at 30°C. To ensure gas and temperature equilibration of samples at the start of the assay, all the dispensing steps were carried out at 30°C using prewarmed solutions and a Multio-Blok heater (Barnstead/LabLine, Melrose Park, IL) holding the microplate.
RESULTS
Analysis of Oxygen Consumption in Isolated Rat Liver Mitochondria: Comparison of Polarographic and Oxygen-Sensitive Fluorescence-Based Methods
To characterize the performance of the proposed fluorescence-based mitochondrial oxygen consumption assay, rat liver mitochondria were analyzed at 30°C in a standard 96-well plate. Both glutamate/malate- and succinate-driven respiratory activity were assessed in basal and ADP-activated states (state 2 and state 3, respectively). The results obtained were compared with traditional polarographic measurements carried out in parallel on an Oxytherm system. A typical polarographic trace of such mitochondrial oxygen consumption is presented in Figure 2A. Mitochondria at 0.25 mg/ml were incubated at 30°C with substrate (succinate), and basal oxygen consumption (state 2) was recorded for several minutes. A defined amount of ADP was then added, and maximal respiration (state 3) was recorded. After all ADP was consumed mitochondria returned to basal respiration (state 4).
Figure 2B shows the raw data output of the fluorescence-based analysis in standard microtitre plate using the soluble oxygen-sensitive probes. Initially, probe emission is quenched by dissolved oxygen, present at air-saturated concentrations. Depletion of dissolved oxygen by the activity of the ETC reduces this quenching effect, resulting in an increase in probe fluorescent signal over time and mitochondrial concentration must be carefully optimized to ensure accuracy. A layer of mineral oil is applied on top of each sample to restrict back diffusion of atmospheric oxygen, thereby increasing assay sensitivity and reproducibility. This prevents reagent addition during measurement requiring each respiration state to be measured separately. Using this approach, glutamate/malate- and succinate-driven respiratory activity was analyzed in both state 2 and state 3 for six different mitochondrial concentrations (Fig. 2B). It should be noted that the probe is covalently bound to bovine albumin with a molecular weight of 65 kDa which prevents its uptake into mitochondria eliminating possible interference with organelle function.
Representative profiles of succinate-driven respiration measured in state 2 are presented in Figure 2C showing the effect of increasing protein concentrations on measured probe signal. Under these conditions, a protein concentration of 1 mg/ml causes complete sample deoxygenation in ∼10 min resulting in a sixfold increase in probe emission intensity. Reduced protein concentrations result in lower rates of oxygen consumption, which, in turn, result in reduced rates of probe emission intensity increase. Figure 2D shows the relationship between the rate of increase of probe emission intensity and mitochondrial protein concentration, which is close to linear. Such analysis allows the identification of an appropriate concentration of mitochondrial protein for analysis and displays the low coefficient of variance (CV) values associated with this methodology (< 3%).
Closer inspection of the data presented in Figure 2B allows an examination of probe response to ADP activation of both glutamate/malate- and succinate-driven mitochondrial oxygen consumption. Figure 3A shows probe signal increasing over time in response to basal and ADP-activated glutamate/malate-driven respiration in a sample containing mitochondrial protein at 0.25 mg/ml. Figure 3B shows similar data for succinate-driven respiration in a sample containing mitochondrial protein at 0.125 mg/ml. As is the case for polarographic measurements (Fig. 2A), basal respiration (state 2) is seen to be substantially lower than ADP-driven respiration (state 3). Higher rates of probe signal increase are observed for succinate-driven respiration than for glutamate/malate-driven respiration, while ADP activation causes a more rapid increase in probe signal for glutamate-driven than for succinate-driven respiration. Furthermore, Figure 3 also demonstrates the excellent reproducibility of the method with CV values of less than 5% for quadruplicate analysis.
Using the known analytical relationship between the probe fluorescent signal and the concentration of dissolved oxygen, measured profiles of fluorescence may be converted into oxygen profiles (for detail see “Materials and Methods” section). Such a conversion is presented in Figure 4 and illustrates the decrease in dissolved oxygen concentration caused by basal glutamate/malate-driven respiration at increasing mitochondrial protein concentration. This transformation allows a more direct comparison of the new fluorescence assay with polarographic analysis.
Table 1 presents rates of change of dissolved oxygen calculated from such fluorescence data and compares these rates to the oxygen consumption rates measured using the polarographic system. Both data sets show that mitochondria are well coupled, showing low basal oxygen consumption rates (state 2) and high rates upon the addition of ADP, resulting in ratios (state 3/2) of approximately 3 for succinate-driven and 6 for glutamate/malate-driven respiration, respectively. Absolute rates of oxygen consumption using the fluorescence technique were lower, but yielded ratios of ADP-driven respiration (state 3) to basal respiration (state 2) that were comparable to the ratios obtained using the polarographic method (see “Discussion” section).
. | n . | Substrates . | State 2, mean ± SD (nmols/min/mg) . | State 3, mean ± SD (nmols/min/mg) . | Ratio, state 3/2, mean ± SD . |
---|---|---|---|---|---|
Polarimetry | |||||
Day 1 | 3 | Succinate | 25.0 ± 1.5 | 105.8 ± 2.6 | 4.2 ± 0.14 |
Glu/mal | 10.8 ± 0.84 | 73.2 ± 0.58 | 6.8 ± 0.54 | ||
Day 2 | 3 | Succinate | 23.8 ± 2.8 | 107.4 ± 7.2 | 4.6 ± 0.46 |
Glu/mal | 10.6 ± 2.2 | 74.2 ± 5.6 | 7.1 ± 1.1 | ||
Day 3 | 3 | Succinate | 31.0 ± 2.6 | 125.8 ± 10.2 | 4.1 ± 0.15 |
Glu/mal | 11.6 ± 1.0 | 80.4 ± 7.2 | 6.9 ± 0.16 | ||
Day 4 | 4 | Succinate | 29.2 ± 2.4 | 119.8 ± 6.4 | 4.1 ± 0.33 |
Glu/mal | 13.6 ± 0.52 | 77.2 ± 2.0 | 5.7 ± 0.14 | ||
Fluorescence-based | |||||
Day 1 | 4 | Succinate | 20.1 ± 0.5 | 42.2 ± 1.2 | 2.1 |
Glu/mal | 4.3 ± 0.4 | 26.6 ± 0.4 | 6.2 | ||
Day 2 | 4 | Succinate | 18.5 ± 0.4 | 54.4 ± 1.5 | 2.9 |
Glu/mal | 5.8 ± 0.8 | 37.1 ± 1.2 | 6.3 | ||
Day 3 | 4 | Succinate | 15.5 ± 0.9 | 56.6 ± 1.1 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 37.8 ± 0.8 | 10.4 | ||
Day 4 | 4 | Succinate | 15.5 ± 0.8 | 56.8 ± 1.6 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 27.1 ± 1.2 | 6.9 |
. | n . | Substrates . | State 2, mean ± SD (nmols/min/mg) . | State 3, mean ± SD (nmols/min/mg) . | Ratio, state 3/2, mean ± SD . |
---|---|---|---|---|---|
Polarimetry | |||||
Day 1 | 3 | Succinate | 25.0 ± 1.5 | 105.8 ± 2.6 | 4.2 ± 0.14 |
Glu/mal | 10.8 ± 0.84 | 73.2 ± 0.58 | 6.8 ± 0.54 | ||
Day 2 | 3 | Succinate | 23.8 ± 2.8 | 107.4 ± 7.2 | 4.6 ± 0.46 |
Glu/mal | 10.6 ± 2.2 | 74.2 ± 5.6 | 7.1 ± 1.1 | ||
Day 3 | 3 | Succinate | 31.0 ± 2.6 | 125.8 ± 10.2 | 4.1 ± 0.15 |
Glu/mal | 11.6 ± 1.0 | 80.4 ± 7.2 | 6.9 ± 0.16 | ||
Day 4 | 4 | Succinate | 29.2 ± 2.4 | 119.8 ± 6.4 | 4.1 ± 0.33 |
Glu/mal | 13.6 ± 0.52 | 77.2 ± 2.0 | 5.7 ± 0.14 | ||
Fluorescence-based | |||||
Day 1 | 4 | Succinate | 20.1 ± 0.5 | 42.2 ± 1.2 | 2.1 |
Glu/mal | 4.3 ± 0.4 | 26.6 ± 0.4 | 6.2 | ||
Day 2 | 4 | Succinate | 18.5 ± 0.4 | 54.4 ± 1.5 | 2.9 |
Glu/mal | 5.8 ± 0.8 | 37.1 ± 1.2 | 6.3 | ||
Day 3 | 4 | Succinate | 15.5 ± 0.9 | 56.6 ± 1.1 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 37.8 ± 0.8 | 10.4 | ||
Day 4 | 4 | Succinate | 15.5 ± 0.8 | 56.8 ± 1.6 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 27.1 ± 1.2 | 6.9 |
. | n . | Substrates . | State 2, mean ± SD (nmols/min/mg) . | State 3, mean ± SD (nmols/min/mg) . | Ratio, state 3/2, mean ± SD . |
---|---|---|---|---|---|
Polarimetry | |||||
Day 1 | 3 | Succinate | 25.0 ± 1.5 | 105.8 ± 2.6 | 4.2 ± 0.14 |
Glu/mal | 10.8 ± 0.84 | 73.2 ± 0.58 | 6.8 ± 0.54 | ||
Day 2 | 3 | Succinate | 23.8 ± 2.8 | 107.4 ± 7.2 | 4.6 ± 0.46 |
Glu/mal | 10.6 ± 2.2 | 74.2 ± 5.6 | 7.1 ± 1.1 | ||
Day 3 | 3 | Succinate | 31.0 ± 2.6 | 125.8 ± 10.2 | 4.1 ± 0.15 |
Glu/mal | 11.6 ± 1.0 | 80.4 ± 7.2 | 6.9 ± 0.16 | ||
Day 4 | 4 | Succinate | 29.2 ± 2.4 | 119.8 ± 6.4 | 4.1 ± 0.33 |
Glu/mal | 13.6 ± 0.52 | 77.2 ± 2.0 | 5.7 ± 0.14 | ||
Fluorescence-based | |||||
Day 1 | 4 | Succinate | 20.1 ± 0.5 | 42.2 ± 1.2 | 2.1 |
Glu/mal | 4.3 ± 0.4 | 26.6 ± 0.4 | 6.2 | ||
Day 2 | 4 | Succinate | 18.5 ± 0.4 | 54.4 ± 1.5 | 2.9 |
Glu/mal | 5.8 ± 0.8 | 37.1 ± 1.2 | 6.3 | ||
Day 3 | 4 | Succinate | 15.5 ± 0.9 | 56.6 ± 1.1 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 37.8 ± 0.8 | 10.4 | ||
Day 4 | 4 | Succinate | 15.5 ± 0.8 | 56.8 ± 1.6 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 27.1 ± 1.2 | 6.9 |
. | n . | Substrates . | State 2, mean ± SD (nmols/min/mg) . | State 3, mean ± SD (nmols/min/mg) . | Ratio, state 3/2, mean ± SD . |
---|---|---|---|---|---|
Polarimetry | |||||
Day 1 | 3 | Succinate | 25.0 ± 1.5 | 105.8 ± 2.6 | 4.2 ± 0.14 |
Glu/mal | 10.8 ± 0.84 | 73.2 ± 0.58 | 6.8 ± 0.54 | ||
Day 2 | 3 | Succinate | 23.8 ± 2.8 | 107.4 ± 7.2 | 4.6 ± 0.46 |
Glu/mal | 10.6 ± 2.2 | 74.2 ± 5.6 | 7.1 ± 1.1 | ||
Day 3 | 3 | Succinate | 31.0 ± 2.6 | 125.8 ± 10.2 | 4.1 ± 0.15 |
Glu/mal | 11.6 ± 1.0 | 80.4 ± 7.2 | 6.9 ± 0.16 | ||
Day 4 | 4 | Succinate | 29.2 ± 2.4 | 119.8 ± 6.4 | 4.1 ± 0.33 |
Glu/mal | 13.6 ± 0.52 | 77.2 ± 2.0 | 5.7 ± 0.14 | ||
Fluorescence-based | |||||
Day 1 | 4 | Succinate | 20.1 ± 0.5 | 42.2 ± 1.2 | 2.1 |
Glu/mal | 4.3 ± 0.4 | 26.6 ± 0.4 | 6.2 | ||
Day 2 | 4 | Succinate | 18.5 ± 0.4 | 54.4 ± 1.5 | 2.9 |
Glu/mal | 5.8 ± 0.8 | 37.1 ± 1.2 | 6.3 | ||
Day 3 | 4 | Succinate | 15.5 ± 0.9 | 56.6 ± 1.1 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 37.8 ± 0.8 | 10.4 | ||
Day 4 | 4 | Succinate | 15.5 ± 0.8 | 56.8 ± 1.6 | 3.6 |
Glu/mal | 3.9 ± 0.5 | 27.1 ± 1.2 | 6.9 |
To further assess the fluorescence-based method, mitochondria isolated from liver, heart, and skeletal muscle tissue were prepared concurrently as described in “Materials and Methods” section and subjected to parallel polarographic and fluorescence-based monitoring of oxygen consumption. Glutamate/malate-driven respiration was assessed, as this substrate is more preferred than succinate by heart and muscle, yielding substantially higher increases in respiratory activity on ADP stimulation. Polarographic analysis was carried out on samples containing mitochondrial protein at 0.25 mg/ml, while the higher throughput capabilities of the fluorescence-based approach allowed the analysis of basal and ADP-stimulated respiration at six different protein concentrations. Sample fluorescence-based data are presented in Figure 5. Polarographic analysis revealed state 3 respiration rates for glutamate/malate of 73.2 ± 0.58 (Table 1: Day 1 analysis), 159.6 ± 4.5 (n = 2), and 213.8 ± 8.2 (n = 2) nmol/min/mg for liver, heart, and skeletal muscle–derived mitochondria, respectively. A similar trend was observed using the fluorescence-based approach, with skeletal muscle mitochondria producing the greatest consumption rates. Reduced rates were observed for heart with liver mitochondria showing the lowest levels of oxygen consumption. Reproducible measurements were again evident with CV values of less than 3%.
Monitoring the Effects of Classical Inhibitors and Uncouplers
To examine the ability of these florescence-based oxygen probes to identify compounds, that perturb mitochondrial function, the effects of several classical mitochondrial inhibitors and uncouplers on mitochondrial oxygen consumption were examined. Results were again compared to those obtained from parallel polarographic analysis. The compounds tested, along with site of action and the effect on mitochondrial oxygen consumption, assessed using both methods are presented in Table 2, with representative results obtained using fluorescence method shown in Figures 6–8.
. | . | Glutamate/malate-driven . | . | Succinate-driven . | . | ||
---|---|---|---|---|---|---|---|
Compound . | Mechanism [Reference] . | State 2 . | State 3 . | State 2 . | State 3 . | ||
Rotenone | Inhibits NADH dehydrogenase (Horgan et al., 1968) | Inhibition | Inhibition | No effect | No effect | ||
TTFA | Inhibits succinate dehydrogenase (Kolesova et al., 1989) | No effect | No effect | Inhibition | Inhibition | ||
Antimycin | Inhibits cytochrome b-c1 (Schagger et al., 1995) | Inhibition | Inhibition | Inhibition | Inhibition | ||
KCN | Inhibits cytochrome oxidase (Isom and Way, 1984) | Inhibition | Inhibition | Inhibition | Inhibition | ||
Oligomycin | Inhibits FO/F1 ATPase (Walker et al., 1995) | No effect | Inhibition | No effect | Inhibition | ||
FCCP | Uncoupler (Cunarro and Weiner, 1975) | Increase | No effect | Increase | No effect |
. | . | Glutamate/malate-driven . | . | Succinate-driven . | . | ||
---|---|---|---|---|---|---|---|
Compound . | Mechanism [Reference] . | State 2 . | State 3 . | State 2 . | State 3 . | ||
Rotenone | Inhibits NADH dehydrogenase (Horgan et al., 1968) | Inhibition | Inhibition | No effect | No effect | ||
TTFA | Inhibits succinate dehydrogenase (Kolesova et al., 1989) | No effect | No effect | Inhibition | Inhibition | ||
Antimycin | Inhibits cytochrome b-c1 (Schagger et al., 1995) | Inhibition | Inhibition | Inhibition | Inhibition | ||
KCN | Inhibits cytochrome oxidase (Isom and Way, 1984) | Inhibition | Inhibition | Inhibition | Inhibition | ||
Oligomycin | Inhibits FO/F1 ATPase (Walker et al., 1995) | No effect | Inhibition | No effect | Inhibition | ||
FCCP | Uncoupler (Cunarro and Weiner, 1975) | Increase | No effect | Increase | No effect |
. | . | Glutamate/malate-driven . | . | Succinate-driven . | . | ||
---|---|---|---|---|---|---|---|
Compound . | Mechanism [Reference] . | State 2 . | State 3 . | State 2 . | State 3 . | ||
Rotenone | Inhibits NADH dehydrogenase (Horgan et al., 1968) | Inhibition | Inhibition | No effect | No effect | ||
TTFA | Inhibits succinate dehydrogenase (Kolesova et al., 1989) | No effect | No effect | Inhibition | Inhibition | ||
Antimycin | Inhibits cytochrome b-c1 (Schagger et al., 1995) | Inhibition | Inhibition | Inhibition | Inhibition | ||
KCN | Inhibits cytochrome oxidase (Isom and Way, 1984) | Inhibition | Inhibition | Inhibition | Inhibition | ||
Oligomycin | Inhibits FO/F1 ATPase (Walker et al., 1995) | No effect | Inhibition | No effect | Inhibition | ||
FCCP | Uncoupler (Cunarro and Weiner, 1975) | Increase | No effect | Increase | No effect |
. | . | Glutamate/malate-driven . | . | Succinate-driven . | . | ||
---|---|---|---|---|---|---|---|
Compound . | Mechanism [Reference] . | State 2 . | State 3 . | State 2 . | State 3 . | ||
Rotenone | Inhibits NADH dehydrogenase (Horgan et al., 1968) | Inhibition | Inhibition | No effect | No effect | ||
TTFA | Inhibits succinate dehydrogenase (Kolesova et al., 1989) | No effect | No effect | Inhibition | Inhibition | ||
Antimycin | Inhibits cytochrome b-c1 (Schagger et al., 1995) | Inhibition | Inhibition | Inhibition | Inhibition | ||
KCN | Inhibits cytochrome oxidase (Isom and Way, 1984) | Inhibition | Inhibition | Inhibition | Inhibition | ||
Oligomycin | Inhibits FO/F1 ATPase (Walker et al., 1995) | No effect | Inhibition | No effect | Inhibition | ||
FCCP | Uncoupler (Cunarro and Weiner, 1975) | Increase | No effect | Increase | No effect |
As can be seen in Figure 6, the complex II inhibitor Thenoyl trifluoracetone (TTFA), complex III inhibitor antimycin, and cytochrome oxidase inhibitor Potassium cyanide (KCN) were all seen to significantly reduce basal succinate-driven respiration (Fig. 6A). FCCP, an uncoupler, which dissipates mitochondrial membrane potential, caused an increase in basal oxygen consumption, while the F1/FO-ATPase inhibitor oligomycin had no effect. For ADP-stimulated succinate-driven respiration (Fig. 6B), oligomycin was seen to cause inhibition. Surprisingly, FCCP treatment was observed to induce slight inhibition of respiration. Similar results were observed for glutamate/malate-driven respiration; however, in this instance, rotenone was seen to cause inhibition of both basal and ADP-activated states (data not shown), while the complex II inhibitor TTFA had no effect on either states.
Previous analysis of the effect of the uncoupler FCCP at a fixed concentration revealed an increase in respiratory activity in state 2 and a decrease in state 3. To investigate the implications of this observation for the analysis of uncouplers at single concentrations, the dose dependence of this observed inhibition was assessed. Results presented in Figure 7 illustrate that in state 2, up to a critical concentration of 0.625 nmol/mg, increasing concentrations of FCCP cause increased oxygen consumption, with inhibition being observed above this concentration (Fig. 7A). Conversely in state 3, increasing concentrations of FCCP caused considerable inhibition (Fig. 7B). Figure 8 illustrates the effect of KCN (a cytochrome c oxidase inhibitor) on glutamate/malate-driven state 3 respiration. Unprocessed fluorescence data are presented in Figure 8A, in which increasing concentrations of KCN are seen to reduce respiration rates, thereby providing semiquantitative dose-response information. Conversion of these fluorescence profiles into oxygen profiles (Fig. 8B) allows calculation of corresponding changes in dissolved oxygen and IC50 values. The value calculated for KCN determined using this method (Fig. 5C) is 370 nmol/mg protein which is well below the amount usually used in polarimetry experiments to achieve full inhibition (1 umol/mg protein).
The RST Platform Can be Used for Mechanistic Approaches and to Generate IC50 Values
From the data presented it is clear that, for a comprehensive assessment of a compound's ability to perturb mitochondrial function, an assessment must be carried out on both basal and ADP-activated glutamate/malate- and succinate-driven respiration. In addition, results with FCCP suggest that these assessments should be carried at a number of concentrations. To examine the feasibility of this approach in the assessment of mitochondrial toxicity, five compounds of known mitochondrial toxicity were examined using this fluorescence-based approach. The observed in vivo toxicities and the proposed mechanisms for each drug are summarized in Table 3. One 96-well plate was run to determine the effect of treatment on basal and ADP-activated glutamate/malate-driven respiration and another to determine the effect of treatment on basal and ADP-activated succinate-driven respiration. Repeated experiments gave similar results.
Drug . | Therapeutic application . | Effect on mitochondria . | Observed in vivo toxicity . | Reference . |
---|---|---|---|---|
Diclofenac | Anti-inflammatory | Uncoupler; ANT and ATPase inhibitor; concentration tested: 30uM | Liver toxicant | (Moreno-Sanchez et al., 1999) |
Fenofibrate | Hyperlipidemia | ΔΨm depolarization; complex I inhibitor; inhibition of state 3 with glutamate/malate but not succinate; concentration tested: 75uM | Liver toxicant, hepatic, or renal dysfunction | (Zhou and Wallace, 1999; Brunmair et al., 2004) |
Flutamide | Antiandrogen/prostate cancer | Inhibition of state 3 with glutamate/malate and succinate; increase in state 4 with succinate; concentration tested: 50uM | Liver toxicant | (Fau et al., 1994) |
Tamoxifen | Breast cancer therapy | Uncoupler | Ocular, liver toxicity, agranulocytosis | (Cardoso et al., 2001) |
Quinidine | Antiarrhythmic - Antimalarial | Inhibits mitochondrial potassium channel at 0.1mM; inhibits ATPase; IC50 4.8 ± 0.6mM | Liver toxicity, ventricular tachycardia | (Mironova et al., 1997; Almotrefi, 1993) |
Drug . | Therapeutic application . | Effect on mitochondria . | Observed in vivo toxicity . | Reference . |
---|---|---|---|---|
Diclofenac | Anti-inflammatory | Uncoupler; ANT and ATPase inhibitor; concentration tested: 30uM | Liver toxicant | (Moreno-Sanchez et al., 1999) |
Fenofibrate | Hyperlipidemia | ΔΨm depolarization; complex I inhibitor; inhibition of state 3 with glutamate/malate but not succinate; concentration tested: 75uM | Liver toxicant, hepatic, or renal dysfunction | (Zhou and Wallace, 1999; Brunmair et al., 2004) |
Flutamide | Antiandrogen/prostate cancer | Inhibition of state 3 with glutamate/malate and succinate; increase in state 4 with succinate; concentration tested: 50uM | Liver toxicant | (Fau et al., 1994) |
Tamoxifen | Breast cancer therapy | Uncoupler | Ocular, liver toxicity, agranulocytosis | (Cardoso et al., 2001) |
Quinidine | Antiarrhythmic - Antimalarial | Inhibits mitochondrial potassium channel at 0.1mM; inhibits ATPase; IC50 4.8 ± 0.6mM | Liver toxicity, ventricular tachycardia | (Mironova et al., 1997; Almotrefi, 1993) |
Drug . | Therapeutic application . | Effect on mitochondria . | Observed in vivo toxicity . | Reference . |
---|---|---|---|---|
Diclofenac | Anti-inflammatory | Uncoupler; ANT and ATPase inhibitor; concentration tested: 30uM | Liver toxicant | (Moreno-Sanchez et al., 1999) |
Fenofibrate | Hyperlipidemia | ΔΨm depolarization; complex I inhibitor; inhibition of state 3 with glutamate/malate but not succinate; concentration tested: 75uM | Liver toxicant, hepatic, or renal dysfunction | (Zhou and Wallace, 1999; Brunmair et al., 2004) |
Flutamide | Antiandrogen/prostate cancer | Inhibition of state 3 with glutamate/malate and succinate; increase in state 4 with succinate; concentration tested: 50uM | Liver toxicant | (Fau et al., 1994) |
Tamoxifen | Breast cancer therapy | Uncoupler | Ocular, liver toxicity, agranulocytosis | (Cardoso et al., 2001) |
Quinidine | Antiarrhythmic - Antimalarial | Inhibits mitochondrial potassium channel at 0.1mM; inhibits ATPase; IC50 4.8 ± 0.6mM | Liver toxicity, ventricular tachycardia | (Mironova et al., 1997; Almotrefi, 1993) |
Drug . | Therapeutic application . | Effect on mitochondria . | Observed in vivo toxicity . | Reference . |
---|---|---|---|---|
Diclofenac | Anti-inflammatory | Uncoupler; ANT and ATPase inhibitor; concentration tested: 30uM | Liver toxicant | (Moreno-Sanchez et al., 1999) |
Fenofibrate | Hyperlipidemia | ΔΨm depolarization; complex I inhibitor; inhibition of state 3 with glutamate/malate but not succinate; concentration tested: 75uM | Liver toxicant, hepatic, or renal dysfunction | (Zhou and Wallace, 1999; Brunmair et al., 2004) |
Flutamide | Antiandrogen/prostate cancer | Inhibition of state 3 with glutamate/malate and succinate; increase in state 4 with succinate; concentration tested: 50uM | Liver toxicant | (Fau et al., 1994) |
Tamoxifen | Breast cancer therapy | Uncoupler | Ocular, liver toxicity, agranulocytosis | (Cardoso et al., 2001) |
Quinidine | Antiarrhythmic - Antimalarial | Inhibits mitochondrial potassium channel at 0.1mM; inhibits ATPase; IC50 4.8 ± 0.6mM | Liver toxicity, ventricular tachycardia | (Mironova et al., 1997; Almotrefi, 1993) |
Changes in mitochondrial respiration upon the addition of drugs were analyzed using an eight-point two-fold dilution series. The four drugs were also analyzed at the highest concentration using polarimetry. It has been reported previously that, for some drugs, the observed effect can change with increasing protein concentration (Moreno-Sanchez et al., 1999); therefore, all drug concentrations are normalized to mitochondrial protein concentration and are presented in units of nmol/mg. Using polarimetry, diclofenac (500 nmol/mg protein) showed a 300% increase in basal glutamate-driven respiration (state 2, Table 4). No further increase in oxygen consumption was observed upon ADP addition (state 3). This treated state 3 rate is however 40% lower than the untreated state 3 rate. Diclofenac treatment was seen to increase basal succinate-driven respiration to a maximum, with addition of ADP returning oxygen consumption to untreated basal respiration rates.
. | State 2 . | . | . | . | Ratio . | . | . | ||
---|---|---|---|---|---|---|---|---|---|
Drugs . | Pre Drug . | Post Drug . | State 3 . | State 4 . | State 3/2 . | State 3/4 . | Drug (nmol/mg protein) . | ||
Glutamate/malate | |||||||||
DMSO | 10.0 | 10.4 | 64.5 | 10.8 | 6.2 | 6.0 | 0.5% vol/vol | ||
Tamoxifen | 7.8 | 13.2 | 7.2 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 8.4 | 11.0 | 8.8 | n/a | n/a | n/a | 500 | ||
Flutamide | 8.7 | 10.2 | 8.3 | n/a | n/a | n/a | 500 | ||
Diclofenac | 10.4 | 46.2 | 41.6 | n/a | n/a | n/a | 500 | ||
Quinidine | 9.0 | 12.9 | 51.7 | 9.4 | 4.0 | 5.5 | 500 | ||
Succinate | |||||||||
DMSO | 24.0 | 25.7 | 95.4 | 19.2 | 3.7 | 4.79 | 0.5% vol/vol | ||
Tamoxifen | 24.2 | 20.9 | 11.4 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 22.8 | 24.1 | 36.2 | n/a | n/a | n/a | 500 | ||
Flutamide | 26.2 | 127.1 | 96.5 | n/a | n/a | n/a | 500 | ||
Diclofenac | 27.9 | 94.8 | 23.8 | n/a | n/a | n/a | 500 | ||
Quinidine | 25.4 | 31.1 | 89.8 | 24.8 | 2.9 | 3.62 | 500 |
. | State 2 . | . | . | . | Ratio . | . | . | ||
---|---|---|---|---|---|---|---|---|---|
Drugs . | Pre Drug . | Post Drug . | State 3 . | State 4 . | State 3/2 . | State 3/4 . | Drug (nmol/mg protein) . | ||
Glutamate/malate | |||||||||
DMSO | 10.0 | 10.4 | 64.5 | 10.8 | 6.2 | 6.0 | 0.5% vol/vol | ||
Tamoxifen | 7.8 | 13.2 | 7.2 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 8.4 | 11.0 | 8.8 | n/a | n/a | n/a | 500 | ||
Flutamide | 8.7 | 10.2 | 8.3 | n/a | n/a | n/a | 500 | ||
Diclofenac | 10.4 | 46.2 | 41.6 | n/a | n/a | n/a | 500 | ||
Quinidine | 9.0 | 12.9 | 51.7 | 9.4 | 4.0 | 5.5 | 500 | ||
Succinate | |||||||||
DMSO | 24.0 | 25.7 | 95.4 | 19.2 | 3.7 | 4.79 | 0.5% vol/vol | ||
Tamoxifen | 24.2 | 20.9 | 11.4 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 22.8 | 24.1 | 36.2 | n/a | n/a | n/a | 500 | ||
Flutamide | 26.2 | 127.1 | 96.5 | n/a | n/a | n/a | 500 | ||
Diclofenac | 27.9 | 94.8 | 23.8 | n/a | n/a | n/a | 500 | ||
Quinidine | 25.4 | 31.1 | 89.8 | 24.8 | 2.9 | 3.62 | 500 |
. | State 2 . | . | . | . | Ratio . | . | . | ||
---|---|---|---|---|---|---|---|---|---|
Drugs . | Pre Drug . | Post Drug . | State 3 . | State 4 . | State 3/2 . | State 3/4 . | Drug (nmol/mg protein) . | ||
Glutamate/malate | |||||||||
DMSO | 10.0 | 10.4 | 64.5 | 10.8 | 6.2 | 6.0 | 0.5% vol/vol | ||
Tamoxifen | 7.8 | 13.2 | 7.2 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 8.4 | 11.0 | 8.8 | n/a | n/a | n/a | 500 | ||
Flutamide | 8.7 | 10.2 | 8.3 | n/a | n/a | n/a | 500 | ||
Diclofenac | 10.4 | 46.2 | 41.6 | n/a | n/a | n/a | 500 | ||
Quinidine | 9.0 | 12.9 | 51.7 | 9.4 | 4.0 | 5.5 | 500 | ||
Succinate | |||||||||
DMSO | 24.0 | 25.7 | 95.4 | 19.2 | 3.7 | 4.79 | 0.5% vol/vol | ||
Tamoxifen | 24.2 | 20.9 | 11.4 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 22.8 | 24.1 | 36.2 | n/a | n/a | n/a | 500 | ||
Flutamide | 26.2 | 127.1 | 96.5 | n/a | n/a | n/a | 500 | ||
Diclofenac | 27.9 | 94.8 | 23.8 | n/a | n/a | n/a | 500 | ||
Quinidine | 25.4 | 31.1 | 89.8 | 24.8 | 2.9 | 3.62 | 500 |
. | State 2 . | . | . | . | Ratio . | . | . | ||
---|---|---|---|---|---|---|---|---|---|
Drugs . | Pre Drug . | Post Drug . | State 3 . | State 4 . | State 3/2 . | State 3/4 . | Drug (nmol/mg protein) . | ||
Glutamate/malate | |||||||||
DMSO | 10.0 | 10.4 | 64.5 | 10.8 | 6.2 | 6.0 | 0.5% vol/vol | ||
Tamoxifen | 7.8 | 13.2 | 7.2 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 8.4 | 11.0 | 8.8 | n/a | n/a | n/a | 500 | ||
Flutamide | 8.7 | 10.2 | 8.3 | n/a | n/a | n/a | 500 | ||
Diclofenac | 10.4 | 46.2 | 41.6 | n/a | n/a | n/a | 500 | ||
Quinidine | 9.0 | 12.9 | 51.7 | 9.4 | 4.0 | 5.5 | 500 | ||
Succinate | |||||||||
DMSO | 24.0 | 25.7 | 95.4 | 19.2 | 3.7 | 4.79 | 0.5% vol/vol | ||
Tamoxifen | 24.2 | 20.9 | 11.4 | n/a | n/a | n/a | 300 | ||
Fenofibrate | 22.8 | 24.1 | 36.2 | n/a | n/a | n/a | 500 | ||
Flutamide | 26.2 | 127.1 | 96.5 | n/a | n/a | n/a | 500 | ||
Diclofenac | 27.9 | 94.8 | 23.8 | n/a | n/a | n/a | 500 | ||
Quinidine | 25.4 | 31.1 | 89.8 | 24.8 | 2.9 | 3.62 | 500 |
The data generated using the fluorescence-based approach show that diclofenac increased state 2 respiration of both glutamate/malate- and succinate-driven respiration in a dose-dependent manner, while state 3 was seen to be inhibited (Fig. 9A). Fenofibrate showed complete inhibition of glutamate/malate-driven respiration when analyzed by polarimetry at 500 nmol/mg protein (Table 4). The inhibition of succinate-driven respiration was > 50% at that concentration with no changes observed in state 2 respiration. The data generated using the oxygen-sensing probes showed almost identical results (Fig. 9B). No inhibition of state 2 respiration was observed with either substrate within the first 30 min. Prolonged assay time showed that basal glutamate-driven respiration seem to decrease slightly at the two highest concentrations tested. Both succinate- and glutamate/malate-driven respiration were clearly inhibited with the latter showing greater sensitivity. The third drug analyzed, flutamide, showed complete inhibition of glutamate-driven respiration and uncoupling of succinate-driven respiration at 500 nmol/mg, using polarographic analysis (Table 4). Data generated using oxygen-sensing probes confirmed that basal glutamate/malate-driven respiration was inhibited at the highest concentration; however, with decreasing concentration this inhibition was seen to change to uncoupling (Fig. 9C).
The fourth drug analyzed, tamoxifen, completely inhibited glutamate/malate- and succinate-driven ADP-activated respiration. No effect was observed on basal respiration with either substrate (Table 4). The experiment using the oxygen-sensing probes showed an uncoupling effect on basal glutamate/malate-driven respiration, while at high tamoxifen concentrations ADP-activated respiration was seen to be inhibited (Fig. 9D). The fifth drug analyzed, quinidine, showed the least effect on mitochondrial respiration. The very slight increase in state 4 and slight decrease in state 3 resulted in somewhat lower respiratory control rations (RCRs) than the DMSO control with both glutamate/malate and succiante (Table 4). The same was observed in the oxygen-sensing probes (Fig. 9E).
Assay Volume Reduction and Application to In Vivo Studies
To use the new assay system in a screening cascade in early drug development, it needs to be amenable to a 384-well plate format, which is currently adopted by most pharmaceutical companies as standard. In this study, we successfully used the fluorescence-based oxygen consumption assay with mitochondria in 384-well plates. The results were very comparable (data not shown), and this transformation did not require any modifications to the procedure, except for the reduction of all dispensing volumes fourfold bringing the total sample volume to 50 ul (compared to 200 ul for the 96-well plates). The use of robotics and automated liquid handling may be adequate in this case, to avoid long dispensing times for the plate (normally should be kept within 10–20 min) and pipetting errors with small volumes.
The feasibility of analyzing mitochondria from multiple animals was also examined by mimicking the final part of an in vivo study during drug development. Liver mitochondria were isolated from six rats in parallel and analyzed simultaneously on a 96-well plate to assess their basal and maximum respiration with succinate and glutamate/malate as the substrate. RCRs were calculated and compared to those obtained by polarimetry (Table 5). RCRs on succinate were almost identical between the polarographic- and fluorescence-based analysis. Values for glutamate-driven respiration were slightly higher in the fluorescence-based assay. This was probably due to the fact that complex I activity diminishes over time and the polarographic analysis took several hours, whereas the fluorescence-based assay was completed within 1 h.
. | Succinate . | . | Glutamate/malate . | . | ||
---|---|---|---|---|---|---|
Animal # . | Polarimtry . | Fluorescence-based . | Polarimtry . | Fluorescence-based . | ||
1 | 3.7 | 3.9 | 5.8 | 7.4 | ||
2 | 4.0 | 4.2 | 5.6 | 6.1 | ||
3 | 3.6 | 4.3 | 5.2 | 6.6 | ||
4 | 3.8 | 5.5 | 5.4 | 7.3 | ||
5 | 3.9 | 4.0 | 5.2 | 7.1 | ||
6 | 3.9 | 4.5 | 5.7 | 7.8 | ||
Mean | 3.82 | 4.37 | 5.48 | 7.04 | ||
SD | 0.15 | 0.58 | 0.26 | 0.61 | ||
CV | 3.9 | 13.2 | 4.67 | 8.69 |
. | Succinate . | . | Glutamate/malate . | . | ||
---|---|---|---|---|---|---|
Animal # . | Polarimtry . | Fluorescence-based . | Polarimtry . | Fluorescence-based . | ||
1 | 3.7 | 3.9 | 5.8 | 7.4 | ||
2 | 4.0 | 4.2 | 5.6 | 6.1 | ||
3 | 3.6 | 4.3 | 5.2 | 6.6 | ||
4 | 3.8 | 5.5 | 5.4 | 7.3 | ||
5 | 3.9 | 4.0 | 5.2 | 7.1 | ||
6 | 3.9 | 4.5 | 5.7 | 7.8 | ||
Mean | 3.82 | 4.37 | 5.48 | 7.04 | ||
SD | 0.15 | 0.58 | 0.26 | 0.61 | ||
CV | 3.9 | 13.2 | 4.67 | 8.69 |
. | Succinate . | . | Glutamate/malate . | . | ||
---|---|---|---|---|---|---|
Animal # . | Polarimtry . | Fluorescence-based . | Polarimtry . | Fluorescence-based . | ||
1 | 3.7 | 3.9 | 5.8 | 7.4 | ||
2 | 4.0 | 4.2 | 5.6 | 6.1 | ||
3 | 3.6 | 4.3 | 5.2 | 6.6 | ||
4 | 3.8 | 5.5 | 5.4 | 7.3 | ||
5 | 3.9 | 4.0 | 5.2 | 7.1 | ||
6 | 3.9 | 4.5 | 5.7 | 7.8 | ||
Mean | 3.82 | 4.37 | 5.48 | 7.04 | ||
SD | 0.15 | 0.58 | 0.26 | 0.61 | ||
CV | 3.9 | 13.2 | 4.67 | 8.69 |
. | Succinate . | . | Glutamate/malate . | . | ||
---|---|---|---|---|---|---|
Animal # . | Polarimtry . | Fluorescence-based . | Polarimtry . | Fluorescence-based . | ||
1 | 3.7 | 3.9 | 5.8 | 7.4 | ||
2 | 4.0 | 4.2 | 5.6 | 6.1 | ||
3 | 3.6 | 4.3 | 5.2 | 6.6 | ||
4 | 3.8 | 5.5 | 5.4 | 7.3 | ||
5 | 3.9 | 4.0 | 5.2 | 7.1 | ||
6 | 3.9 | 4.5 | 5.7 | 7.8 | ||
Mean | 3.82 | 4.37 | 5.48 | 7.04 | ||
SD | 0.15 | 0.58 | 0.26 | 0.61 | ||
CV | 3.9 | 13.2 | 4.67 | 8.69 |
Parameter . | Polarographic method . | Fluorescence method . |
---|---|---|
Equipment | Special Clark-type electrode and chamber | Standard microtitre plates, standard fluorescent reader, and oxygen-sensitive probe (all commercial products) |
Stirring requirement | Yes | No |
Measurement temperature | 30°C | 30°C |
Assay volume | 0.25–3 ml | 0.2 ml (96-well plate) 0.05 ml (384-well plate) |
Protein concentration | 0.25–0.5 mg/ml | 0.063–0.25 mg/ml |
Amount of protein per assay | 0.5–1 mg | 0.01–0.05 mg |
Assay cycle time | 10–20 min | 10–60 min |
Assay sample throughput | < 5 per hour | > 96 per hour |
Determination of respiration rates and RCRs | Yes, RCR2/3 and RCR3/4 | Adequate estimate of rates and RCR2/3. Correction is required to determine absolute values. |
Effector addition | Possible (syringe) | Currently not possible |
IC50 generation | Tedious | Straight forward |
Miniaturization | Difficult | Straight forward (384-well) |
Automation | Not possible currently | Standard liquid handling and detection equipment |
Statistics | Tedious | Straight forward |
Compound interferences | Difficult to compensate. Possible poisoning of the electrode. | Most can be accounted for or overcome. Probe is stable, disposable. |
Parameter . | Polarographic method . | Fluorescence method . |
---|---|---|
Equipment | Special Clark-type electrode and chamber | Standard microtitre plates, standard fluorescent reader, and oxygen-sensitive probe (all commercial products) |
Stirring requirement | Yes | No |
Measurement temperature | 30°C | 30°C |
Assay volume | 0.25–3 ml | 0.2 ml (96-well plate) 0.05 ml (384-well plate) |
Protein concentration | 0.25–0.5 mg/ml | 0.063–0.25 mg/ml |
Amount of protein per assay | 0.5–1 mg | 0.01–0.05 mg |
Assay cycle time | 10–20 min | 10–60 min |
Assay sample throughput | < 5 per hour | > 96 per hour |
Determination of respiration rates and RCRs | Yes, RCR2/3 and RCR3/4 | Adequate estimate of rates and RCR2/3. Correction is required to determine absolute values. |
Effector addition | Possible (syringe) | Currently not possible |
IC50 generation | Tedious | Straight forward |
Miniaturization | Difficult | Straight forward (384-well) |
Automation | Not possible currently | Standard liquid handling and detection equipment |
Statistics | Tedious | Straight forward |
Compound interferences | Difficult to compensate. Possible poisoning of the electrode. | Most can be accounted for or overcome. Probe is stable, disposable. |
Parameter . | Polarographic method . | Fluorescence method . |
---|---|---|
Equipment | Special Clark-type electrode and chamber | Standard microtitre plates, standard fluorescent reader, and oxygen-sensitive probe (all commercial products) |
Stirring requirement | Yes | No |
Measurement temperature | 30°C | 30°C |
Assay volume | 0.25–3 ml | 0.2 ml (96-well plate) 0.05 ml (384-well plate) |
Protein concentration | 0.25–0.5 mg/ml | 0.063–0.25 mg/ml |
Amount of protein per assay | 0.5–1 mg | 0.01–0.05 mg |
Assay cycle time | 10–20 min | 10–60 min |
Assay sample throughput | < 5 per hour | > 96 per hour |
Determination of respiration rates and RCRs | Yes, RCR2/3 and RCR3/4 | Adequate estimate of rates and RCR2/3. Correction is required to determine absolute values. |
Effector addition | Possible (syringe) | Currently not possible |
IC50 generation | Tedious | Straight forward |
Miniaturization | Difficult | Straight forward (384-well) |
Automation | Not possible currently | Standard liquid handling and detection equipment |
Statistics | Tedious | Straight forward |
Compound interferences | Difficult to compensate. Possible poisoning of the electrode. | Most can be accounted for or overcome. Probe is stable, disposable. |
Parameter . | Polarographic method . | Fluorescence method . |
---|---|---|
Equipment | Special Clark-type electrode and chamber | Standard microtitre plates, standard fluorescent reader, and oxygen-sensitive probe (all commercial products) |
Stirring requirement | Yes | No |
Measurement temperature | 30°C | 30°C |
Assay volume | 0.25–3 ml | 0.2 ml (96-well plate) 0.05 ml (384-well plate) |
Protein concentration | 0.25–0.5 mg/ml | 0.063–0.25 mg/ml |
Amount of protein per assay | 0.5–1 mg | 0.01–0.05 mg |
Assay cycle time | 10–20 min | 10–60 min |
Assay sample throughput | < 5 per hour | > 96 per hour |
Determination of respiration rates and RCRs | Yes, RCR2/3 and RCR3/4 | Adequate estimate of rates and RCR2/3. Correction is required to determine absolute values. |
Effector addition | Possible (syringe) | Currently not possible |
IC50 generation | Tedious | Straight forward |
Miniaturization | Difficult | Straight forward (384-well) |
Automation | Not possible currently | Standard liquid handling and detection equipment |
Statistics | Tedious | Straight forward |
Compound interferences | Difficult to compensate. Possible poisoning of the electrode. | Most can be accounted for or overcome. Probe is stable, disposable. |
DISCUSSION
Methods of assessing the potential of an NCE to induce mitochondrial dysfunction are currently required within drug development if mitochondrial toxicity–induced drug attrition or drug withdrawal is to be avoided. While oxygen has long been acknowledged as a key indicator of mitochondrial function, analysis of this parameter in the assessment of mitochondrial toxicity has been hampered by the limited throughput of traditional polarographic techniques (Marroquin et al., 2005). In this study, we present a fluorescence-based high throughput alternative to polarographic assessment of mitochondrial oxygen consumption and examine how such an approach would be best employed in the assessment of drug-induced mitochondrial toxicity.
Measurements are carried out in standard microtitre plates using recently developed dispensable phosphorescent oxygen-sensitive probes with simultaneous monitoring of multiple samples on a conventional fluorescence plate reader, preferably in time-resolved fluorescence mode. Although the absolute rates of oxygen consumption are lower when assessed using the fluorescence technique (Table 1), the effects of substrate and ADP addition are in accord between the two techniques, as reflected by the concordance between the ratios. Classical mitochondrial assessments entail analysis of O2 consumption after addition of substrates (state 2), rate upon ADP addition (state 3), and final return to basal state after all ADP has been phosphorylated (state 4). Classical ratiometric analysis provides insight, with the RCR of state 3/state 4 reflecting coupling condition and phosphorylation efficiency (Estabrook, 1967). The power of polarography lies in its ability to provide paired comparisons of these states, although throughput is severely constrained. Sequential additions are technically difficult in the fluorescence protocol, so that state 2 (respiration with substrate but no exogenous ADP) is substituted for state 4 which yields a nontraditional ratiometric index. However, the validity of the fluorescence system is not unduly undermined by using state 2 in this calculation, as shown by concordance between ratios in Table 1, while throughput is increased by orders of magnitude. In practice, compound effects are expressed as a percentage of controls on the same plate rather than RCR. The excellent reproducibility of the fluorescence-based method (with CV values of less than 3% being typical) also allows low numbers of replicates to be used during analysis and can facilitate single point measurements.
The method was shown to be applicable to analysis of mitochondria derived from different tissues, with cardiac muscle, skeletal muscle, and liver mitochondria all being successfully analyzed. This ability to analyze different tissues can be important in the assessment of drug-induced toxicity as some drugs can exhibit tissue specificity both in vitro and in vivo. For example, in vitro experiments with the chemotherapeutic doxorubicin and certain antivirals are generally carried out using heart mitochondria (Lund and Wallace, 2004; Wallace, 2003), whereas diagnosis of certain mitochondrial genetic diseases is carried out with mitochondria isolated from muscle biopsies (Taylor et al., 2004). We have shown that mitochondria from both these tissues can be reliably analyzed using the fluorescence-based approach and that very small amounts of biomaterial are required, particularly for analysis in state 3, as these mitochondrial preparations are particularity active. This is beneficial as both heart and muscle tissues are usually available in much smaller quantities and yield significantly less mitochondria than liver tissue.
To achieve the convenience of measurement in standard microtitre plates, a level of sensitivity is sacrificed. Due to the incomplete sample sealing in the microplate formats and the associated back diffusion of atmospheric oxygen, the rates of change of dissolved oxygen measured using this approach are lower than the actual consumption rates measured using a sealed polarograhic system. Better agreement between the two methods can potentially be achieved, by applying more complex modeling of the fluorescent system and processing of data generated. This work, which is outside the scope of this study, is already underway. However, the observation that both fully and partially sealed systems produce the same general trends allows the open fluorescence system to be used to assess the effects of various treatments on ETC function.
The assessment of classical inhibitors and uncouplers of oxidative phosphorylation illustrates that the fluorescence method may be used to predict drug toxicity and contribute to the elucidation of the mechanisms involved. In addition, the state of mitochondrial activation (state 2 or 3) and concentration of the NCE used can have a considerable effect on such an analysis. For example, one can easily screen for inhibition of ATP synthesis using ADP-activated mitochondria and an NCE at a single concentration in triplicate. However, to test for inhibitors of complex I and complex II, both glutamate-malate- and succinate-driven respiratory activity must be analyzed. Furthermore, compounds with an uncoupling mechanism similar to FCCP may be missed (i.e., produce false negatives) or misinterpreted as having inhibitory, rather than uncoupling activity, if tested only at a single concentration. For example, at low concentrations FCCP functions as a specific proton ionophore, uncoupling both the pH and electrical components of the mitochondrial membrane potential. However, at higher concentrations, FCCP and other uncouplers such as lipophilic fatty acids, physically disrupt membrane integrity with consequent loss of electron transfer efficiency (Fig. 7). Likewise, compounds with mechanisms similar to oligomycin (inhibition of F1/FO-ATPase) may also be missed if only state 2 respiration is analyzed. For mechanistic studies, it is therefore advisable to test drugs at multiple concentrations on both basal and ADP-activated succinate- and glutamate/malate-driven respiration, a matrix for which polarography is not well suited, compared to the fluorescence technique described here.
Fluorescent respiration profiles generated by this method (i.e., unprocessed raw data) provide semiquantitative information on degree of inhibition or uncoupling, while conversion to concentration of dissolved oxygen allows the generation of IC50 values. Such values, although of limited value from a mechanistic standpoint, are highly useful in that they allow the ranking of compounds in accordance with calculated toxicity potency. This value may then be related to in vivo potency with respect to a specific target (i.e., safety margin). Compounds may also be compared to known toxins for further characterization.
Some compounds may interfere with the oxygen probe employed here, affecting fluorescence signals detected from the sample, e.g., due to quenching, autofluorescence, or optical effects (O‘Donovan et al., 2005). To account for such cases, a brief inspection of raw fluorescence data (i.e., simple visual check of initial intensities and shapes of respiration curves) is recommended, as it allows identification of possible interferences. The latter may be overcome either through appropriate data analysis of fluorescence data or through the use of a dispensable but particle-based probe (O’Donovan et al., 2005). Blind conversion of fluorescence profiles into oxygen profiles may occasionally lead to considerable errors in calculated oxygen consumption rates and IC50. Oxygen electrodes are also susceptible to chemical interferences and poisoning, and these effects are often difficult to both detect and to compensate for. In contrast to electrode-based analysis which can become unreliable at low oxygen concentrations, the fluorescence-based oxygen-sensitive probes are efficient over the entire physiological oxygen range, particularly at low oxygen concentrations (Papkovsky, 2004).
The inherent advantages of the fluorescence approach are particularly evident from the analysis of the effects of drug treatment. The results obtained with fenofibrate provide additional information on how fenofibrate inhibits mitochondrial function. Our results showing glutamate/malate-driven respiration to be more sensitive support previous work (Brunmair et al., 2004) demonstrating that fenobibrate inhibits complex I activity. These researchers however observed no effect on succinate-driven respiration, an observation which is at odds both with our observations and previous reports (Zhou and Wallace, 1999). Since fenofibrate is seen to inhibit both, succinate- and glutamate-driven respiration, the mechanism involved must be more than a standard inhibition of complex I. We tested fenofibrate for the ability to inhibit complex V, but did not see any inhibition (data not shown) illustrating that further work will be required to fully understand the mechanism involved. The effects we report for diclofenac also concur with previous reports (Moreno-Sanchez et al., 1999) where membrane potential collapse was observed at very high treatment concentrations (> 300 nmol/mg), with inhibition observed at lower concentrations.
The effects of flutamide on mitochondrial function have not been examined in detail to date. We found flutamide to be a very potent mitochondrial toxin, an observation supported by previous reports (Fau et al., 1994). We have recently identified flutamide to be a complex I inhibitor with an approximate IC50 of 76uM (Marroquin et al., 2005), and the data presented here add an additional level of detail. These data illustrate that basal glutamate/malate-driven respiration is inhibited at high concentrations; however, this mechanism is seen to change from inhibition to uncoupling as the treatment concentration decreases.
Tamoxifen has been reported to be a complex III inhibitor and has also been shown to inhibit cytochrome oxidase, an inhibition which is never observed to exceeded 60%. We recently also identified tamoxifen to inhibit complex V with an approximate IC50 of 3uM (Marroquin et al., 2005). It was furthermore reported that tamoxifen could dissipate mitochondrial membrane potential. Polarometric analysis revealed that 500 nmol/mg protein completely inhibited glutamate/malate/ADP-driven respiration as well as ADP-driven respiration with succinate. No effect was seen on basal respiration with either substrates. Fluorescence-based analysis added additional information, in that it showed that ADP-stimulated glutamate/malate-driven respiration was more sensitive to inhibition than ADP-stimulated respiration with succinate suggesting that tamoxifen might even have more targets than previously reported.
Quinidine has been shown to inhibit mitochondrial ATPase but in the millimolar range, while our exposure concentrations were in the micromolar range. Therefore, it was not surprising to see very little effect on mitochondrial respiration. Tissue levels for quinidine must get very high for mitochondrial-derived toxicity to occur.
Overall, the fluorescence-based platform has allowed us to quickly characterize the effect of drugs on glutamate/malate- and succinate-driven basal and ADP-activated respiration. We found that all data correlated with published results and that we were able to generate additional data which contribute to a more thorough understanding of mechanisms involved for several toxicants. The central observation is that a more detailed picture of the effect of drug treatment may be generated by analyzing both substrates over a range of drug concentrations and that this type of analysis can be done rapidly. Using this approach, hundreds of NCEs per day can be screened in triplicate at a single concentration, while IC50 values may be generated for approximately 25 compounds/day. Within the lead optimization paradigm this throughput is adequate. However, some programs may require more rigorous toxicity screening and hence the feasibility of running this assay in a 384-well format was also examined and proved successful. In this case, the use of standard HTS robotics and liquid handling equipment is advisable to reduce plate-processing time and avoid pipetting errors.
Another area where such analysis could provide particularly relevant information is in the testing of the function of mitochondria isolated from a series of treated animals. By analyzing state 2 and state 3 with different substrates, one can get a very good indication as to whether organ-specific mitochondrial function is impaired. This type of safety analysis could, for example, be carried out for discovery programs with possible liabilities, such as HIV or diabetes. For this approach to be viable, it is necessary to be able to analyze mitochondria isolated from multiple control animals (untreated) in parallel with high reproducibility, if subtle differences between treated and untreated animals are to be differentiated above assay noise. Data of this quality were generated during this study, with analysis of mitochondria isolated from six rats' analyses in basal and ADP-activated glutamate/malate- and succinate-driven respiration giving highly reproducible results. In instances where the amount of biomaterial is extremely limited such as in vivo studies examining mitochondrial dysfunction of soleus muscle, more advanced high-sensitivity, low-volume platforms may be used which employ the same oxygen probes and fluorescence-detection methodology (O'Mahony et al., 2005). Further validation of this approach is envisaged using animals treated with model compounds such as acetaminophen.
Overall, the fluorescence-based oxygen consumption assay is seen to be particularly suited to high-throughput analysis of the potential of NCEs to cause mitochondrial dysfunction. The approach can provide detailed and specific information about possible mechanisms of toxicity based on measurements of basal (state 2) and ADP-activated (state 3) respiration. To perform respirometric analysis in standard microplates, a mineral oil seal is required, thereby preventing subsequent reagent addition and requiring different respiration states to be analyzed separately. A comparison of the main performance characteristics of the fluorescence mitochondrial oxygen consumption assay with the established polarographic assay is given in Table 6.
The advantages of the fluorescence method are the simplicity of the measurement procedure, the high-sample throughput and low bulk requirements. Results can be visualized in real time and analyzed either semiquantitatively using standard plate reader software or processed to oxygen concentrations for quantitative analysis. The method therefore demonstrated the capacity to accelerate the development of new drugs and medicines and to improve drug attrition rates due to mitochondrial toxicity.
These authors contributed equally to the work presented.
References
Almotrefi, A. A. (
Amacher, D. E. (
Boelsterli, U. (
Brand, M. D., Buckingham, J. A., Esteves, T. C., Green, K., Lambert, A. J., Miwa, S., Murphy, M. P., Pakay, J. L., Talbot, D. A., and Echtay K. S., (
Brunmair, B., Lest, A., Staniek, K., Gras, F., Scharf, N., Roden, M., Nohl, H., Waldhausl, W., and Furnsinn, C. (
Cardoso, C. M. P., Custodio, J. B. A., Almeida, L. M., and Moreno, A. J. M. (
Cunarro, J., and Weiner, M. W. (
Estabrook, R. W. (
Fau, D., Eugene, D., Berson, A., Letteron, P., Fromenty, B., Fisch, C., and Pessayre, D. (
Haugland, R. P. (
Heisler, C. R. (
Horgan, D. J., Ohno, H., Singer, T. P., and Casida, J. E. (
Hynes, J., Floyd, S., Soini, A. E., O'Connor, R., and Papkovsky, D. B. (
Isom, G. E., and Way, J. L. (
Jaeschke, H., Gores, G. J., Cederbaum, A. I., Hinson, J. A., Pessayre, D., and Lemasters, J. J. (
Kolesova, G. M., Vishnivetskii, S. A., and Iaguzhinskii, L. S. (
Lapidus, R. G., and Sokolove, P. M. (
Lewis, W., Day, B. J., and Copeland, W. C. (
Li, N., Ragheb, K., Lawler, G., Sturgis, J., Rajwa, B., Melendez, J. A., and Robinson, J. P. (
Lund, K. C., and Wallace, K. B. (
Marroquin, L., Stevens, G., and Will, Y. (
Mattson, M. P., and Kroemer, G. (
Messer, J. I., Jackman, M. R., and Willis, W. T. (
Mironova, G. D., Grigoriev, S. M., Skarga, Y., Negoda, A. E., and Kolomytkin, O. V. (
Moreno-Sanchez, R., Bravo, C., Vasquez, C., Ayala, G., Silveira, L. H., and Martinez-Lavin, M. (
Moyle, G. (
O'Donovan, C., Hynes, J., Yashunski, D., and Papkovsky, D. B. (
O‘Mahony, F. C., O’Donovan, C., Hynes, J., Moore, T., Davenport, J., and Papkovsky, D. B. (
Papkovsky, D. B. (
Scaduto, R. C. J., and Grotyohann, L. W. (
Schagger, H., Brandt, U., Gencic, S., and von Jagow, G. (
Sullivan, P. G., and Brown, M. R. (
Taylor, R. W., Schaefer, A. M., Barron, M. J., McFarland, R., and Turnbull, D. M. (
Walker, J. E., Collinson, I. R., Van Raaij, M. J., and Runswick, M. J. (
Wallace, K. B. (
Waterhouse, N. J., Ricci, J.-E., and Green, D. R. (
Wodnicka, M., Guarino, R. D., Hemperly, J. J., Timmins, M. R., Stitt, D., and Pitner, J. B. (
Author notes
*Luxcel Biosciences Ltd., G.17, Lee Maltings, Cork, Ireland; and †Pfizer Global R&D, Safety Sciences, 10646 Science Centre Drive, San Diego, California 92121
Comments