Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 29.
Published in final edited form as: J Huntingtons Dis. 2022;11(2):141–151. doi: 10.3233/JHD-220539

XJB-5-131 Is a Mild Uncoupler of Oxidative Phosphorylation

Zhiyin Xun a, Peter Wipf b, Cynthia T McMurray a,*
PMCID: PMC9798833  NIHMSID: NIHMS1856716  PMID: 35404288

Abstract

Background:

Mitochondria (MT) are energy “powerhouses” of the cell and the decline in their function from oxidative damage is strongly correlated in many diseases. To suppress oxygen damage, we have developed and applied XJB-5-131 as a targeted platform for neutralizing reactive oxygen species (ROS) directly in MT. Although the beneficial activity of XJB-5-131 is well documented, the mechanism of its protective effects is not yet fully understood.

Objective:

Here, we elucidate the mechanism of protection for XJB-5-131, a mitochondrial targeted antioxidant and electron scavenger.

Methods:

The Seahorse Flux Analyzer was used to probe the respiratory states of isolated mouse brain mitochondria treated with XJB-5-131 compared to controls.

Results:

Surprisingly, there is no direct impact of XJB-5-131 radical scavenger on the electron flow through the electron transport chain. Rather, XJB-5-131 is a mild uncoupler of oxidative phosphorylation. The nitroxide moiety in XJB-5-131 acts as a superoxide dismutase mimic, which both extracts or donates electrons during redox reactions. The electron scavenging activity of XJB-5-131 prevents the leakage of electrons and reduces formation of superoxide anion, thereby reducing ROS.

Conclusion:

We show here that XJB-5-131 is a mild uncoupler of oxidative phosphorylation in MT. The mild uncoupling property of XJB-5-131 arises from its redox properties, which exert a protective effect by reducing ROS-induced damage without sacrificing energy production. Because mitochondrial decline is a common and central feature of toxicity, the favorable properties of XJB-5-131 are likely to be useful in treating Huntington’s disease and a wide spectrum of neurodegenerative diseases for which oxidative damage is a key component. The mild uncoupling properties of XJB-5-131 suggest a valuable mechanism of action for the design of clinically effective antioxidants.

Keywords: Antioxidant, metabolism, mitochondrial, oxidative stress, reactive oxygen species

INTRODUCTION

Mitochondria (MT) are the primary intracellular sources of reactive oxygen species (ROS) and the main target for oxidative damage [1-3]. Oxidative stress and defective MT have been implicated in various pathological conditions such as Alzheimer’s disease [4, 5], Huntington’s disease (HD) [6-8], and Parkinson’s disease [9], among other neurodegenerative disorders [10, 11] and diseases arising from excessive oxidation [12-15]. While other naturally occurring generic antioxidants correct for some of the deleterious effects of ROS, they often lack specificity [16, 17] and their efficacy is marginal or ineffective in human clinical trials [18, 19]. Untargeted antioxidants can also negatively interfere in cytosolic ROS signaling cascades and in the activation of host defenses [20-23]. Thus, there has been considerable interest in developing MT-targeted ROS scavengers as more effective therapeutic agents [24-30].

XJB-5-131 is a novel bi-functional synthetic conjugate of a nitroxide free radical species. Its mitochondrial targeting moiety mimics the membrane lipid affinity of the antibiotic gramicidin S [25, 27, 31-33]. XJB-5-131 is beneficial in a variety of pathologies associated with increased oxidative damage, in addition to HD [29]. For instance, XJB-5-131 suppresses reperfusion injury [34], mitigates damage to brain tissue during microdialysis [35], prevents cardiolipin oxidation and reduces neuronal death in traumatic brain injury [33, 36], and mitigates age-related disk degeneration [37] and whole-body irradiation [38], among others. Over the years, we have reported that XJB-5-131 significantly ameliorates disease features in mouse models of HD [39-41]. Collectively, a wealth of in vitro and in vivo studies indicates that XJB-5-131 holds great promise for the development of novel therapeutics for disorders involving oxidative stress [29].

Despite these significant beneficial properties, the mechanism of action of XJB-5-131 remains speculative. XJB-5-131 is capable of a diverse range of redox processes [29]. ROS are highly reactive transient species and are converted to hydrogen peroxide by superoxide dismutase [42]. In this single-electron transfer process, the oxidation state of the metal cation in superoxide dismutase increases from n to n + 1 [42]. The nitroxide moiety in XJB-5-131 can serve as a superoxide dismutase-mimic and participates in a similar reduction-oxidation (redox) mechanism (Fig. 1a). This electron scavenging activity prevents superoxide radical anion formation, neutralizes ROS when it is formed, and prevents the occurrence of organic peroxides and mtDNA damage. Additionally, in other redox states, a hydroxylamine state can reduce a radical species by hydrogen atom donation, and the oxoammonium cation state can participate in a two-electron transfer reaction [42]. Thus, there are multiple mechanisms that could contribute to its protective effects in MT, but the mechanism is unknown.

Fig. 1.

Fig. 1.

Open in a new tab

XJB-5-131 reduces mitochondrial reactive oxygen species (ROS). A) XJB-5-131 undergoes redox equilibration among states. The XJB-5-131 nitroxide contains a nitroxyl group with an unpaired electron. The structures in Fig. 1 indicate redox cycling states of the XJB-5-131 nitroxide. XJB-5-131 is composed of a stable nitroxide radical and an alkene peptide isostere modification of the Leu-d-Phe-Pro-Val-Orn segment of the cyclopeptide antibiotic gramicidin S. The nitroxide radical (i) can equilibrate with the oxoammonium cation (ii) and hydroxylamine (iii) states. B) Chemistry of an ROS inducer, 2,3-dimethoxy-1,4-naphthoquinone (DMNQ). C) XJB-5-131 reduces the mitochondrial H2O2 level under basal conditions and upon exposure to an ROS inducer, DMNQ. Data are mean ± SEM (n = 3). *p <0.05, **p < 0.01 using the unpaired two-tailed Student’s t-test.

Here we investigate the mode of action of XJB-5-131, and the mechanisms by which it exerts its beneficial effects in MT. We provide evidence that XJB-5-131 significantly reduces MT H2O2 (peroxide) level under both basal conditions and upon exposure to a ROS-inducer, 2,3-dimethoxy-1,4-naphthoquinone (DMNQ) [43, 44]. Furthermore, we did not detect a change in mitochondrial electron flow. Rather, XJB-5-131 mildly uncouples oxidative phosphorylation to increase the oxygen consumption rate and reduces cellular ROS without affecting ATP production. We favor a model in which a significant antioxidant property of XJB-5-131 is independent from the direct ROS scavenging and organic radical neutralization effects of its nitroxide substructure.

MATERIALS AND METHODS

Animals

All procedures involving animals were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Protocols were approved by the Lawrence Berkeley National Laboratories Animal Welfare and Research Committee. All animal work was conducted according to national and international guidelines. All MT were isolated from C57Bl6J mice (Jackson Labs, Bar Harbor, ME, USA). Unless otherwise indicated, the MT were isolated between animals of 50–60 weeks, most often 57-week-old animals.

Crude mitochondrial isolation

MT from mouse forebrains were rapidly transferred into ice-cold mannitol sucrose (MS) buffer (225 mmol/L mannitol, 75 mmol/L sucrose, 5 mmol/L HEPES, 1 mmol/L EGTA, 0.1% fatty acid free BSA, pH 7.4) and homogenized. MT were isolated from 57-week-old C57Bl6J mice. The homogenates were centrifuged at 1,300×g for 3 min at 4°C. The supernatants were then centrifuged for 10 min at 21,000×g at 4°C and the resulting pellets containing crude MT were used for further assays. Mitochondrial protein concentration was determined using the Bio-Rad Bradford assay (BioRad Laboratories).

Mitochondrial oxygen consumption rate (OCR) measurements

The respiration of isolated MT was measured to investigate the electron flow through the electron transport chain (ETC) [45]. Mitochondrial OCR was measured with a Seahorse XF96 extracellular flux analyzer (Seahorse Bioscience, Billerica, MA) [6, 39-41]. Mitochondrial pellets were resuspended in the mitochondrial assay solution (MAS), 75 mM sucrose, 225 mM mannitol, 10 mM KH2PO4, 5 mM MgCl2, 5 mM HEPES, 1 mM EGTA, and 0.2% (w/v) fatty acid-free BSA, pH 7.2 at 37°C). 140 μL suspensions of MT (2 μg protein/well) were plated on a pre-chilled Seahorse PS 96-well microplate (Seahorse Bioscience, Billerica, MA). The plate was centrifuged at 3,220×g for 20 min at 4°C. Subsequently, the plate was incubated in 37°C (without CO2) for 8–10 min and transferred to the XF flux analyzer to investigate sequential electron flow through different complexes of the ETC. The order of additions for substrates in the assay protocol was 1) Rotenone (complex I (CI) inhibitor) (4 μM), 2) succinate (complex II (CII) substrate) (50 mM), and 3) Antimycin A (complex III (CIII) inhibitor) (2 μg/mL) and a mixture of Ascorbate (10 mM) and N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) (100 μM) as substrates for complex IV (CIV). Each datapoint is the sum of 15 measurements and expressed as mean ± standard error.

A second set of experiments was designed to investigate mitochondrial coupling status [46]. Specifically, mitochondrial respiration (of 2 μg MT/well) was measured in the order of a) coupled state with substrate present. MT were held in state II (10 mM succinate and 2 μM rotenone, inorganic phosphates in basal respiration buffer) (State II basal respiration), b) State III was initiated by the addition of ADP (phosphorylating respiration, in the presence of ADP and substrate), c) State IVo was induced with the addition of oligomycin, and d) State IIIu (maximal uncoupler-stimulated respiration) was induced with the addition of FCCP (4 μM). Each datapoint is the sum of 15 measurements and expressed as mean ± standard error.

XJB-5-131 synthesis and in vitro treatment

XJB-5-131 was synthesized as described previously [25, 27]. XJB-5-131 (0.2, 1, and 10 μM) was incubated with isolated brain MT for 40 min and 10 min prior to the mitochondrial OCR experiment and the peroxide assay, respectively.

Measurement of mitochondrial ROS (H2O2)

Mitochondrial H2O2 (peroxide) was determined by using the Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Briefly, the isolated brain mitochondrial pellet was suspended in the MAS buffer (MAS), 75 mM sucrose, 225 mM mannitol, 10 mM KH2PO4, 5mM MgCl2, 5mM HEPES, 1 mM EGTA, and 0.2% (w/v) fatty acid-free BSA, pH 7.2 at 37°C). Various concentrations of XJB-5-131 or the ROS-inducer DMNQ was added to incubate for 10 min. Amplex Red Hydrogen Peroxide working solution containing 10 μM Amplex Red and 10 U/mL hydrogen peroxidase was added 10 min prior to the reading of the fluorescent intensity on a TECAN plate reader with the excitation and emission wavelength of 560 nm and 590 nm, respectively.

Statistical analysis

Values were expressed as mean ± standard error of the mean (SEM), unless otherwise stated. p-values were obtained from the unpaired two-tailed Student’s t-test.

RESULTS

XJB-5-131 reduces ROS in MT

XJB-5-131 is a bi-functional conjugate that comprises a spin label, 4-amino-2,2,6,6-tetramethyl piperidine-1-oxyl, conjugated to a mitochondrial targeted alkene peptide isostere (Fig. 1a) [25, 27]. The targeting portion of the molecule uses the β-turn moiety of the Leu-d-Phe-Pro-Val-Orn segment of the cyclopeptide antibiotic gramicidin S, which allows it to localize to the mitochondrial membrane. The nitroxide (i) can participate in single- and double-electron/proton exchange equilibria involving an oxoammonium cation (ii) and a hydroxylamine (iii) (Fig. 1a), thus responding to the redox environment and contributing to redox homeostasis [42].

To test whether XJB-5-131 acted mainly as an ROS scavenger in cells, we isolated MT from 57-week mouse brains and measured the formation of H2O2 after treatment with DMNQ in the presence or absence of XJB-5-131 (Fig. 1b, c). DMNQ is a non-thiol-capturing and non-alkylating redox cycling quinone that causes continuous intracellular generation of H2O2 and subsequent oxidative damage to MT [43, 44]. Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine) reacts in a 1:1 ratio with peroxide and its level is determined by the fluorescent intensity at 571 nm. If XJB-5-131 eliminates superoxide radical anion (O2•−) formation, then the compound should reduce the peroxide concentration. Indeed, the quantified amount of H2O2 was significantly lower at all three concentrations of XJB-5-131 (i.e., 0.2, 1, and 10 μM, in the presence of 0 μM, 1 μM, or 10 μM DMNQ) relative to the untreated control (Fig. 1c). At 10 μM DMNQ, the peroxide reduction was still measurable but not statistically significant. It is possible that at high concentration of the redox cycling quinone, a large amount of superoxide radical anion is converted to hydrogen peroxide and overwhelms the scavenging properties of the MT localized XJB-5-131. In fact, nitroxides are known to be more effective in protecting against superoxide radical anion, O2•−, relative to hydrogen peroxide, H2O2. The decrease in Amplex Red reported H2O2 after XJB-5-131 treatment confirmed the mild ROS scavenging properties of XJB-5-131.

XJB-5-131 does not affect mitochondrial electron flow

Nitroxides can prevent the formation of ROS, particularly superoxide, due to electron capture and conversion to the hydroxylamine state after proton transfer (Fig. 1a) [42]. To test whether XJB-5-131 had a similar mechanism, we pretreated MT from 57-week-old brains of C57Bl6J animals with XJB-5-131 for 40 min, and sequentially added substrates or inhibitors of ETC complexes, as previously reported (Fig. 2a) [39-41]. Inhibition of CI with rotenone attenuated electron transport, which was restored by addition of succinate (a CII substrate) (Fig. 3a). Inhibition of CIII with antimycin A stopped the electron flow at Complex III and reduced OCR, which was rescued by co-treatment with an artificial electron donor, N1, N1, N1, N1-tetramethyl-1,4-phenylene diamine (TMPD/ascorbate). TMPD is an electron donor for CIV, and ascorbate (Asc) keeps TMPD in its reduced state [46] (Fig. 3a). Pre-treatment with XJB-5-131 had no measurable impact on substrate-dependent electron transport along the ETC (Fig. 3a).

Fig. 2.

Fig. 2.

Open in a new tab

Schematic diagram of substrates and inhibitors of the electron transport chain. A) Schematic diagram of substrates and inhibitors for Complex I, II, III, and IV of the mitochondrial electron transport chain (ETC) (electron transport image, Ilya Belevich, University of Helsinki). Pyruvate, succinate (suc), duroquil, ascorbate (Asc) and N1, N1, N1, N1-tetramethyl-1,4-phenylene diamine (TMPD), and ADP serve as substrates for complex I-V, respectively. Rotenone (Rot), malonate (mal), antimycin A (AA), inhibit complex I, II, and III, respectively. Oligomycin inhibits ATP synthase (complex V). Treatment with Inhibitors decreases OCR while treatment with substrate increases OCR. B) Theoretical schematic diagram of mitochondrial respiratory states that represents oxygen consumption in coupled and uncoupled states. Diagram used with permission from Noel Sturm 2020; https://www2.csudh.edu/nsturm/CHE450/25ElectronTransport.htm. The plot in Panel B tracks mitochondrial coupling status. Specifically, mitochondrial respiration is measured in the order of A) coupled state with substrate present and B) phosphorylating respiration, in the presence of substrate. MT are held in state II (10mM succinate and 2 μM rotenone, inorganic phosphates and NADH in basal respiration buffer) (State II basal respiration). Addition of rotenone at this step ensures no backflow to complex I. As much as 30% of the basal metabolic rate arises from proton leak but little oxygen is consumed since there is no ADP added and oxidative phosphorylation does not occur. B) State III is initiated by the addition of ADP (phosphorylating respiration, in the presence of substrate) in state II. Addition of ADP initiates state III, the ADP stimulated respiration of isolated coupled mitochondria. Oxygen consumption slows as ADP is depleted and ATP rises. Sate IV ends as ATP is high and the residual oxygen is consumed.

Fig. 3.

Fig. 3.

Open in a new tab

XJB-5-131 does not affect mitochondrial electron flow. A) MT in assay buffer were pre-incubated with various concentrations of XJB-5-131, and OCR from electron flow was probed using substrates and inhibitors. MT were resuspended in assay buffer: 70 mM sucrose, 200 mM mannitol, 10mM KH2PO4, 2 mM HEPES, pH 7.2, 5 mM MgCl2, 1 mM EGTA, 0.2% BSA (FA free), 2 mM L-malate, 4 mM ADP, 10 mM sodium pyruvate (10 μm). Arrows indicate the added substrate or inhibitor as follows: (Rot) rotenone (2 μM final); (Suc) succinate (10mM final); (AA) antimycin A (4 mM final); ascorbate and N,N,N9,N9-tetramethyl-p-phenylenediamine (TMPD) (10 mM and 100 μm, respectively). There were no significant effects of XJB-5-131 on electron flow through the ETC. Each datapoint was derived from 15 independent measurements of OCR acquired in the same buffer on the same day in the same instrument and the values were expressed as mean ± standard error (SEM). B) At the start of the experiment, MT were resuspended in assay buffer (70 mM sucrose, 200 mM mannitol, 10 mM KH2PO4, 2 mM HEPES, pH 7.2, 5mM MgCl2, 1 mM EGTA, 0.2% BSA (FA free), 2 mM L-malate, 4 mM ADP, 10 mM sodium pyruvate (10 μm)) with or without XJB-5-131 at the indicated concentrations. ETC substrates were added as indicated by the arrows, except that TMPD/Asc was either added or left out at the last injection step. TMPD/Asc added to untreated MT was an effective substrate for electron flow to CIV, indicating that it was active, but XJB-5-31 alone had no capacity to transfer electrons. Each datapoint was derived from 15 independent measurements of OCR acquired in the same buffer on the same day in the same instrument and the values were expressed as mean ± standard error (SEM). p-values were obtained from the unpaired two-tailed Student’s t-test for the significance of the OCR difference between XJB-5-131 treated and TMPD+XJB-5-131 treated MT. ****p < 0.0001. C) Same as (B) except that TMPD/Asc was added alone or together with indicated concentrations of XJB-5-131 at the last injection. TMPD/Asc was able to stimulate oxygen consumption at CIV whether or not XJB-5-131 was present. D) Same as in (C) except that TMPD was added without Asc. Asc ensures that the TMPD is reduced and continues to donate electrons. Since Asc was absent, CIV was stimulated to a lesser degree than for TMPD/Asc (compare to C).

It was possible that XJB-5-131 itself promoted electron flow and masked apparent substrate effects on electron transport. Thus, we tested the impact of XJB-5-131 electron transport to CIV (Fig. 3b). At the start of the experiment, MT were resuspended in assay buffer (70 mM sucrose, 200 mM mannitol, 10 mM KH2PO4, 2mM HEPES, pH 7.2, 5 mM MgCl2, 1 mM EGTA, 0.2% BSA (FA free), 2 mM L-malate, 4 mM ADP, 10 mM sodium pyruvate (10 μm)) without XJB-5-131 or with XJB-5-131 at the indicated concentrations. ETC substrates for CI-III were added as in (a), indicated by the arrows, and TMPD/Asc (substrate for CIV) was either added or left out at the last injection. OCR increased after the addition of TMPD/Asc, as expected, but in its absence, XJB-5-131 alone was not capable of passing electrons to complex IV and did not increase OCR.

As a complement, in a third experiment, substrates and inhibitors of CIII were sequentially added to MT in the assay buffer as in (Fig. 3a) in the absence of XJB-5-131. At the last injection, TMPD/Asc alone or TMPD/Asc and XJB-5-131 of various concentrations were added (Fig. 3c). Electron flow did not require pre-treatment with XJB-5-131 and OCR from substrate-driven CIV occurred, whether or not XJB-5-131 was present. The experiment in Fig. 3c was repeated for TMPD alone (Fig. 3d). Ascorbate (Asc) keeps TMPD in its reduced state to transport electrons to CIV [46] (Fig. 3a). OCR was lower than TMPD in the absence of Asc (Fig. 3d). Collectively, XJB-5-131 had no obvious impact on electron transfer along the ETC.

XJB-5-131 induces mild-uncoupling of MT

ETC uncoupling tends to decrease mitochondrial ROS production and increase oxygen consumption [47-50]. Since it did not affect electron flow from CI to CIV (Fig. 3), we examined whether XJB-5-131 had uncoupling activity using a Seahorse assay [46]. To test this idea, isolated brain MT from 57 week C57Bl6J mice were pretreated with 0.2, 1, and 10 μM XJB-5-131 for 40 min, followed by sequential addition of ADP, oligomycin, and carbonylcyanide p-trifluoromethoxyphenylhydrazone (FCCP) (see Fig. 2a) [51]. The impact of XJB-5-131 on state II, III, IVO, and IIIu respiration was measured (Fig. 2b).

States II and IV are direct indicators of “uncoupled states” [47-50]. State II is a basal respiration which begins when NADH provides an electron source. An elevation of state II OCR would occur only if XJB-5-131 dissipated the proton gradient, i.e., from proton leak. To measure it, MT, in assay buffer, were held in respiratory state II with succinate and rotenone in the absence of ADP or ATP (see Fig. 2b). Addition of rotenone inhibits complex I and prevents backward electron flow from Complex II (Fig. 2a). Since the oxphos cannot occur if ADP is absent, OCR in State II is normally reduced [49, 50]. In contrast, we found that XJB-5-131 induced a dose-dependent OCR increase in state II (Fig. 4a, b). The results suggested that XJB-5-131 impaired the proton gradient relative to untreated MT alone (Fig. 4a, b). State III is the ADP stimulated respiration of isolated coupled mitochondria. However, the rise in OCR after ADP addition was reduced by XJB-5-13 relative to untreated MT. The result suggested that elevation of OCR in uncoupled state II reduced the maximum change in OCR that was observed in state III upon addition of ADP.

Fig. 4.

Fig. 4.

Open in a new tab

XJB-5-131 induces mild uncoupling of mitochondria. A) OCR in mitochondria that were untreated or pretreated with XJB-5-131. Representative profiles of 40-min pre-incubation of XJB-5-131 on oxygen consumption rate (OCR) during respiration. Initial conditions are MT in assay buffer as in Fig. 3, plus added inorganic phosphate and succinate (as in Fig. 3). Addition of ADP initiates mitochondrial state III respiration. CI-CIV substrate concentrations were as indicated in Fig. 3; with an additional treatment of fluoro-carbonyl cyanide phenylhydrazone (FCCP) (2 μM). Oligomycin (Oligo) is an inhibitor of ATP synthase and injection of Oligo blocks oxidative phosphorylation and OCR decreases. Fluoro-carbonyl cyanide phenylhydrazone (FCCP) is a mitochondrial uncoupler which destroys the proton gradient and induces uncoupled respiration (IIIU). Oligomycin (Oligo) is an inhibitor of ATP synthase. In that state, OCR increases but does not produce ATP. Antimycin A (AA) is an inhibitor of Complex III and addition of AA disables the electron transport chain and OCR decreases to the minimum. A) Untreated MT are coupled and exhibit high State III respiration (gray). The presence of XJB-5-131 results in a dose-dependent increase in mitochondrial state II and IVO respiration, indicating a mild uncoupling effect (XJB-5-131 dose is indicated by color key). Due to the uncoupling, OCR in state III is lower with dose of XJB-5-131. B) Quantification of OCR results in A). Each was OCR acquired in 15 independent experiments performed in the same buffer on the same day in the same instrument, yielding small errors. The 15 datapoints overlap significantly. *p < 0.05 for the difference between untreated MT and 1.0 μM FCCP-treated in state II, using the unpaired two-tailed Student’s t-test. C) The same as in (A) except that MT were pre-incubated for 40-min with FCCP. FCCP titration included 0.125, 0.25, 0.5, 1, and 2 μM, as indicated in the standard curve. FCCP uncoupling results in a dose-dependent increase in State II and state IVo. In the presence of 2 μM FCCP, mitochondria are essentially uncoupled, which increases OCR in state II and IVO to a rate equal to that of State III. D) Quantification of (C) is as described in (B). *p < 0.05 for the difference between untreated MT and 1.0 μM FCCP-treated in state II, using the unpaired two-tailed Student’s t-test.

Uncoupling was also suggested by the increase in respiration state IV0. Oligomycin will block complex V conversion of ADP to ATP without effects on the proton gradient, and OCR will decrease to the level of substrate alone. However, if ATP synthase is inhibited by oligomycin treatment, a residual oxygen consumption would occur solely due to proton leak. Indeed, OCR decreased in untreated MT after oligomycin addition, as expected, but in the presence of XJB-5-131, OCR in state IVo was higher relative to substrate alone (Fig. 4a, b). The uncoupling effect of XJB-5-131 increased with dose. Collectively, XJB-5-131 appeared to induce a mild uncoupled state, which was evident by the small, but significant increases in both State II and state IVo respiration (Fig. 4a, b). In a final experiment, we followed the impact of XJB-5-131 on the maximum OCR in State IIIu induced by the potent oxidative phosphorylation uncoupler, FCCP (2 μM) (Fig. 4a, b). FCCP destroys the proton gradient (Fig. 2a) and promotes continuous passage of electrons, which is accompanied by a substantial increase in OCR. The increase reaches a maximum in state IIIu when oxidative phosphorylation is fully uncoupled (Fig. 4a, b). We expected that the maximum rise OCR induced by FCCP would be reduced by XJB-5-131 if it also had uncoupling activity, i.e., the maximum OCR would sum the impact of both uncouplers. In untreated MT, FCCP increased OCR to a maximum in state IIIu. However, the magnitude of the rise was reduced in the presence of XJB-5-131 in a dose dependent manner (Fig. 4a, b). The result was specific for the state IIIu respiratory activity since inhibition of complex III with antimycin attenuated OCR (Fig. 4a, b). As a control, we repeated the experiment in (c) for FCCP alone (Fig. 4c, d). Even at low concentrations (below 2.0 μM), uncoupling was strong and overall OCR level was higher than untreated MT in states II and IVo. However, in both states, the dose-dependent differences in OCR were smaller between untreated and treated MT in response to FCCP (Fig. 4a, b). The same effects were observed on the maximum ADP-dependent OCR in state III and on OCR from the fully uncoupled state IIIu. At the highest dose of FCCP tested (2 μM), MT uncoupling was so strong that OCR was not only high, but it was also relatively insensitive to injection of ADP and was not further stimulated by a second injection of FCCP, as indicated (Fig. 4c, d). Collectively, these results supported a mild uncoupling mechanism for XJB-5-131.

DISCUSSION

Over the years, we have reported that XJB-5-131 significantly ameliorates disease markers in a knock-in mouse model of HD [39-41]. Others have demonstrated beneficial effects of XJB-5-131 on suppressing toxicity from lethal hemorrhagic shock [29], reperfusion injury [34], damage to brain tissue during microdialysis [35], cardiolipin oxidation and neuronal death in traumatic brain injury [33, 36], and whole-body irradiation [38]. Nevertheless, the mechanism by which XJB-5-131 exerted its beneficial effects was not fully understood. Here, we provide evidence that XJB-5-131 in the concentration range of 0.2 to 10 μM is a mild uncoupler of oxidative phosphorylation. XJB-5-131 is both an electron scavenger (Fig. 1) and an antioxidant. Superoxide (O2−)2 and H2O2 are produced by leaks of electrons from donor redox centers along the mitochondrial ETC by either one-electron or two-electron reduction of oxygen. Strong uncouplers such as FCCP destroy the proton gradient and suppress ROS at the expense of ATP production, which is ultimately toxic. We show here that XJB-5-131 is a mild uncoupler and the beneficial properties arise from redox chemistry. Because of the nitroxide radical moiety in its structure, XJB-5-131 captures the electrons escaping from the ETC during proton leak, thereby suppressing ROS, i.e., via production of superoxide radical anion (O2•−) and peroxide [42, 47, 48] (Fig. 1c). At the same time, we find that XJB-5-131 has no impact on electron transfer along the ETC (Fig. 3). Our results support a model in which the redox superoxide dismutase-like properties of XJB-5-131 return electrons to the ETC. Thus, XJB-5-131 suppresses the mitochondrial O2−/H2O2 generation without impairing the electron transport that drives CV, the ATP synthase. We propose that the mild uncoupling activity underlies the remarkable efficacy of XJB-5-131 in both in vitro and in vivo models.

Indeed, compounds that reduce ROS production but maintain sufficient ATP production are known to be therapeutically valuable in mouse and in humans [51-54]. For example, high concentrations of FCCP are toxic but a low concentration of FCCP (100 nM being the optimal concentration) is cardioprotective [51]. In other work, we have reported that XJB-5-131 acts directly on MT and increases their copy number, promotes neuronal survival, prevents weight loss and motor function decline in mouse models of HD [39-41]. Furthermore, in studies from other laboratories, treatment with 2,4-dinitrophenol yields a statistically significant higher post-thaw survival of rhesus monkey sperm [55], in flies [56], and in yeast [57]. We propose a model in which the mild uncoupling property of XJB-5-131 exerts its protective effect by reducing oxidative damage without significant reductions in ATP production.

Based on the action of other known uncouplers, the mild activity of XJB-5-131 is beneficial in two ways. Protonophores (such as FCCP and DNP) diffuse across the lipid bilayer driven, in part, by the membrane potential, and dissipate some of the gradient (reducing ROS) by transporting protons back into MT [58-61]. However, most uncouplers are charged, and the positive or negative state and their effects are sensitive to the membrane potential [59, 60]. For example, when hydrophobic penetrating cations are targeted to MT, in combination with low (non-toxic) concentrations of protonophores, uncoupling is stimulated [59-61]. However, one can envision conditions in which the charge has undesirable effects on the MT membrane, leading to depolarization, unacceptable increases in ROS damage or diminished ATP production, if uncoupling is too low or too high, respectively [62, 63]. In contrast, XJB-5-131 is neutral and cannot act as a proton shuttle. Therefore, it does not alter the membrane gradient (Fig. 3), nor is its entry sensitive to the membrane potential. Thus, the mild uncoupling effect of neutral XJB-5-131, on its own, is novel and particularly attractive as a therapeutic agent since it can enter and decrease ROS even in depolarized MT.

Mitochondria homeostasis is maintained by balancing the need for ATP production at the deleterious effects of oxidative damage. We provide new evidence that XJB-5-131 at concentrations ≤ 10 μM markedly reduces both basal and DMNQ-induced peroxide levels (Fig. 1) without affecting mitochondrial electron flow (Fig. 3). The mild uncoupling property of XJB-5-131 suggests a valuable mechanism of action for the design of clinically effective antioxidants and warrants further consideration as a clinically relevant candidate.

ACKNOWLEDGMENTS

This work was supported by the National Institutes of Health grants GM119161 (to CTM), AG070972 (to CTM), and NS060115 (to CTM), and GM067082 (to PW).

Footnotes

CONFLICT OF INTEREST

There are no conflicts or competing interests by the authors.

REFERENCES

  • [1].Hu L, Wang Z, Carmone C, Keijer J, Zhang D. Role of oxidative DNA damage and repair in atrial fibrillation and ischemic heart disease. Int J Mol Sci. 2021;22(8):3838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Luo J, Mills K, le Cessie S, Noordam R, van Heemst D. Ageing, age-related diseases and oxidative stress: What to do next? Ageing Res Rev. 2020;57:100982. [DOI] [PubMed] [Google Scholar]
  • [3].Yousefzadeh M, Henpita C, Vyas R, Soto-Palma C, Robbins P, Niedernhofer L. DNA damage-how and why we age? eLife. 2021;10:e62852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Cai Q, Jeong YY. Mitophagy in Alzheimer’s disease and other age-related neurodegenerative diseases. Cells. 2020;9(1):150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Jové M, Mota-Martorell N, Torres P, Ayala V, Portero-Otin M, Ferrer I, et al. The causal role of lipoxidative damage in mitochondrial bioenergetic dysfunction linked to Alzheimer’s disease pathology. Life. 2021;11(5):388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Polyzos AA, Lee DY, Datta R, Hauser M, Budworth H, Holt A, et al. Metabolic reprogramming in astrocytes distinguishes region-specific neuronal susceptibility in Huntington mice. Cell Metab. 2019;29(6):1258–73.e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Polyzos AA, McMurray CT. The chicken or the egg: Mitochondrial dysfunction as a cause or consequence of toxicity in Huntington’s disease. Mech Ageing Dev. 2017;161(Pt A):181–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Sharma A, Behl T, Sharma L, Aelya L, Bungau S. Mitochondrial dysfunction in Huntington’s disease: Pathogenesis and therapeutic opportunities. Curr Drug Targets. 2021;22(14):1637–67. [DOI] [PubMed] [Google Scholar]
  • [9].Trinh D, Israwi AR, Arathoon LR, Gleave JA, Nash JE. The multi-faceted role of mitochondria in the pathology of Parkinson’s disease. J Neurochem. 2020;156(6):715–52. [DOI] [PubMed] [Google Scholar]
  • [10].Cotticelli MG, Xia S, Lin D, Lee T, Terrab L, Wipf P, et al. Ferroptosis as a novel therapeutic target for Friedreich’s ataxia. J Pharmacol Exp Ther. 2019;369(1):47–54. [DOI] [PubMed] [Google Scholar]
  • [11].Harley J, Clarke BE, Patani R. The interplay of RNA binding proteins, oxidative stress and mitochondrial dysfunction in ALS. Antioxidants. 2021;10(4):552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Horváth B, Magid L, Mukhopadhyay P, Bátkai S, Rajesh M, Park O, et al. A new cannabinoid CB2 receptor agonist HU-910 attenuates oxidative stress, inflammation and cell death associated with hepatic ischaemia/reperfusion injury. Bri J Pharmacol. 2012;165(8):2462–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Shang L, Ren H, Wang S, Liu H, Hu A, Gou P, et al. SS-31 protects liver from ischemia-reperfusion injury via modulating macrophage polarization. Oxid Med Cell Longevity. 2021;2021:6662156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Basso M, Ratan RR. Transglutaminase is a therapeutic target for oxidative stress, excitotoxicity and stroke: A new epigenetic kid on the CNS block. J Cereb Blood Flow Metab. 2013;33(6):809–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Peng W, Cai G, Xia Y, Chen J, Wu P, Wang Z, et al. Mitochondrial dysfunction in atherosclerosis. DNA Cell Biol. 2019;38(7):597–606. [DOI] [PubMed] [Google Scholar]
  • [16].Ghorbani M, Izadi Z, Jafari S, Casals E, Rezaei F, Aliabadi A, et al. Preclinical studies conducted on nanozyme antioxidants: Shortcomings and challenges based on US FDA regulations. Nanomed. 2021;16(13):1133–51. [DOI] [PubMed] [Google Scholar]
  • [17].Hidalgo-Gutiérrez A, González-García P, Díaz-Casado ME, Barriocanal-Casado E, López-Herrador S, Quinzii CM, et al. Metabolic targets of Coenzyme Q10 in mitochondria. Antioxidants. 2021;10(4):520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].A randomized clinical trial of coenzyme Q10 and GPI-1485 in early Parkinson disease. Neurol. 2007;68(1):20–8. [DOI] [PubMed] [Google Scholar]
  • [19].Mangialasche F, Solomon A, Winblad B, Mecocci P, Kivipelto M. Alzheimer’s disease: Clinical trials and drug development. Lancet Neurol. 2010;9(7):702–16. [DOI] [PubMed] [Google Scholar]
  • [20].Chandel NS, Tuveson DA. The promise and perils of antioxidants for cancer patients. N Engl J Med. 2014;371(2):177–8. [DOI] [PubMed] [Google Scholar]
  • [21].Kawagishi H, Finkel T. Unraveling the truth about antioxidants: ROS and disease: Finding the right balance. Nat Med. 2014;20(7):711–3. [DOI] [PubMed] [Google Scholar]
  • [22].Krakauer T. Inflammasomes, autophagy, and cell death: The trinity of innate host defense against intracellular bacteria. Mediators Inflamm. 2019;2019:2471215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Ristow M. Unraveling the truth about antioxidants: Mitohormesis explains ROS-induced health benefits. Nat Med. 2014;20(7):709–11. [DOI] [PubMed] [Google Scholar]
  • [24].da Silva HNM, Covatti C, da Rocha GL, Mizobuti DS, Mâncio RD, Hermes TA, et al. Oxidative stress, inflammation, and activators of mitochondrial biogenesis: Tempol targets in the diaphragm muscle of exercise trained-mdx mice. Front Physiol. 2021;12:649793. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Kagan VE, Wipf P, Stoyanovsky D, Greenberger JS, Borisenko G, Belikova NA, et al. Mitochondrial targeting of electron scavenging antioxidants: Regulation of selective oxidation vs random chain reactions. Adv Drug Delivery Rev. 2009;61(14):1375–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Scialo F, Sanz A. Coenzyme Q redox signalling and longevity. Free Radical Biol Med. 2021;164:187–205. [DOI] [PubMed] [Google Scholar]
  • [27].Wipf P, Xiao J, Jiang J, Belikova NA, Tyurin VA, Fink MP, et al. Mitochondrial targeting of selective electron scavengers: Synthesis and biological analysis of hemigramicidin–TEMPO conjugates. J Am Chem Soc. 2005;127(36):12460–1. [DOI] [PubMed] [Google Scholar]
  • [28].Wu M, Zhang C, Xie M, Zhen Y, Lai B, Liu J, et al. Compartmentally scavenging hepatic oxidants through AMPK/SIRT3-PGC1α axis improves mitochondrial biogenesis and glucose catabolism. Free Radical Biol Med. 2021;168:117–28. [DOI] [PubMed] [Google Scholar]
  • [29].Wipf P, Polyzos AA, McMurray CT. A double-pronged sword: XJB-5-131 is a suppressor of somatic instability and toxicity in Huntington’s disease. J Huntingtons Dis. 2022;11(1):3–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Apostolova N, Victor VM. Molecular strategies for targeting antioxidants to mitochondria: Therapeutic implications. Antioxid Redox Signal. 2015;22(8):686–729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Fink MP, Macias CA, Xiao J, Tyurina YY, Jiang J, Belikova N, et al. Hemigramicidin–TEMPO conjugates: Novel mitochondria-targeted anti-oxidants. Biochem Pharmacol. 2007;74(6):801–9. [DOI] [PubMed] [Google Scholar]
  • [32].Hoye AT, Davoren JE, Wipf P, Fink MP, Kagan VE. Targeting mitochondria. Acc Chem Res. 2008;41(1):87–97. [DOI] [PubMed] [Google Scholar]
  • [33].Ji J, Kline AE, Amoscato A, Samhan-Arias AK, Sparvero LJ, Tyurin VA, et al. Lipidomics identifies cardiolipin oxidation as a mitochondrial target for redox therapy of brain injury. Nat Neurosci. 2012;15(10):1407–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Zhao Z, Wu J, Xu H, Zhou C, Han B, Zhu H, et al. XJB-5-131 inhibited ferroptosis in tubular epithelial cells after ischemia-reperfusion injury. Cell Death Dis. 2020;11(8):629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Nesbitt KM, Jaquins-Gerstl A, Skoda EM, Wipf P, Michael AC. Pharmacological mitigation of tissue damage during brain microdialysis. Anal Chem. 2013;85(17):8173–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Di Pietro V, Yakoub KM, Caruso G, Lazzarino G, Signoretti S, Barbey AK, et al. Antioxidant therapies in traumatic brain injury. Antioxidants. 2020;9(3):260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Nasto LA, Robinson AR, Ngo K, Clauson CL, Dong Q, St Croix C, et al. Mitochondrial-derived reactive oxygen species (ROS) play a causal role in aging-related intervertebral disc degeneration. J Orthop Res. 2013;31(7):1150–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Epperly MW, Fisher R, Zhang X, Hou W, Shields D, Wipf P, et al. Fanconi anemia mouse genotype-specific mitigation of total body irradiation by GS-nitroxide JP4-039. In Vivo. 2020;34(1):33–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Polyzos A, Holt A, Brown C, Cosme C, Wipf P, Gomez-Marin A, et al. Mitochondrial targeting of XJB-5-131 attenuates or improves pathophysiology in HdhQ150 animals with well-developed disease phenotypes. Hum Mol Genet. 2016;25(9):1792–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Polyzos AA, Wood NI, Williams P, Wipf P, Morton AJ, McMurray CT. XJB-5-131-mediated improvement in physiology and behaviour of the R6/2 mouse model of Huntington’s disease is age- and sex- dependent. PLoS One. 2018;13(4):e0194580–e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Xun Z, Rivera-Sánchez S, Ayala-Peña S, Lim J, Budworth H, Skoda EM, et al. Targeting of XJB-5-131 to mitochondria suppresses oxidative DNA damage and motor decline in a mouse model of Huntington’s disease. Cell Rep. 2012;2(5):1137–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Sheng Y, Abreu IA, Cabelli DE, Maroney MJ, Miller A-F, Teixeira M, et al. Superoxide dismutases and superoxide reductases. Chem Rev. 2014;114(7):3854–918. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Ishihara Y, Shiba D, Shimamoto N. Enhancement of DMNQ-induced hepatocyte toxicity by cytochrome P450 inhibition. Toxicol Appl Pharmacol. 2006;214(2):109–17. [DOI] [PubMed] [Google Scholar]
  • [44].Parry JD, Pointon AV, Lutz U, Teichert F, Charlwood JK, Chan PH, et al. Pivotal role for two electron reduction in 2,3-Dimethoxy-1,4-naphthoquinone and 2-Methyl-1,4-naphthoquinone metabolism and kinetics in vivo that prevents liver redox stress. Chem Res Toxicol. 2009;22(4):717–25. [DOI] [PubMed] [Google Scholar]
  • [45].Salabei JK, Gibb AA, Hill BG. Comprehensive measurement of respiratory activity in permeabilized cells using extracellular flux analysis. Nat Protoc. 2014;9(2):421–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Rogers GW, Brand MD, Petrosyan S, Ashok D, Elorza AA, Ferrick DA, et al. High throughput microplate respiratory measurements using minimal quantities of isolated mitochondria. PloS One. 2011;6(7):e21746–e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Jastroch M, Divakaruni AS, Mookerjee S, Treberg JR, Brand MD. Mitochondrial proton and electron leaks. Essays Biochem. 2010;47:53–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].Mookerjee SA, Divakaruni AS, Jastroch M, Brand MD. Mitochondrial uncoupling and lifespan. Mech Ageing Dev. 2010. 131(7-8):463–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Echtay KS, Roussel D, St-Pierre J, Jekabsons MB, Cadenas S, Stuart JA, Harper JA, Roebuck SJ, Morrison A, Pickering S, Clapham JC, Brand MD. Superoxide activates mitochondrial uncoupling proteins. Nature. 2002;415(6867):96–9. [DOI] [PubMed] [Google Scholar]
  • [50].Toime LJ, Brand MD. Uncoupling protein-3 lowers reactive oxygen species production in isolated mitochondria. Free Radical Biol Med. 2010;49(4):606–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [51].Brennan J, Southworth R, Medina R, Davidson S, Duchen M, Shattock M. Mitochondrial uncoupling, with low concentration FCCP, induces ROS-dependent cardioprotection independent of KATP channel activation. Cardiovasc Res. 2006;72(2):313–21. [DOI] [PubMed] [Google Scholar]
  • [52].Caldeira da Silva CC, Cerqueira FM, Barbosa LF, Medeiros MHG, Kowaltowski AJ. Mild mitochondrial uncoupling in mice affects energy metabolism, redox balance and longevity. Aging Cell. 2008;7(4):552–60. [DOI] [PubMed] [Google Scholar]
  • [53].Goedeke L, Shulman GI. Therapeutic potential of mitochondrial uncouplers for the treatment of metabolic associated fatty liver disease and NASH. Mol Metab. 2021;46:101178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [54].Shrestha R, Johnson E, Byrne FL. Exploring the therapeutic potential of mitochondrial uncouplers in cancer. Mol Metab. 2021;51:101222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [55].Dong Q, Tollner TL, Rodenburg SE, Hill DL, VandeVoort CA. Antioxidants, Oxyrase, and mitochondrial uncoupler 2,4-dinitrophenol improved postthaw survival of rhesus monkey sperm from ejaculates with low cryosurvival. Fertil Steril. 2010;94(6):2359–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [56].Padalko VI. Uncoupler of oxidative phosphorylation prolongs the lifespan of Drosophila. Biochemistry (Mosc). 2005;70(9):986–9. [DOI] [PubMed] [Google Scholar]
  • [57].Barros MH, Bandy B, Tahara EB, Kowaltowski AJ. Higher respiratory activity decreases mitochondrial reactive oxygen release and increases life span in Saccharomyces cerevisiae. J Biol Chem. 2004;279(48):49883–8. [DOI] [PubMed] [Google Scholar]
  • [58].Bestman JE, Stackley KD, Rahn JJ, Williamson TJ, Chan SS. The cellular and molecular progression of mitochondrial dysfunction induced by 2,4-dinitrophenol in developing zebrafish embryos. Differentiation. 2015;89(3-4):51–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [59].Antonenko YN, Khailova LS, Knorre DA, Markova OV, Rokitskaya TI, Ilyasova TM, et al. Penetrating cations enhance uncoupling activity of anionic protonophores in mitochondria. PLoS One. 2013;8(4):e61902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [60].Chopineaux-Courtois V, Reymond F, Bouchard G, Carrupt PA, Testa B, Girault HH. Effects of charge and intramolecular structure on the lipophilicity of nitrophenols. J Am Chem Soc. 1999;121(8):1743–7. [Google Scholar]
  • [61].Ozaki S, Kano K, Shirai O. Electrochemical elucidation on the mechanism of uncoupling caused by hydrophobic weak acids. Phys Chem Chem Phys. 2008;10(30):4449–55. [DOI] [PubMed] [Google Scholar]
  • [62].Demine S, Renard P, Arnould T. Mitochondrial uncoupling: A key controller of biological processes in physiology and diseases. Cells. 2019;8(8):795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [63].Abeti R, Parkinson MH, Hargreaves IP, Angelova PR, Sandi C, Pook MA, et al. Mitochondrial energy imbalance and lipid peroxidation cause cell death in Friedreich’s ataxia. Cell Death Dis. 2016;7:e2237. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES