A novel mitochondrial complex I ROS inhibitor partially improves muscle regeneration in adult but not old mice

It is unclear whether mitochondrial dysfunction and redox stress contribute to impaired age-related muscle regenerative capacity. Here we characterized a novel compound, BI4500, that inhibits the release of reactive oxygen species (ROS) from the quinone site in mitochondrial complex I (site IQ). We tested the hypothesis that ROS release from site IQ contributes to impaired regenerative capacity in aging muscle. Electron transfer system site-specific ROS production was measured in adult and aged mouse isolated muscle mitochondria and permeabilized gastrocnemius fibers. BI4500 inhibited ROS production from site IQ in a concentration-dependent manner (IC50 = ∼985 nM) by inhibiting ROS release without impairing complex I-linked respiration. In vivo BI4500 treatment decreased ROS production from site IQ. Muscle injury and sham injury were induced using barium chloride or vehicle injection to the tibialis anterior (TA) muscle in adult and aged male mice. On the same day as injury, mice began a daily gavage of 30 mg/kg BI4500 (BI) or placebo (PLA). Muscle regeneration (H&E, Sirius Red, Pax7) was measured at 5 and 35 days after injury. Muscle injury increased centrally nucleated fibers (CNFs) and fibrosis with no treatment or age effect. There was a significant age by treatment interaction for CNFs at 5- and 35-days post injury with significantly more CNFs in BI adults compared to PLA adults. Muscle fiber cross-sectional area (CSA) recovered significantly more in adult BI mice (−89 ± 365 μm2) compared to old PLA (−599 ± 153 μm2) and old BI (−535 ± 222 μm2, mean ± SD). In situ TA force recovery was measured 35 days after injury and was not significantly different by age or treatment. Inhibition of site IQ ROS partially improves muscle regeneration in adult but not old muscle demonstrating a role for CI ROS in the response to muscle injury. Site IQ ROS does not contribute to impaired regenerative capacity in aging.


Introduction
Skeletal muscle is a highly plastic tissue, capable of adapting to various stressors encountered throughout life [1,2]. Aging is associated with declines in adaptive muscle plasticity related to muscle function and mass [3][4][5], which exacerbate muscle loss, low quality of life and increased risk of morbidity and mortality [6]. Maintaining muscle mass and function are therefore critical for maintaining independence and quality of life in older adults. Muscle satellite cells (MuSCs) are tissue-resident stem cells responsible for the regeneration of new muscle through gene programs that regulate proliferation, differentiation and fusion of new myotubes [7]. These gene programs are turned on in response to various models of muscle injury, as well as exercise [8]. However, aging leads to MuSC senescence preventing activation of the proliferation program, decreasing regenerative capacity, and increasing susceptibility to age related comorbidities and dysfunction [9].
One potential explanation for the declining repair capacity of skeletal muscle satellite cells is age-associated mitochondrial dysfunction and concomitant redox stress [10]. Aging mitochondria have increased reactive oxygen species (ROS) production, shifting the cellular milieu to a more oxidized state, which could alter satellite cell activation and contribute to declining regenerative capacity [10,11]. Several studies (caption on next page) G. Pharaoh et al. have demonstrated that targeting redox active compounds to mitochondria or genetic over expression of antioxidant enzymes in mitochondria can restore age-related declines in function and the redox proteome to young levels [12][13][14]. Therefore, targeting mitochondria has therapeutic potential for developing and testing new pharmaceuticals to improve age-related muscle plasticity and regeneration.
We previously reported mitochondrial electron transport system (ETS) complex I protein abundance and function are affected by aging in mouse skeletal muscle [15]. ROS released from complex I sites I F and I Q in skeletal muscle mitochondria account for about half of the total ROS production during rest and the vast majority of mitochondrial ROS during exercise [16]. ROS produced from these sites by reverse electron transport (RET) drives tissue damage after ischemia-reperfusion injury and during aging, and limiting ROS released from complex I has become a major research topic to improve lifespan and healthspan [17][18][19]. The complex relationship between CI ROS production and muscle differentiation has been studied in vitro, but it is unknown how CI ROS production contributes to muscle differentiation in vivo [20][21][22][23]. Here we characterized a novel compound (BI4500) that inhibits superoxide (O 2 − ) release from RET at mitochondrial complex I Q (Graphical Abstract).
We then used this compound to test the hypothesis that treatment with a RET complex I inhibitor would improve muscle regeneration after injury in adult and old mice compared to placebo.

Animal husbandry
C57Bl6/J male and female mice were purchased from The Jackson Laboratory and housed at the University of Washington. All mice were maintained at 21 • C on a 14/10 light/dark cycle at at 30-70% humidity and given standard mouse chow (LabDiet PicoLab® Rodent Diet 20) and water ad libitum unless otherwise specified. This study was reviewed and approved by the University of Washington Institutional Animal Care and Use Committee (IACUC).

BI4500
BI4500 is an indolinone composed of C 20 H 27 N 3 O 3 and was manufactured by Boehringer Ingelheim International with greater than 99.5% purity.

Direct H 2 O 2 scavenging assay
To test whether BI4500 had direct H 2 O 2 scavenging capabilities, a plate-based Amplex UltraRed assay was used identical to the sitespecific isolated mitochondria ROS assay but without biological samples. BI concentrations of 0, 1, and 10 μM were incubated with a standard curve of H 2 O 2 in triplicate.

Mitochondrial isolation and superoxide release from sites in complex I
The gastrocnemius, quadriceps femoris, and tibialis anterior (TA) muscles were dissected from both hindlimbs, and mitochondrial isolation was performed by differential centrifugation. The production of O 2 − and hydrogen peroxide (H 2 O 2 ) from the ETS complex I sites was measured using Amplex UltraRed (Molecular Probes, Eugene, OR) as described in previous publications with some modifications [24,25]. ROS production from site I Q reverse was measured in mitochondria from 12 mo adult and 24-26 mo old female mice using 5 mM succinate and 0.3 μM S3QEL1.2 normalized to the inhibitor control of 5 mM succinate, 0.3 μM S3QEL1.2, and 1 μM nigericin at the same concentrations of BI4500 in duplicate (Fig. 1A). Succinate as a substrate stimulates O 2 − release from sites I Q , I F , and III Qo [26]. S3QEL1.2 (Sigma SML1554) was added to specifically inhibit site III Qo . Nigericin (Sigma N7143) collapses the pH gradient necessary for I Q O 2 − release. After 8 weeks of control chow or BI4500 chow feeding, ROS production from specific sites was measured in 11-12 mo adult and 24-26 mo old female mice using the substrate and inhibitors described in columns in Table S1 in triplicate (Fig. 1C). Detailed methods are available in the supplemental information.

Muscle fiber preparation and respirometry and fluorometry
Several permeabilized red gastrocnemius muscle fiber bundles per animal were prepared from 6 mo adult and 29 mo old female mice as previously described with some modifications (Fig. 1B) [27][28][29]. Fibers were permeabilized and washed with vehicle, 1, or 10 μM BI4500. To measure site-specific hydroperoxide production in permeabilized muscle fibers stimulated with succinate, we developed a SUIT protocol designed to isolate ROS production from specific sites using sequential additions of succinate, S3QEL1.2, nigericin, and ADP. Detailed methods were previously described, and additional details are available in the supplemental information [27][28][29].

BI chow feeding and oral gavage pharmacokinetic studies
We fed adult (9-10 mo) and aged (22-24 mo) female mice control chow for 2 weeks then BI4500 chow (580 mg BI4500 per kg chow) for 8 weeks (Figs. S1A and 1C). Age-and sex-matched control mice were maintained on control chow. The mice were sacrificed at 12 months and 24-26 months of age. Plasma was collected to measure BI45000 concentration and mitochondria were isolated from the muscles of these mice and used to measure ROS production. ROS production was compared to age-and sex-matched mouse fed standard chow. A single dose and once-daily oral gavage pharmacokinetic studies at 30 mg/kg BI4500 were performed (Fig. 1D). Additional details are available in supplemental methods. BI4500 concentrations were determined using Fig. 1. A) I Q ROS dose response: Gastrocnemius (gastroc), quadriceps femoris (quad), and tibialis anterior (TA) muscles were dissected from adult 12 mo and old 24-26 mo mice, their mitochondria were isolated, and then ROS production from site I Q was measured with a dose response of BI45000. B) CI-Linked respiration and ROS: the gastrocnemius (gastroc) muscle was dissected from adult 6 mo or old 29 mo female mice and fibers were mechanically separated, permeabilized, and washed. During the permeabilization and wash steps, fibers were exposed to vehicle or BI4500. CI-linked respiration and ROS production was measured. C) Chow feeding study: Adult 9-10 mo and old 22-24 mo female mice were fed control chow or BI4500 chow (580 mg/kg) for 8 weeks. Mitochondria were isolated from their muscle and CI site-specific ROS production was measured. Plasma BI4500 concentration was determined. D) Oral Gavage PK study: 4 mo male mice were given daily oral gavage of BI4500 (30 mg BI4500/kg body mass) for 7 days. Plasma BI4500 concentration was determined after 7 days. E) Muscle regeneration study design: Adult and aged animals were treated with BI4500 daily for 5 days or 35 days after muscle injury. mass spectrometry by PharmaCadence Analytical Services (Hatfield, Pennsylvania).

Muscle injury and regeneration study with BI4500 oral gavage
We used intramuscular sham and barium chloride (BaCl 2 ) injections to injure TA muscles of adult (9-11mo) and old (26mo) animals ( Fig. 1E). Animals began daily oral gavage treatment with either placebo control (PLA) or 30 mg/kg/day BI4500 compound (BI) the same day as muscle injury. Treatment was administered daily for either 5 days or 35 days and the TA muscle was collected. The cohort that was treated for 35 days underwent in situ muscle function testing followed by TA collection for histological measures. Additional details are available in supplemental methods.

In situ TA muscle function
On day 35, animals were anesthetized with 4% isoflurane induction and placed on a heat pad under 1-2% isoflurane to maintain anesthesia during the procedure. Skin was removed from the ankle to the knee, followed by careful dissection of the surrounding connective tissue superficial to the TA. The distal TA tendon was cut below the ankle and sutured to the force transducer. Electrodes were placed subcutaneously behind the knee to stimulate the peroneal nerve. Following electrode placement, a voltage titration was performed to identify maximal twitch force (6-10V) followed by a force frequency test. Force was measured every minute with increasing frequency from 10 to 200 Hz. Maximal TA force was recorded from the force frequency curve. Both TA muscles were tested in each animal with the injured limb always being tested first to control for an order effect of muscle stimulation.

Histology
After TA muscle collection, muscles were immediately placed in 0.4% paraformaldehyde -PBS solution for 2 h at 4 • C followed by incubation in 20% sucrose-PBS at 4 • C overnight. Muscles were removed from sucrose-PBS and immediately frozen in OCT using liquid N 2 cooled isopentane. Frozen tissue was stored at − 80 • C until sectioning. Eight μm cross sections were cut using a Leica CM1950 (Leica Biosystems, Deer Park IL) and placed on glass slides. Slides were stored at − 80 • C until staining.

Staining & imaging
H&E, Sirius red, or immunohistochemistry protocols were used to characterize muscle morphology, centrally nucleated fibers, fibrosis, and Pax7 positive cells, respectively, using the protocols previously described [30]. Detailed methods are available in the supplemental information.

Statistical analysis
Graphing and statistical analysis were performed using GraphPad Prism for OS X (GraphPad Software, San Diego, California USA). For all statistical tests, alpha levels were set to p < 0.05. Detailed methods are available in the supplemental information.

BI4500 inhibits ROS production from site I Q in a dose-dependent manner
We tested whether BI4500 prevents ROS release from site I Q in muscle mitochondria from adult and old mice with a dose-response of BI4500 (Fig. 1A). BI4500 inhibits ROS release from site I Q in both adult and aged isolated muscle mitochondria with an IC 50 of ~985 nM and complete inhibition in aged mitochondria around 5-10 μM (Fig. 2).
We measured site-specific ROS production in gastrocnemius muscle fibers from adult (6-8 mo) and aged (29 mo) mice after pre-treatment ex vivo during the fiber permeabilization and wash steps with vehicle or 1 μM BI4500 (Fig. 1B). Incubation with vehicle or BI4500 during fiber permeabilization and washing results in ~1 h of treatment before measuring ROS production. This dose showed inhibitory effects on site I Q ROS production in isolated mitochondria, while avoiding the potential for interference observed at 10 μM in the Amplex Red reaction (Figs. S2E-F). BI4500 significantly decreased total ROS production with succinate (p < 0.05 BI4500 effect) (Fig. 3A). In muscle fibers from aged mice, BI4500 decreased ROS production from site I Q (p < 0.05 by t-test) (Fig. 3A). Site I F was unaffected by BI4500 treatment (Fig. 3A). Interestingly, BI4500 pre-treatment decreased the elevated basal hydroperoxide production observed in aged fibers (p < 0.05 age effect, p < 0.01 BI4500 x age effect), which has previously been correlated with the extent of muscle atrophy in various models including aging (Fig. 3A) [31]. This is likely due to micromolar doses of BI4500 displaying inhibitory effects on lipoxygenases. A selectivity study of the compound showed greater than 75% inhibition of ALOX5 and ALOX15 lipoxygenases at 10 μM. ALOX15 partially accounts for increased basal hydroperoxides in aging after spontaneous loss of neuromuscular innervation [31]. BI4500 pre-treatment increased the efficiency of electron transport resulting in less ROS produced per oxygen consumed (p < 0.05 BI4500 effect) (Fig. 3B).

Ex vivo treatment with BI4500 does not significantly impair mitochondrial respiration in permeabilized muscle fibers
We measured mitochondrial respiration in permeabilized gastrocnemius muscle fibers from adult (6-8 mo) and aged (29 mo) mice pretreated ex vivo with vehicle, 1 μM, or 10 μM BI4500 during the fiber permeabilization and wash steps to identify if the compound inhibits mitochondrial respiration (Figs. 1B and 3C). Pre-treatment with BI4500 did not significantly decrease respiration capacities under any substrate condition in either age group (Fig. 3C). Respiration capacities were all significantly decreased with age (p < 0.05 age effect).

BI4500 has low direct superoxide and H 2 O 2 scavenging
We determined whether BI4500 directly scavenges O 2 − and H 2 O 2 in assay conditions without any biological samples. BI4500 has SOD

Eight weeks of BI chow feeding decreases ROS production from site I Q in isolated mitochondria
We fed adult (9-10 mo) and aged (22-24 mo) female mice BI4500 chow for 8 weeks (Fig. 1C). There was no difference in food consumption after switching to BI chow (Fig. S1A). Unfortunately, concentrations of BI4500 were below the limit of quantitation for the plasma samples. Mitochondria were isolated from the combined hindlimb muscles (gastrocnemius, quadriceps, and TA) of BI chow fed mice as well as from age-and sex-matched mice fed control chow to measure site-specific ROS production (Table S1). Despite the low plasma PK values, animals fed BI chow had significantly lower ROS production from site I Q compared to controls (p < 0.0001 BI4500 effect), while site I F reverse was higher (p < 0.05 age effect, p < 0.0001 BI4500 effect) and site I F forward was unchanged (Fig. 4A). Total combined ROS produced by sites I F and I Q were similar across ages and treatments (Fig. 4B). 8-week consumption of BI4500 chow did not significantly affect body mass or hindlimb muscle masses (Fig. 4C-D). , and representative images (right) of Pax7 stained TA cross sections from BaCl 2 injured and sham muscles five days after muscle injury. Pax7 (red) was used as a marker of satellite cells, Laminin (green), and DAPI (blue). Nuclei were counted as satellite cells if DAPI and Pax7 overlapped (magenta), illustrated by the white arrows. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Mean ± SEM. All samples from adult (9-11 mo) and old (27 mo) male mice. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Contribution of I Q ROS production to muscle regeneration
Following these mechanistic studies, we wanted to test the role of complex I Q on muscle regeneration after injury in adult and aged mice. We hypothesized that BI4500 (BI) treatment would improve muscle fiber regeneration in aged animals compared to placebo control (PLA).
To test muscle regeneration, we performed barium chloride (BaCl 2 ) injury on one TA muscle and sham injury on the contralateral TA muscle of adult and aged male mice and simultaneously began a daily oral gavage treatment with BI4500 (Fig. 1E). We changed to oral gavage administration, since compound concentrations in plasma following chow treatment were below detectable limits. Plasma concentrations of BI4500 after one week of 30 mg/kg once daily oral gavage were 18.4 ± 8.2 μM (n = 7, 1 h post dose, mean ± SD), which was in range of the observed Cmax in previous PK studies after oral dosing (Fig. 1D).

Muscle regeneration five days post-injury
Five days after muscle injury there was a significant increase in the number of CNFs of all injured limbs compared to sham controls illustrated in Fig. 5A. There was a significant interaction effect between age and BI4500 treatment (p = 0.045 by Two-Way ANOVA) for the increase in CNFs (Δ) in the injured versus sham limbs (Fig. 5A). Specifically, there was a larger increase in CNFs in the injured limb compared to sham in adult animals treated with BI compared to adult animals treated with PLA (p = 0.043 by Tukey's post-hoc test), while there was no significant difference between the increase in CNFs of the injured limb in PLA and BI treated old animals. Additional quantification of H&E was not performed at 5 days due to the severe muscle damage and clear necrosis at this timepoint. Picrosirius red staining was used to assess muscle fibrosis. Baseline tissue fibrosis was elevated in the sham muscles from old animals consistent with previous reports (Fig. 5B) [32,33]. BaCl 2 injection significantly increased fibrosis in all groups five days after injury (Fig. 5B). There were no significant differences in the increased (Δ) fibrosis response to injury with age or BI4500 treatment among the groups (Fig. 5B). Pax7 staining for satellite cell abundance in response to muscle injury shows no significant differences between adult and old animals or a BI4500 effect (Fig. 5C). This is likely due to the large variation in satellite cell proliferation in response to injury.

Muscle regeneration 35 Days post-injury
Muscle cross sections were analyzed by H&E staining 35 days after injury to assess muscle morphology and regeneration. CNFs remained elevated in the injured limbs even 35 days after injury in all groups (Fig. 6A), suggesting the injured muscle was not fully recovered from the BaCl 2 injections. There was a main effect of age on CNFs (p < 0.001 age effect), with adult mice showing a greater percent increase of CNFs than old mice (Fig. 6A). There was also a BI4500 treatment effect (p = 0.005) and a BI4500 x age interaction (p = 0.016). The age by treatment interaction effects were driven by the adult BI treated group, which had a significantly higher increase in CNFs than each of the other groups (Fig. 6A). These results suggest that BI4500 increased CNFs in adult but not old animals 35 days post injury.
Thirty-five days after muscle injury, muscle fiber cross sectional area (CSA) was still lower in injured limbs compared to sham except for BI treated adult animals ( Fig. 6B-C). CSA showed significant effects of age (p = 0.014) which was expected given that generally aging is associated with lower muscle mass and impaired recovery capacity from injury. There was also a BI4500 x age interaction (p = 0.029) but no main effect of treatment on CSA. Significantly higher CSA differences in BI adults compared to old PLA (p = 0.024) and old BI animals (p = 0.049) indicate that the age by treatment interaction was being driven by the adult BI treated animals.
TA muscle fibrosis remained elevated 35 days after injury compared to sham (Fig. 7A paired comparisons). However, there were no significant differences in the increase in fibrosis between groups in response to injury (Fig. 7A). Pax7 positive staining showed no major age or treatment differences in response to the muscle injury thirty-five days after injury (Fig. 7B). Finally, there were no significant differences in maximal TA muscle force recovery by age or treatment group (Fig. 8A-B).

Discussion
Here we characterized a novel compound targeted to mitochondria. BI4500 decreases ROS production from ETS site I Q not by direct ROS scavenging but by inhibiting ROS release (Graphical Abstract). The conclusion that the BI compound specifically inhibits ROS production from site I Q and is not a general complex I inhibitor is supported by the absence of any effect of acute in vitro treatment on forward or reverse ROS production from the I F site. In contrast to the in vitro results, the inhibition of I Q RET ROS production with 8 weeks of in vivo treatment with BI4500 was associated with a significant elevation of I F RET oxidant production. These results suggest the in vivo treatment resulted in altered structure and function of complex I and support an important signaling role for mitochondrial oxidant production in determining ETS function. Unlike classical site I Q inhibitors like rotenone, BI4500 inhibits I Q ROS release without inhibiting complex I-linked respiration.
It has previously been established that other compounds that inhibit ROS release from site I Q during RET without impairing forward electron flow during respiration do so by binding away from the quinone pocket and modulating the quinone-redox reaction through structural changes in the binding pocket [34]. Conversely, quinone-site inhibitors like rotenone bind directly in the site I Q binding pocket and increase ROS production from site I F by modifying the oxidation-reduction state of Flavin mononucleotide (FMN) [34,35]. The binding site of BI4500 to complex I is currently unknown, but it is possible that the increase in ROS production at site I F during RET could occur either through modifying the oxidation-reduction state of FMN or by inducing structural changes in complex I.
ROS release from site I Q by RET contributes to both physiological and pathological processes in mammals. RET occurs during aging in flies, and treatment with RET inhibitors extends lifespan and healthspan [19]. RET occurs in several tissues in vivo after ischemia-reperfusion injury due to the buildup of succinate [17]. During reperfusion, the resupply of oxygen and abundance of succinate causes high ROS production from RET that results in tissue damage, and treatments that decrease RET O 2 − release improve response to ischemia-reperfusion injury [18]. Supporting the contribution of CI ROS production to cellular damage in vivo, oocytes maintain cellular viability and fitness for decades by complete elimination of active complex I, which practically eliminates mitochondrial ROS production [36].
The relationship between CI ROS production and muscle stem cell differentiation is complex, and the role of CI ROS production in muscle differentiation and regeneration in vivo is unknown. Increased CI ROS production occurs during muscle myoblast differentiation [20]. − production and inhibiting the mitochondrial ROS production with treatment of mitochondrially targeted catalase inhibited muscle differentiation. Further experiments were able to delineate that CI RET specifically was responsible for the redox signaling driving muscle differentiation [23]. Conversely, increased ROS production can also impair differentiation [22]. It has been proposed that ROS production tightly controls muscle stem cell differentiation with a modest increase inducing differentiation but a large increase impairing differentiation (Graphical Abstract) [21]. Barium chloride (BaCl 2 ) is an established injury model for studying impaired regenerative capacity of aged muscle [37]. By using BI4500 and BaCl 2 injections, we were able to probe the role of I Q ROS in muscle regeneration of adult and old muscle in vivo. Interestingly, we found significantly higher numbers of CNFs within 5 days after injury in adult BI4500-treated animals (Figs. 5A and 6A). This suggests there is a role for mitochondrial ROS production in inhibiting the formation of centrally nucleated fibers. The increase in CNFs in adult BI4500-treated animals contributed to a full recovery in mean fiber CSA at 35 days post injury. However, this did not translate to a difference in tissue fibrosis or force recovery at 35 days post-injury.
Generally, there were no differences between old placebo and old BI treated animals for any metric of muscle regeneration except force recovery. Increased CNFs and CSA recovery observed in adult BI-treated animals were not observed in old BI-treated animals. These results do not support the hypothesis that inhibition of RET O 2 − release from mitochondrial site I Q would improve muscle recovery in aged muscle. Based on the working hypothesis that the level of induction for CI ROS tightly regulates muscle differentiation, it is possible that no effect of BI4500 treatment was observed in old muscle due to the lower agerelated capacity we observed to produce ROS from site I Q caused by lower CI-linked respiratory capacities (Figs. 2, 3B-C, 4A).

Limitations
We initially opted for chow delivery of the BI4500 to minimize the handling of the mice in this study. Despite the significant effects of this delivery on the target the plasma levels of the compound were below the level of quantitation and the single value for muscle compound concentration was below the target level. Because of this we changed to delivery by oral gavage for the muscle injury aspect of this study to achieve a higher peak circulating compound content. The once a day bolus dose instead of the more spread out dose available from chow feeding and the daily handling and stress that accompanies oral gavage could confound these results and limit our conclusions to oral gavage delivery method.
It is unclear why we did not see an increase in Pax7 positive nuclei in injured muscle at 5 days after injury given previous reports showing an increase in satellite cell number after injury [38]. One potential explanation is the timeline of the recovery relative to the severity of the injury. The BaCl 2 injury appears very severe in H&E cross sections, with very little muscle tissue present. Others using cardiotoxin as an injury model have shown less severe pathology at a similar timeframe of recovery [39]. The severity of the injury could be affecting the timeline of recovery, which could explain the lack of differences in Pax7 positive cells between injured limbs.

Conclusions
We have demonstrated that BI4500 is a specific inhibitor of complex I Q RET. CI RET is implicated in multiple disease and pathological conditions including ischemia-reperfusion. This compound provides another tool to dissect the importance of site-specific mitochondrial ROS production and may have potential as a therapeutic approach for preventing pathology associated with ischemia-reperfusion injury.

Funding
Funding was provided by Boehringer Ingelheim and the National Institute on Aging (NIA) grants (T32AG066574, F32AG074655, P01AG001751).

Declaration of competing interest
Antoni Filareto, Jens Markus Borghardt, and Michael Franti are employees of Boehringer Ingelheim. All other authors declare that they have no commercial or financial conflicts of interest. The authors worked with Boehringer Ingelheim on the design of this study. Otherwise, the funders played no role in the decision to publish this manuscript.

Data availability
Data will be made available on request.