Intermittent treatment with elamipretide preserves exercise tolerance in aged female mice

Abstract

The pathology of aging impacts multiple organ systems, including the kidney and skeletal and cardiac muscles. Long-term treatment with the mitochondrial-targeted peptide elamipretide has previously been shown to improve in vivo mitochondrial function in aged mice, which is associated with increased fatigue resistance and treadmill performance, improved cardiovascular diastolic function, and glomerular architecture of the kidney. However, elamipretide is a short tetrameric peptide that is not orally bioavailable, limiting its routes of administration. This study tested whether twice weekly intermittent injections of elamipretide could recapitulate the same functional improvements as continuous long-term infusion. We found that intermittent treatment with elamipretide for 8 months preserved exercise tolerance and left ventricular mass in mice with modest protection of diastolic function and skeletal muscle force production but did not affect kidney function as previously reported using continuous treatment.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11357-023-00754-0.

Introduction

Mitochondria play a key role in the biology of aging, and age-related mitochondrial dysfunction contributes to age-related cellular, tissue, and end-organ degeneration, thereby increasing the susceptibility to age-related chronic disease. This is particularly relevant in tissues with a high energetic demand, such as skeletal and cardiac muscles, where reduced function contributes significantly to the loss of mobility [1], exercise intolerance [2], and an associated decline in quality of life [3, 4]. As a result of the important role of mitochondrial-related energy production in aging heart and skeletal muscles, there is increasing interest in mitochondrial-targeted interventions that have the potential to reverse and slow the decline in sarcopenia, cardiac dysfunction, exercise intolerance, and reduced healthspan. Previous work has demonstrated that the mitochondrial-targeted peptide elamipretide (ELAM), previously referred to as Bendavia, MTP-131, and SS-31 is an effective intervention to reverse skeletal muscle [5, 6], cardiac [7], kidney [8], retina [9], and cognitive dysfunction in aged mice [10]. Clinical trials in aged humans demonstrated improvements to in vivo mitochondrial energetics after a single treatment [11]. This improvement was short-lived, consistent with the short half-life of the compound in rats [12].

In order to compare the ability of ELAM to slow aging or increase lifespan to other more established aging interventions, an extended duration of treatment is necessary. Aging remains the greatest risk factor for the development of multiple diseases, including heart disease [13, 14], cancer [15], and sarcopenia [16]. Therefore, slowing the aging process has the potential to improve public health, increase life expectancy, and expand the years over which people can expect to live independent and productive lives. There are two strategies for testing compounds that can improve function with age. The first is to test whether a compound is capable of reversing age-related decline when administered late in life after the effects of age are already apparent. The second strategy is to determine whether or not an intervention can slow the aging process when the agent is started early before significant dysfunction is present and then continued with advancing age.

Currently, ELAM is not available as an oral formulation because of its poor oral bioavailability. However, human clinical studies evaluating the effects of ELAM have used daily, subcutaneous dosing (NCT03323749). While this route of administration has been shown to be effective, we sought to investigate whether a regimen of 2×/week for 8 months would reproduce or improve the beneficial effects we have observed in the skeletal and cardiac function of aged mice with 8 weeks of continuous treatment.

Methods

Animals

All experiments performed in this study were reviewed and approved by the University of Washington Institutional Animal Care and Use Committee (IACUC). Female C57Bl/6J mice were procured from the National Institute on Aging aged mouse colony. Mice were kept at 21 °C on a 14/10 light-dark cycle and given chow and water ad lib.

Elamipretide administration

Animals were weighed once per week to determine injection volumes for the week. Mice were given a twice weekly dose on Monday and Thursday of either saline or 3 mg/kg of ELAM in saline via intraperitoneal injection for 8 months.

Treadmill performance

Mice were tested at baseline prior to initiation of twice-weekly injections and every 4 weeks thereafter. All mice were acclimated to the treadmill twice prior to each endurance test. During acclimation day 1, animals were loaded onto the treadmill at 0° incline with an active shockplate delivering 0.15 amps at 1 Hz. Mice were allowed to free roam for 2 min and then run for 2 min, accelerating from 0 meters(m)/minute(min) to 10 m/min, then held at a constant speed of 10 m/min. During acclimation day 2, animals underwent the same steps as day 1 with a maximum speed of 20 m/min. Finally, on test days, mice were placed on a treadmill at 10° incline and run for 5 min, accelerating from 0 m/min to 30 m/min, then run to exhaustion at 30 m/min. Exhaustion was defined as inability to remount the treadmill while receiving 5 consecutive shocks plus light physical prodding. Treadmill acclimation and endurance tests were all performed between 8 p.m. and 2 a.m. to align with the natural active period for these mice.

In vivo muscle mechanics and analyses

Mice were tested at baseline prior to initiation of 2× per week injections and at 8, 16, and 28 week during treatment. Animals were anesthetized using 2–4% isoflurane in 1 L/min O₂ and maintained on a heated platform at 37 °C using a thermally controlled water circulator. The right hind leg was held in position at the kneecap, and the foot was taped to a servomotor arm for torque force measurement and linear conversion (Aurora Scientific, Aurora, ON, Canada). Stimulation was controlled by and measured using Dynamic Muscle Control software v.5.500 (Aurora Scientific). For fatigue tests, the gastrocnemius (GAS) muscle was stimulated using a S88X dual channel stimulator (Grass Astro-Med Inc.) through the tibial nerve at 180 Hz every other second for 2 min followed by a 1- and 5-min post-fatigue tetanic test to evaluate recovery. Force and fatigue analysis was performed using Dynamic Muscle Analysis software v. 5.300 (Aurora Scientific).

Dissection and tissue partitioning

Animals were cervically dislocated and dissected for tissue processing. Part of the red gastrocnemius and the apex of the heart were immediately processed for skeletal and cardiac muscle respirometry and peroxide production, as described below. All other skeletal muscles were snap frozen in liquid N₂ for future biochemical assays. The remainder of the heart was snap frozen, and the kidneys were processed for histology and stoichiometry as described below.

Echocardiography

Echocardiography was performed as previously described [7]. In brief, animals were anesthetized using 0.5–1% isofluorane delivered via 1 L/min O₂. Imaging was performed using a Siemens Acuson CV-70 equipped with a 13 MHz probe. Results were analyzed using mixed-effects ANOVA with correction for multiple comparisons using Sidak’s post hoc test. Ea/Aa and Left Ventricular Mass (LVM) are directly measured values and LVMI is derived from LVM normalized to body surface area.

Skeletal and cardiac muscle respirometry and peroxide production

Immediately following cervical dislocation, GAS and heart were dissected and placed on ice. Approximately 2–3 mg of the red GAS was removed, fibers were gently separated with forceps, and permeabilized using 50 ug/mL saponin at 4 °C for 40 min. During permeabilization, approximately 10 mg of the apex of the heart was removed and homogenized. Permeabilized fibers and heart homogenate were added to a 2 mL chamber of an Oxygraph 2K dual respirometer/fluorometer (Oroboros Instruments, Innsbruck, Austria) at 37 °C and stirred at 750 rpm during titrations of substrates, uncouplers, and inhibitors. Following the addition of sample, superoxide dismutase (5 u/mL), amplex red (10 uM), and horseradish peroxidase (0.5 u/mL) were added to the chambers. For evaluation of forward electron flow, titrations were performed in the following order (concentrations): malate (5 mM), pyruvate (5 mM), and glutamate (10 mM); sub-saturating ADP (50 uM), saturating ADP (2.5 mM), succinate (20 mM), rotenone (0.5 uM), antimycin A (2.5 uM), TMPD (0.5 mM), ascorbate (2 mM), and KCN (1 mM). For evaluation of reverse electron flow, titrations were performed in the following order (concentrations): succinate (1, 5, 10, and 20 mM), and saturating ADP (2.5 mM).

Mitochondrial dynamics western bots

Snap frozen gastrocnemius was pulverized under liquid nitrogen with a mortar and pestle. Ground tissue was extracted on ice in CellLytic MT (Sigma C3228) with phosphatase and protease inhibitors using a bullet blender (Next Advance BBX24B) and centrifuged. Ten ug of protein extracts were separated using SDS-PAGE gels for each western blot (Criterion 3450034). Proteins were detected using anti-DRP1 (1:1000, Cell Signaling 8570); anti-FIS1 (1:1000, Abcam ab156865); anti-MFF (1:1000, 17090-1-AP); anti-MFN1 (1:1000, Abcam ab221661); anti-MFN2 (1:1000, Cell Signaling 9482); anti-Mid51 (1:1000, Thermo 20164-1-AP); anti-OPA1 (1:1000, Cell Signaling 80471). Protein sample homogenates used were consistent for each blot, and all labeled protein bands were normalized to total protein loaded as measured by Ponceau S staining.

Kidney histology and stoichiometry

At sacrifice the right kidney from each animal was decapsulated, bifurcated, and fixed at 4 °C in 4% paraformaldehyde for 16 h, followed by 70% ethanol before paraffin embedding. Immunohistochemistry for podocyte p57 expression (rabbit anti-p57; Santa Cruz Biotechnology, Santa Cruz, CA) was performed on 4 um sections using antigen-retrieval EDTA buffer pH 6 and overnight incubation at 1:800. Antibody was visualized with diaminobenzidine (Sigma) and counterstained with periodic acid Schiff stain (PAS), including hematoxylin to visualize nuclei. At least 40 glomeruli from the outer cortex and 20 glomeruli from the juxtamedullary cortex were scored from each kidney (n = 5 ELAM treated, n = 4 control) for the presence of sclerosis on a scale of 0 (no injury); 1 (minor structural changes or matrix deposition); 2–3 (25–75% sclerotic); 4 (fully sclerosed or necrotic). Podocytes were measured in micrographs using ImageJ software and the Venkatereddy–Wiggins method [17] to calculate the number, diameter, and density within the tuft for each glomerulus (n = 3 ELAM treated, n = 2 control). Individual student’s t-tests showed no significant differences between ELAM and control for any kidney parameter.

Results

Body and tissue mass

Starting weights for the saline and ELAM-treated groups were not significantly different (28.3 ± 1.7 g vs. 28.9 ± 2.2 g). Two times per week treatment with ELAM starting at 20 months of age had no effect on body mass throughout the 8-month treatment period, analyzed by area under the curve (Fig. 1A), nor was there a difference in end-of-treatment weights. There was no effect of treatment on survival, with 8 saline mice and 9 ELAM mice dying of natural causes over the treatment period (Fig. 1B). Due to the small sample sizes, it was not possible to make any statements regarding differences in the cause of death. Consistent with no differences in body mass, there was also no effect of treatment on the skeletal muscle mass of the GAS and the tibialis anterior (TA) muscles and no effect on cardiac hypertrophy (Fig. 1C–E).

Fig. 1.

Mass and survival. A Weekly masses over the course of treatment normalized to initial body masses N = 20 at baseline and N = 8–10 at endpoint. B Survival curves of animals that did not reach endpoint. C Tibialis anterior mass normalized to tibia length N = 11–12. D Gastrocnemius mass normalized to tibia length N = 11–12. E Heart mass normalized to tibia length N = 9–10. All data expressed as means ± SE

Exercise tolerance

To assess the effect of treatment and age on whole-body performance, we tested exercise tolerance by running mice to exhaustion on a treadmill every 4 weeks. There was no difference in running time at baseline between the groups. With a mixed-effect ANOVA analysis, there was a highly significant effect of both age and treatment on total running time. The improvement in exercise tolerance was apparent after the first 4 weeks of treatment with ELAM and persisted throughout the 28-week treatment period (Fig. 2).

Fig. 2.

Treadmill running times. Elamipretide significantly preserves treadmill endurance. Data expressed as means ± SE. Mixed effects ANOVA, with Sidak’s correction for multiple comparisons, p < 0.05. N = 20 at baseline; N = 10–12 at endpoint

Skeletal muscle and cardiac function

To assess the effect of aging and treatment on skeletal muscle function, we measured the maximum force-time integral and fatigue resistance in the GAS muscle at 8, 16, and 28 weeks throughout the treatment period. Despite the improved treadmill exercise tolerance in the ELAM group and previous work demonstrating increased fatigue resistance after continuous 8-week treatment [6], there was no effect of treatment on GAS fatigue resistance nor force recovery following fatigue at any time point (Fig. 3A–D). We evaluated total force generation as the total force generated over one maximum tetanic stimulus, otherwise known as the maximum force time integral (FTI). There was no significant effect of age on the maximum FTI of the GAS throughout the study period (p = 0.07) and no significant interaction of age with treatment. Although the effect was not significant with a mixed model ANOVA, multiple testing revealed a modest improvement in maximum FTI at week 16 only (P = 0.05) in the ELAM group (Fig. 3E). There was also no effect of treatment on absolute peak force or peak force normalized to body weight (Supplemental Fig. 1A and B).

Fig. 3.

In vivo skeletal muscle function. A Force time integral (FTI) of fatigue normalized to resting force at baseline before entering study, N = 19–20 (B) at 8 weeks, N = 16–18 (C) at 16 weeks, N = 15–17 (D) and at endpoint of study, N = 13 (E) Maximum force time integral over the course of the study. Data expressed as means ± SE. Mixed effects ANOVA, with Sidak’s correction for multiple comparisons, p < 0.05

To assess whether the 28-week intermittent treatment reproduces the reversal of cardiac dysfunction, we performed echocardiography to measure Ea/Aa ratio and left ventricular mass index LVMI on the mice at baseline, 16, 24, and 28 weeks. We have previously reported that ELAM treatment reverses diastolic dysfunction in aged mice [7]. Similar to the maximum FTI analysis, the treatment effect was not significantly different with a mixed model ANOVA (Fig. 4A) p = 0.07) and three was no significant interaction between age and treatment. However, multiple testing suggested preservation of cardiac function manifested in a significant difference between the control and ELAM groups in the Ea/Aa ratio at 28 weeks (P = 0.05). In contrast, the LVMI indicated a significant effect of treatment overall with no significant effect at any one timepoint p < 0.01 (Fig. 4B).

Fig. 4.

Cardiac effects. A Ea/Aa ratio over the course of the study, N = 7–12. B Left ventricular mass index over the course of the study, N = 7–12. Data expressed as means ± SE. Mixed effects ANOVA, with Sidak’s correction for multiple comparisons, p < 0.05

Mitochondrial function

To test the effect of long-term intermittent treatment of ELAM on mitochondrial function and mitochondrial peroxide production, we analyzed permeabilized muscle fibers from the red portion of the gastrocnemius and homogenate taken from the apex of the heart using a combination fluorometer/respirometer (Oroboros Instruments, Innsbruck, AU) at the end of 28 weeks. We tested forward electron flow and reverse electron flow to assess maximal mitochondrial respiration as well as peroxide production. There was no significant effect of treatment on maximum mitochondrial respiration in either permeabilized fibers or heart homogenate (Fig. 5A and B). However, treatment with ELAM did decrease amplex red signal in permeabilized GAS fibers in forward electron flow but not in heart homogenate (Fig. 5C and D). There was no effect on respiration in either tissue during reverse electron flow (Supplemental Fig. 2A and B), but there was a significant decrease in amplex red signal produced by permeabilized GAS fibers in reverse electron flow using mixed-effect ANOVA (p < 0.002) but not in heart (Supplemental Fig. 2C and D). It should be noted that the total amplex red signal includes not only reactive oxygen species generated by the electron transport chain but also a large portion by lipid peroxides [18]. To evaluate mitochondrial fusion and fission, we measured protein expression of dynamin-related protein 1 (DRP1), mitochondrial fission 1 (Fis1), mitochondrial fission factor (MFF), mitofusin 1 (MFN1), mitofusin 2 (MFN2), mitochondrial dynamin-like GTPase (OPA1), and mitochondrial dynamics protein Mid51 (Mid51) in gastrocnemius muscle collected at endpoint. Although MFN1, OPA1, and Mid51 show a trend toward decrease by treatment with ELAM, this was largely driven by increased labeling and variation in some saline-treated samples, but there was no significant difference in any of the measured proteins (Supplemental Fig. 3).

Fig. 5.

Mitochondrial respiration and amplex red signal. A Respiration in permeabilized gastrocnemius fibers in forward flow, N = 8–10. B Respiration in heart homogenate in forward flow, N = 6–8. C Peroxide production in permeabilized gastrocnemius fibers in forward flow, N = 8–10. D Peroxide production in heart homogenate in forward flow, N = 6–8. Data expressed as means ± SE. Mixed effects ANOVA, with Sidak’s correction for multiple comparisons, p < 0.05

Kidney pathology

Kidneys are susceptible to predictable age-accumulated injury in both humans and rodents [19]. With age, kidney function declines, and this is accompanied by pathological changes to the microanatomy, most notably focal accumulation of sclerotic lesions in the glomeruli [20, 21] and nephron loss [22]. Loss of specialized glomerular epithelial cells, podocytes, are thought to precede and instigate much of the glomerular dysfunction accumulated with age [23, 24]. Podocytes are non-proliferative cells that are required for the formation of the kidney filtration slit and, thus, are paramount to kidney function. Previously, we demonstrated that 8 weeks of continuous systemic ELAM treatment in male mice treated from 24 to 26 months of age was beneficial to the kidney by significantly reducing the accumulation of age-induced glomerulosclerosis [8]. Additionally, some standard measures of podocyte damage were significantly reduced, suggesting cell-specific protection. In the same study, 28-month-old female mice treated with ELAM for the same 8-week duration exhibited reduced cellular senescence in all kidney compartments. In that study, female kidneys were not also assessed for glomerulosclerosis; therefore, we sought to determine whether intermittent extended treatment could similarly protect aged female mice from age-accumulated sclerosis. Kidneys were harvested at sacrifice and stained to visualize and quantify microanatomy. Although indicators of kidney aging were observed in all mice, scoring for glomerular injury did not demonstrate any differences between intermittent ELAM injection and untreated female mice (Fig. 6A and B). Further analysis of podocytes for cell number (podocytes per glomerular tuft), cellular hypertrophy (nuclear diameter), cell density (cell number per glomerular tuft area), and glomerular hypertrophy showed no response to ELAM as compared to control animals (Supplemental Fig. 2C–F). Thus, kidney improvements, which were detected in male mice with 8-week continuous ELAM intervention, were not observed with intermittent intervention in female mice.

Fig. 6.

Kidney pathology. A Kidney glomeruli were scored for aging glomerulosclerosis. Because glomerular size differs by kidney region, glomeruli were separated for scoring by cortical location as either occupying the outer cortex or near the medulla (juxtamedullary/JM). B Example images of immunohistochemistry of p57 to identify podocyte nuclei (dark brown, DAB) and counter stained with periodic acid Schiff (PAS) and hematoxylin (blue nuclei). Relative scores, for example, images shown. C Average number of podocytes per individual glomerular tuft. D Average diameter of podocyte nuclei within all glomeruli analyzed within each mouse. E Average volume of glomerular tuft. F Average density of podocytes within all glomeruli scored for each compartment was calculated as the number of podocytes normalized to tuft volume. All control and ELAM-treated groups were analyzed by individual Student’s t-test, and no significant differences were detected. N = 4–6

Discussion

This study was designed as a pilot to test whether long-term, intermittent treatment with the mitochondrial-targeted peptide ELAM would slow the decline in skeletal muscle and cardiac function when started early in the aging process. We have previously demonstrated that continuous infusion with ELAM using osmotic minipumps for 8 weeks improves skeletal muscle fatigue resistance and in vivo mitochondrial function [6], reverses cardiac diastolic dysfunction [7], and improves tolerance in mouse models of aging when administered late in life [5]. Furthermore, ELAM has also been demonstrated to be effective in multiple models of chronic disease [8, 10, 25], making it a promising mitochondrial-targeted intervention for slowing the aging process and reducing frailty. Due to its peptide structure, ELAM is rapidly degraded in the gut environment. ELAM has a short elimination half-life of 2–4 h [12], so previous preclinical work with ELAM has focused primarily on daily administration [5, 9] or continuous delivery by osmotic minipumps [6–8]. These approaches are limiting for long-term studies due to the extensive handling of the animals necessary for daily injections, eye drops administration [9], or the multiple surgeries that are required to remove and replace osmotic pumps to maintain delivery for several months. This limitation makes it difficult to compare to other aging interventions, such as those in the National Institute on Aging Intervention Testing Program (ITP). Recent evidence in older humans demonstrated that a single treatment with ELAM improved in vivo mitochondrial function immediately after treatment and returned to placebo level a week later [11]. However, this study suggested that effects on skeletal muscle fatigue resistance were elevated a week after treatment. In an attempt to test whether we could reduce the necessary animal handling required for a long-term study, we tested whether twice-a-week treatment with ELAM starting early in life would be sufficient to reproduce or improve our previous results with 8-week treatment on aging cardiac and skeletal muscle function.

Intermittent treatment with ELAM preserved exercise tolerance and delayed decline in skeletal muscle force and cardiac function, but did not lead to improvements in skeletal muscle fatiguability or cardiac function as demonstrated with 8 weeks of continuous delivery [6, 7]. However, the preserved exercise tolerance observed starting at 4 weeks (the first time point after baseline) is consistent with the rapid effects of ELAM treatment on mitochondrial function and whole-body performance previously demonstrated within 7 days of daily treatment [5]. Despite the improved systemic effect on treadmill running, the intermittent treatment did not appear to prevent the loss of body or skeletal muscle mass and morbidity over this age range.

Treadmill exercise tolerance is a complex phenotype involving aspects of motivation as well as physiological capacity. The main physiological capacities that contribute to exercise tolerance are cardiac function and muscle fatigue resistance [26–28], both of which are improved with continuous ELAM treatment [6, 7]. The absence of an effect of ELAM on muscle fatigue resistance of the GAS was surprising given our previous work, and the modest improvement in muscle force production is consistent with other studies. Continuous pump treatment for 8 weeks and 4 months of daily injections, both starting at 24 months in C57Bl/6 mice, failed to significantly increase muscle force production, despite improvement in mitochondrial function or redox stress. It is now clear that the loss of muscle mass and force production is initiated by the decline in neuromuscular junctions (NMJ) well before significant muscle deficits are observed [29]. The disruption of NMJ leads to elevated mitochondrial oxidative stress in the myofiber, which leads to a feedforward cycle causing further damage to the NMJ [30, 31]. Although the ability of ELAM to reduce mitochondrial oxidative stress is well established, the twice-per-week treatment used in the current study, even when initiated early in the process, does not appear to be sufficient to prevent this cycle.

The primary aging cardiac phenotypes in the C57Bl/6 mice are diastolic dysfunction and cardiac hypertrophy. Previous work demonstrated that ELAM reversed diastolic dysfunction as assessed by the early (E) to late (A) ventricular filling, otherwise known as the E′/A′ ratio, with 8 weeks of continuous treatment [7], and this effect persisted for two weeks following the end of treatment [7]. In contrast, the intermittent treatment only demonstrated modest preservation of diastolic dysfunction at 28 weeks of treatment. There are two potential explanations for the greater length of treatment required in this study. The first is that the intermittent treatment required a longer period to drive the cardiac remodeling underlying improved cardiac relaxation and the persistent effect of treatment [7]. The alternative is that there was no significant cardiac dysfunction present until later in the treatment period. Multiple studies have demonstrated that ELAM has no effect on a healthy, well-functioning heart. This interpretation is supported by the stable E′/A′ ratio over the first 24 weeks of this study, with values similar to what has been reported for young C57Bl/6 mice. Left ventricular mass index is used as a parameter to indicate cardiac hypertrophy. We found that the ability of intermittent treatment to contribute to partial remodeling of the aged heart is supported by the decreased LVMI, although this did not translate to reduced cardiac hypertrophy based on total heart mass.

Conclusions

This pilot study demonstrated that, although twice weekly treatment with ELAM for 8 months led to the preservation of whole-body exercise tolerance and partial improvement of cardiac aging phenotypes, it only reproduced a portion of the significant benefits previously observed on skeletal and cardiac function with continuous 8-week treatment at a late age. Thus, continuous ELAM treatment (daily injection or osmotic pump delivery) remains the most effective delivery method to observe long-term benefits of ELAM on health span in aging mice.

Abbreviations

GAS

Gastrocnemius

TA

Tibialis anterior

ELAM

Elamipretide

Funding

This work was supported by the National Institute of Health grants P01 AG001751 and T32 AG000057, the University of Washington Nathan Shock Center P30 AA013280, and the University of Washington Center for Translational Muscle Research P30 AR074990.

Declarations

Conflict of interest

The authors declare no competing interests.