Increased skeletal muscle mitochondrial free radical production in peripheral arterial disease despite preserved mitochondrial respiratory capacity

Subjects

A total of 22 subjects, 12 HC and 10 patients with mild PAD, classified as Fontaine Stage II (Norgren et al., 2007), were recruited to participate in this study. Inclusion criteria for patients with PAD were a clinical diagnosis with evidence of femoropopliteal occlusive disease (ABI ≤ 0.90), intermittent claudication, and no prior PAD-related surgical interventions. Arterial lesions in the patients with PAD were localized to the distal femoral and/or popliteal arteries and characterized by moderate calcification. All HC were normotensive (<140/90 mmHg), not taking any medication recognized to alter blood flow or metabolism, and no evidence of overt cardiovascular disease. All subjects were recruited based upon being moderately physically active (assessed by both interview and accelerometry). Females were postmenopausal and not taking any form of estrogen-replacement therapy, minimizing the hormonal influences. Subjects reported to the laboratory for testing in a fasted state (>8 h postprandial) and refrained from caffeine or strenuous exercise prior to the studies (>24 h). All subjects performed a 10-m walk test as a clinical assessment of overall physical function during the initial screening visit, as previously described (McCully, Leiper, Sanders, & Griffin, 1999; Peters, Fritz, & Krotish, 2013).

Muscle Biopsy

A percutaneous needle biopsy was obtained from the anteromedial aspect of the gastrocnemius muscle belly, ~10 cm distal to the tibial tuberosity at a depth of ~1.0 cm under sterile conditions. In the patients with PAD, biopsies were performed on the most symptomatic leg, whereas biopsies in the HC were performed on the right leg. Immediately following removal of the muscle sample (~150 mg) from the leg, ~20% of the sample was immersed in ice-cold biopsy preservation fluid (BIOPS) for respiratory analysis (Park et al., 2014), and the remaining muscle sample was snap frozen and stored at −80°C for histological and biochemical analyses, as well as electron paramagnetic resonance (EPR) spectroscopy.

Mitochondrial Respiration

Muscle samples were prepared and permeabilized, as described by Park et al. (Park et al., 2014). Briefly, BIOPS-immersed fibers were carefully separated with fine-tip forceps and subsequently bathed in a BIOPS-based saponin solution (50μg saponin/ml BIOPS) for 30 minutes. Following saponin treatment, muscle fibers were rinsed twice in ice-cold mitochondrial respiration fluid (MIR05) for 10 minutes each rinse. The wet weight of the muscle sample (~3–4 mg) was then assessed on a calibrated scale.

The muscle fibers were then placed in two temperature-controlled respiration chambers (Oxytherm, Hansatech Instruments. Norfolk, UK) in 2 ml of MIR05 solution and warmed to 37°C, allowing duplicate measurements. After allowing the muscle 10 minutes to equilibrate, State 2 respiration was assessed with malate (2 mM) + glutamate (10 mM), followed by State 3 complex I (State 3:CI) respiration with ADP (5 mM) (State 3 CI). Subsequently, succinate (10 mM) was added to assess State 3 complex I and II glycolytic oxidation rate (State 3 S CI+II) followed by octanoyl carnitine (1.5 mM) to assess State 3:CI+CII fatty acid oxidation rate (State 3 O CI+II). Next, cytochrome C (10 μM) was added to the bath to verify membrane integrity, as recommended by Pesta et al. (Pesta & Gnaiger, 2012). Finally, rotenone (0.5mM) was added to inhibit CI for the assessment of State3:CII respiration (State 3 CII). Step changes in respiration lasted as long as required to produce a steady state respiration rate, typically ~3 minutes in duration. Background respiration inherent to the experimental setup was measured and taken into account when assessing overall mitochondrial respiration. The respiration data for each of the two separately assessed muscle fiber samples was then averaged. The rate of O₂ consumption was measured as picomoles of O₂ per second and then expressed relative to muscle sample mass (pmol·s⁻¹·mg wet wt⁻¹). Control ratios, including the respiratory control ratio (RCR), and substrate control ratios for the addition of octanoyl carnitine (SCR_O) and succinate (SCR_S), were calculated from the respiratory flux measurements, as described previously (Kraunsoe et al., 2010).

CS activity

Citrate synthase (CS) activity was assessed as an indicator of mitochondrial content and volume (Larsen et al., 2012; Meinild Lundby et al., 2018; Pipinos, Swanson, et al., 2008). After the respiration measurements, the same muscle samples (~3–4 mg wet weight) were homogenized in a buffer containing 250 mM sucrose, 40 mM KCl, 2 mM EGTA, and 20 mM Tris·HCl (Qiagen, Hilden, Germany) and a subsequent CS activity assay was performed using a spectrophotometer (Biotek Instruments, Winooski, VT), as previously described (Picard et al., 2008).

Muscle fiber type

Immunofluorescence was used to determine muscle fiber type (Myosin Heavy Chain I: primary antibody BA-F8, secondary antibody Alexa Fluor 350; Myosin Heavy Chain IIa/IIx: primary antibody SC-71, secondary antibody Alexa Fluor 488, Abcam, Cambridge, MA) and capillarity (ab96884, Abcam, Cambridge, MA) with an Imager A2 microscope with Axiocam and accompanying software (Zeiss, Jena, Germany), as described in detail by Bloemberg and Quadrilatero (Bloemberg & Quadrilatero, 2012). Primary antibodies used for fiber typing were from Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, IA). Secondary antibodies were from Invitrogen (Thermo Fisher Scientific, Waltham, MA).

Free radicals, inflammation, and oxidative stress

Free radical measurements were assessed by EPR spectroscopy on muscle tissue and whole blood samples using an EMX X-band spectrometer (Bruker, MA). The frozen muscle tissue samples were placed in a micro-centrifuge tube containing 150 μL of the spin trap 1-hydroxy-4-[2-(triphenylphosphonio)-acetamido]-2,2,6,6-tetramethylpiperidine (mitoTempo-H) (0.5 mmol/L) (Enzo Life Sciences San Diego, CA) for 60 minutes at 37° C. After 60 minutes, the samples were placed on ice and 50 μL of the solution was loaded into a capillary tube for EPR spectroscopy analysis. The EPR spectroscopy scan was run with a center field at approximately g = 2.004 and the area under the curve of the EPR spectroscopy signal was calculated by double integration. Whole blood samples were collected and spin-trapped (Richardson et al., 2007). Briefly, 1.5 ml of venous blood was collected into a vacutainer containing 0.5 ml of the spin trap α-phenyl-N-tert-butylnitrone (PBN) (0.0140 mol/L). After centrifugation, the PBN adduct (300 μl) was pipetted into a precision-bore quartz EPR sample tube (Wilmad, Vineland, NJ). EPR spectroscopy was then performed at 21°C. The area under the curve of the EPR spectroscopy signal was calculated by double integration.

Venous blood samples, for the analysis of blood born markers of inflammation and oxidative stress, were centrifuged to allow the collection of the plasma and stored at −80°C until analysis. Concentrations of the pro-inflammatory cytokines C-reactive protein (CRP), TNFα, IL-6, and the anti-inflammatory cytokine IL-10, were determined by solid phase sandwich ELISA (R&D Systems, Minneapolis, MN, USA). Superoxide dismutase (SOD) activity was assessed as previously described (Wheeler, Salzman, Elsayed, Omaye, & Korte, 1990). Colorimetric ELISA assays were used to detect lipid peroxidation by plasma malondialdehyde (MDA) levels (Bioxytech LPO-586, Foster City, CA) and protein oxidation by protein carbonyl (PC) levels (R&D Systems, Minneapolis, MN).

To analyze skeletal muscle protein oxidation, tissue PC content was measured after treatment with 2.4-dinitrophenylhydraziine (DNPH), which forms labeled protein hydrazine derivatives (Mecocci et al., 1999). Briefly, two 200 μl samples were transferred to two 2 ml plastic tubes, a sample tube and a control tube. 800 μl of DNPH was then added to the sample tube and 800 μl of 2.5M hydrochloride was added to the control tube. Both the control and sample tube were incubated in the dark at room temperature for one hour and vortexed every 15 min. 1 ml of 20% and 10% trichloroacetic acid (TCA) was added and afterwards both tubes were centrifuged at 10,000 g for 10 min at 4°C. The pellets formed were then washed three times with 1 ml of ethanol/ethyl acetate mixture (1:1). The final pellets were dissolved in 500 μl of guanidine hydrochloride by vortexing and the solution was centrifuged at 10,000 g for an additional 10 min. PC content was determined spectrophotometrically at 370 nm and calculated using a molar absorption coefficient of 22,000 M⁻¹ cm⁻¹ (Pansarasa, Bertorelli, Vecchiet, Felzani, & Marzatico, 1999).

The relative abundance of the target proteins for IL-6 and 4-HNE was determined in skeletal muscle samples by immunoblotting. Briefly, tissue samples were homogenized 1:12 (wt/vol) in an ice-cold buffer containing 5 mM Tris (pH 7.5) and 5 mM EDTA (pH 8.0) with a protease inhibitor cocktail (1). Homogenates were centrifuged at 1500 g for 10 min at 4°C. After centrifugation, the supernatant was collected and the protein concentration was determined using the Bradford technique. Proteins from the supernatant fraction were separated via polyacrylamide gel electrophoresis, transferred onto a polyvinylidene difluoride membrane, and incubated with primary and secondary antibodies specific to the proteins of interest (Kwon, Tanner, et al.). Membranes were exposed on a ChemiDoc XRS (Bio-Rad) and quantified with Image Lab software (Bio-Rad). The specific antibodies used to detect skeletal muscle proteins IL-6 and 4-HNE included: Anti-Interleukin 6 (IL6) (ab6672, Abcam, Cambridge, MA) and 4-hydroxynonenal (4-HNE) (ab46545, Abcam, Cambridge, MA). The protein abundance of each protein was normalized to beta-actin (ab8227, Abcam, Cambridge, MA), which served as a loading control (Kwon, Smuder, et al.).

Lower Leg Volume and Muscle Mass

Lower leg volume was calculated based on lower leg circumference (three sites: distal, middle, and proximal), lower leg length, and skinfold measurements, with muscle mass then being calculated based upon the expected volume ratio of the major muscles of the lower leg (Jones & Pearson, 1969). This method has previously been confirmed to provide a valid estimate for muscle mass across a spectrum of individuals with normal muscle mass and severe muscle atrophy (Layec, Venturelli, Jeong, & Richardson, 2014).

Mitochondrial function

Mitochondrial respiration in the HC and patients with PAD is displayed in Figure 1. There were no differences between groups for State 2 (HC: 8 ± 1; PAD 8 ± 2 pMol.mg.s⁻¹), State 3CI (HC: 18 ± 1; PAD 20 ± 1 pMol.mg.s⁻¹), State 3:CII (HC: 16 ± 1; PAD 16 ± 1 pMol.mg.s⁻¹), or State 3:CI+II respiration using either octanoyl carnitine (HC: 19 ± 1; PAD 21 ± 2 pMol.mg.s⁻¹) or succinate (HC: 25 ± 2; PAD 26 ± 2 pMol.mg.s⁻¹). Furthermore, no differences between the HC and patients with PAD were observed in CS activity (HC: 38 ± 6; PAD: 40 ± 7 nmol.min.mg tissue⁻¹) (p = 0.84). The RCR (HC: 4.7 ± 1.9; PAD: 4.7 ± 1.9), SCR_S (HC: 1.3 ± 0.2; PAD: 1.3 ± 0.3), and SCR_O (HC: 1.1 ± 0.3; PAD: 1.1 ± 0.3) were not different between groups.

Open in a separate window

Figure 1:

Mitochondrial respiration in healthy controls (HC) and patients with peripheral arterial disease (PAD).

CI, complex I; CII, complex II; O, octanoyl carnitine; S, succinate. Data presented as mean ± SD.

Inflammation, free radicals, and oxidative stress

Plasma markers of inflammation, free radical production, and oxidative stress are displayed in Table 4. Patients with PAD exhibited significantly greater plasma concentrations of TNF-α and IL-6 compared to the HC (p < 0.05). There were no significant differences between groups for the EPR spectroscopy detection of ROS in whole blood or plasma concentrations of CRP, IL-10, SOD, MDA, or PC. Patients with PAD had significantly increased levels of intramuscular pro-inflammatory cytokine IL-6 (~75% greater) (Figure 2A) and the intramuscular lipid peroxidation marker 4-HNE (~94% greater) (Figure 2B), expressed as a percentage of the HC (p < 0.05). Intramuscular PC were significantly elevated in patients with PAD compared to the HC (HC: 2.12 ± 0.10; PAD: 3.44 ± 0.15 nmol/mg protein, p < 0.05) (Figure 2C). In addition, the EPR spectroscopy signal representative of intramuscular, mitochondrial-derived ROS production was significantly greater in patients with PAD compared to the HC (HC: 1.00 ± 0.36; PAD 4.31 ± 0.97 AU per mg tissue, p < 0.05) (Figure 2D).

Open in a separate window

Figure 2:

The presence of intramuscular inflammatory marker IL-6 (panel A), oxidative stress markers 4-HNE and protein carbonyls (panels B and C, respectively), and mitochondrial ROS production (panel D) in healthy controls (HC) and patients with peripheral arterial disease (PAD).

Representative Western Blot images are presented for IL-6 and 4-HNE (panels A and B, respectively). Panel A, HC (n = 9) and PAD (n = 7); Panels B and C, HC (n = 9) and PAD (n = 7); Panel D, HC (n = 12) and PAD (n = 10). Data presented as mean ± SEM. *Significant difference between groups, (p < 0.05).

Mitochondrial respiratory function in patients with PAD

Mitochondrial DNA damage has been detected in patients with PAD (Brass & Hiatt, 2000), and is proposed to reflect systemic inflammation and oxidative stress, however, conclusions from prior studies regarding the impact of PAD on mitochondrial function have been equivocal. For example, Hou and colleagues (Hou et al., 2002) reported preserved peak mitochondrial ATP production rates in isolated mitochondria from patients with PAD with similar average ABIs (~0.67) as the patients assessed in the current study (Table 1). However, evidence suggests that the mitochondrial isolation procedure itself may selectively damage the organelles of interest, potentially confounding the assessment of mitochondrial function (Picard et al., 2011). More recently, Lindegaard-Pedersen and colleagues (Lindegaard Pedersen, Baekgaard, & Quistorff, 2017) found no differences in mitochondrial respiratory function assessed in permeabilized skeletal muscle fibers, a method documented to better represent in vivo physiology, between healthy controls and patients with PAD, with the exception of increased sensitivity to the substrate succinate in the patients with PAD. In contrast, Pipinos and colleagues (Pipinos et al., 2006; Pipinos et al., 2003) reported attenuated mitochondrial respiratory capacity from patients with PAD in permeabilized skeletal muscle fibers. However, it should be noted that the subjects had advanced PAD, critical limb ischemia, an average ABI of ~0.37, pain at rest, and a loss of muscle mass in the affected limb of the majority of patients. In fact, as nearly half of the patients in the study of Pipinos and colleagues (Pipinos et al., 2003) were undergoing surgical amputations of the symptomatic limb, the attenuated mitochondrial respiratory capacity may have been a consequence of end-stage sarcopenia and tissue necrosis (Anderson et al., 2009), rather than PAD per se.

Using a similar approach to Pipinos and colleagues (Pipinos et al., 2006; Pipinos et al., 2003), this study revealed preserved mitochondrial respiratory function in permeabilized muscle fibers from patients with PAD compared to reasonably well-matched HC (Figure 1). While these findings conflict with those of Pipinos and colleagues (Pipinos et al., 2006; Pipinos et al., 2003), as already eluded to, the disparity may be attributed to differences in disease severity. Specifically, the patients in the present study were classified as Fontaine stage IIa based on symptoms of intermittent claudication that appeared after walking >200 meters, an average ABI of 0.67 ± 0.10 (Table 1), and the absence of pain at rest (Gardner & Afaq, 2008; Norgren et al., 2007). Using a similar cohort of patients with PAD, Lindegaard-Pedersen and colleagues (Lindegaard Pedersen et al., 2017) reported no differences in State 3 and 4 mitochondrial respiration between healthy controls and patients with PAD. However, the same study reported significantly lower mitochondrial respiratory function in a group of patients with PAD combined with the diagnosis of type 2 diabetes, suggesting the potential role of insulin resistance in impairing compensatory mechanisms that preserve skeletal muscle function in patients with early stage PAD. Thus, the present findings demonstrate that mitochondrial respiratory function, assessed in permeabilized skeletal muscle fibers, is not an obligatory accompaniment to the early stages of PAD.

Differences in physical activity may also contribute to the previously documented decline in mitochondrial function in patients with PAD compared to their healthy counterparts. In this study, participants were carefully matched for physical activity using both accelerometry (Table 1) and questionnaires. Prior studies have associated a decline in mitochondrial function with walking performance in older adults (Coen et al., 2013; Santanasto et al., 2016) and in patients with PAD (Hou et al., 2002), while also demonstrating the beneficial effects of exercise to off-set this age-related decline in mitochondrial function (Joseph, Adhihetty, & Leeuwenburgh, 2015). Furthermore, the efficacy of physical activity to preserve mitochondrial function with age would most likely translate into preserved mitochondrial function with maintained physical activity in patients with PAD. Indeed, prior evidence strongly supports a link between the benefits of exercise therapy, which enhances mitochondrial synthesis and metabolic responses, and quality of life, which, likely in combination, ultimately increases physical activity in patients with PAD (Haas, Lloyd, Yang, & Terjung, 2012). Thus, based on the present findings, it appears that physical activity may play an important role in the preservation of mitochondrial respiratory capacity in the patients with PAD evaluated in this study.

Inflammation in the pathogenesis of PAD

Emerging evidence suggests that elevated systemic inflammation and oxidative stress collectively contribute to muscle dysfunction and sarcopenia in patients with PAD (Ha et al., 2016; Koutakis et al., 2015; Pipinos et al., 2006; Pipinos et al., 2003; Pipinos, Swanson, et al., 2008; Signorelli et al., 2012; Signorelli et al., 2014; Signorelli et al., 2016; Signorelli et al., 2003; Weiss et al., 2013). Furthermore, similar mechanisms involving inflammation, elevated ROS production, and oxidative stress contribute to the vascular sequela associated with impaired macro and microvascular dysfunction in atherosclerosis (Steven, Daiber, Dopheide, Münzel, & Espinola-Klein, 2017). Indeed, the pathogenesis of PAD appears to begin with chronic, repeated cycles of ischemia-reperfusion induced during skeletal muscle contraction as blood flows through an arterial stenosis, eliciting an exaggerated inflammatory response and systemic free radical production (Pipinos, Judge, et al., 2008). Elevations in free radicals trigger numerous pathophysiological pathways that result in systemic oxidative stress, leading to the clinical manifestations of muscle dysfunction and sarcopenia associated with PAD (Hiatt, Armstrong, Larson, & Brass, 2015). Although it is becoming increasingly apparent that the vasculature is a significant source of free radicals (e.g. NADPH oxidase) (Berry et al., 2000; Pipinos, Judge, et al., 2008; Walker et al., 2016), many have proposed that damaged skeletal muscle mitochondria may be the primary source of ROS in PAD (Pipinos, Judge, et al., 2008; Pipinos et al., 2006; Pipinos, Swanson, et al., 2008). As mitochondrial dysfunction may limit muscle function and lead to loss of muscle mass in patients with PAD, understanding the pathophysiological consequences of inflammation, mitochondrial-derived ROS production, and oxidative stress is clinically important.

In accordance with the proposed inflammation and oxidative stress-mediated pathway for the pathogenesis of PAD, this study revealed evidence of elevated systemic inflammation, as evidenced by ~2-fold greater plasma concentrations of TNF-α and IL-6 (Table 4) and a ~75% increase in intramuscular levels of IL-6 (Figure 2A) in patients with PAD compared to the HC. In accordance with these findings, Signorelli and colleagues (Signorelli et al., 2012; Signorelli et al., 2003) have documented elevations in plasma concentrations of TNF-α and IL-6 in patients with PAD at rest and following treadmill walking, although the influence on mitochondrial function and oxidative stress was not assessed. Using magnetic resonance spectroscopy to assess mitochondrial function in vivo, Tecilazich and colleagues (Tecilazich et al., 2013) reported slower phosphocreatine recovery kinetics associated with an increased pro-inflammatory state in diabetics with PAD. However, this impairment in mitochondrial function was evident in the diabetics whether or not they had PAD. To our knowledge, the present study is the first to document an increased pro-inflammatory state within both plasma and skeletal muscle of patients with mild PAD, supporting a systemic inflammatory profile associated with the pathophysiology of the disease.

Mitochondrial-derived ROS and oxidative stress in the pathogenesis of PAD

The heightened inflammatory profile observed in patients with PAD was accompanied by a ~4-fold greater mitochondrial-derived ROS production and a nearly two-fold increase in intramuscular oxidative stress compared to the HC, as evidence by increased lipid peroxidation (4-HNE) and protein oxidation (PC) (Figure 2). To our knowledge, the present study is the first to provide direct evidence of mitochondrial-derived ROS in patients with PAD. Furthermore, there was a significant positive relationship between mitochondrial-derived ROS production and intramuscular protein carbonyl concentrations (Figure 3D), suggesting a link between mitochondrial-derived ROS production and skeletal muscle oxidative stress. In contrast, patients with PAD and the HC exhibited similar levels of systemic free radicals, assessed by EPR spectroscopy in whole blood, corresponding to no difference in systemic levels of oxidative stress, as evidenced by similar plasma markers of lipid peroxidation (MDA) and protein oxidation (protein carbonyls) (Table 4). Thus, the present data suggest that the redox balance within skeletal muscle mitochondria may be more susceptible to increased systemic inflammation, resulting in elevated ROS production, although causality cannot be determined from the current study design. Of note, significant relationships between plasma concentrations of TNF-α and intramuscular protein carbonyl concentrations (Figure 3B), as well as between intramuscular levels of IL-6 and 4-HNE (Figure 3C), suggest a link between systemic inflammation and skeletal muscle oxidative stress, which is potentially mediated by excess mitochondrial-derived ROS production. Conversely, increased mitochondrial-derived ROS during the early stages of PAD may potentially serve to preserve mitochondrial function, mediated by an upregulation of the transcription factors NRF-1, NRF-2, and the PPARgamma coactivator-1alpha (PGC-1α), which are involved in mitochondrial biogenesis (Irrcher, Ljubicic, & Hood, 2009; Lee, Yin, Chi, & Wei, 2002; St-Pierre et al., 2003). While significant interactions between the ABI, mitochondrial-derived ROS, and markers of inflammation and oxidative stress were evident when both groups were combined in the present study, future studies that include a broader range of ABI’s in patients with PAD are needed to determine the relationship between ROS and mitochondrial function with increasing disease severity.

Interestingly, while there were no differences in mitochondrial respiratory function between patients with PAD and the HC, a significant negative correlation suggests a link between State 2, uncoupled, respiration and intramuscular 4-HNE (Figure 3A). Indeed, increased levels of superoxide and lipid peroxides have been documented to activate mitochondrial uncoupling to protect against further damage from ROS (Harper, Bevilacqua, Hagopian, Weindruch, & Ramsey, 2004) and the lipids composing the mitochondrial membrane are highly susceptible to oxidative damage (Chen & Yu, 1994), leading to the production of 4-HNE. In the present study, patients with PAD had greater levels of 4-HNE despite similar State 2 respiration compared to the HC. Therefore, the lipid oxidative damage may be a consequence of an insufficient adaptive response in patients with PAD to increase mitochondrial uncoupling in the face of increased levels of ROS. However, in recognition that our tissue samples were of inadequate mass to provide measurements of 4-HNE in all subjects (HC = 9, PAD = 7), future studies are needed to further elucidate the relationship between lipid peroxidation and the control of uncoupled mitochondrial respiration in patients with PAD.

The authors wish to thank all the subjects who partook in this study for their dedicated participation in this research.

Grants

This work was funded, in part, by the National Heart, Lung, and Blood Institute at the National Institute of Health (K99HL125756, PO1 HL1091830); the Flight Attendant Medical Research Institute (FAMRI), the Veteran’s Administration Rehabilitation Research and Development Service (E6910-R, E1697-R, E1433-P and E9275-L), Senior Research Career Scientist Award (E9275-L), and the Spire Award E1572P.