High-intensity interval training prevents oxidantmediated diaphragm muscle weakness in hypertensive mice

Transcription

1 THE JOURNAL RESEARCH High-intensity interval training prevents oxidantmediated diaphragm muscle weakness in hypertensive mice T. Scott Bowen, 1 Sophia Eisenkolb, Juliane Drobner, Tina Fischer, Sarah Werner, Axel Linke, Norman Mangner, Gerhard Schuler, and Volker Adams Department of Internal Medicine and Cardiology, Leipzig University Heart Center, Leipzig, Germany ABSTRACT: Hypertension is a key risk factor for heart failure, with the latter characterized by diaphragm muscle weakness that is mediated in part by increased oxidative stress. In the present study, we used a deoxycorticosterone acetate (DOCA)-salt mouse model to determine whether hypertension could independently induce diaphragm dysfunction and further investigated the effects of high-intensity interval training (HIIT). Sham-treated (n =11), DOCA-salt-treated (n =11), and DOCA-salt+HIIT-treated (n =15) mice were studied over 4 wk. Diaphragm contractile function, protein expression, enzyme activity, and fiber cross-sectional area and type were subsequently determined. Elevated blood pressure confirmed hypertension in DOCA-salt mice independent of HIIT (P < 5). Diaphragm forces were impaired by 15 2% in DOCA-salt vs. sham-treated mice (P < 5), but this effect was prevented after HIIT. Myosin heavy chain (MyHC) protein expression tended to decrease ( 3%; P = 6) in DOCA-salt vs. sham- and DOCA-salt+HIIT mice, whereas oxidative stress increased (P < 5). Enzyme activity of NADPH oxidase was higher, but superoxide dismutase was lower, with MyHC oxidation elevated by 5%. HIIT further prevented direct oxidant-mediated diaphragm contractile dysfunction (P < 5) after a 3 min exposure to H 2 O 2 (1 mm). Our data suggest that hypertension induces diaphragm contractile dysfunction via an oxidantmediated mechanism that is prevented by HIIT. Bowen, T. S., Eisenkolb, S., Drobner, J., Fischer, T., Werner, S., Linke, A., Mangner, N., Schuler, G., Adams, V. High-intensity interval training prevents oxidant-mediated diaphragm muscle weakness in hypertensive mice. FASEB J. 31, 6 71 (217). KEY WORDS: skeletal muscle exercise heart failure Heart failure (HF) is a major public health concern diagnosed in ;15 million Europeans, with a mortality rate of ;5% over 5 yr (1). Hypertension representsakeyriskfactor for the development of HF, particularly in patients with HF with preserved left ventricular ejection fraction (HFpEF) who represent half of the overall HF population (2, 3). Recent clinical trials have shown that patients with HFpEF do not respond to traditional drug treatments targeting the cardiovascular system (3), which provides a compelling rationale for developing alterative therapeutic approaches, by investigating noncardiac peripheral mechanisms. Weakness of the main respiratory muscle, the diaphragm, is a significant problem in patients withhf,exacerbatingthekeysymptomsofexercise ABBREVIATIONS: CSA, cross-sectional area; DOCA, deoxycorticosterone acetate; LV, left ventricle; GPX, glutathione peroxidase; HF, heart failure; HFpEF, heart failure with preserved left ventricular ejection fraction; HFrEF, heart failure with reduced left ventricular ejection fraction; HIIT, high-intensity interval training; MAFbx, muscle atrophy F-box; MuRF1, muscle RING finger 1; MyHC, myosin heavy chain; ROS, reactive oxygen species; SOD, superoxide dismutase 1 Correspondence: University of Leipzig Heart Center, Strümpellstrasse 39, 4289 Leipzig, Germany. bows@med.uni-leipzig.de doi: 1.196/fj R intolerance, shortness of breath, and fatigue (4 6), while being an independent predictor of prognosis (7). Although HFpEF itself has only more recently been shown to lead to diaphragm dysfunction (6, 8, 9), to our knowledge, it remains unclear whether hypertension can independently induce diaphragm muscle weakness. Although a limited number of studies in hypertensive rats have reported numerous functional and molecular impairments to limb muscles (e.g., soleus) (1 12), the diaphragm represents a more appropriate skeletal muscle for study, because it is constantly recruited and thus resistant to many of the confounding factors typically associated with disuse in the limbs, and more recent data suggest that it is more sensitive to impairments during the development of HF when compared with limb muscle (8, 13). Such evidence, therefore, provides a strong rationale that the diaphragm is appreciably more susceptible to impairments induced by hypertension compared to the limb muscles. Thus, many patients with hypertension may have existing impairments to respiratory muscle function before the onset of HFpEF (i.e., in the preclinical stage) that exacerbate symptoms that until now have been largely overlooked /17/31-6 FASEB

2 In addition, whether similar molecular mechanisms that mediate diaphragm dysfunction in HF may act in hypertension remains unknown. For example, a key mechanism contributing to HF-induced diaphragm muscle weakness is an increase in reactive oxygen species (ROS; e.g., superoxide anion and H 2 O 2 ) (14), as produced by the key sources of NADPH oxidase (15), xanthine oxidase(16),and the mitochondria (17), which consequently mediate a cascade of post-translational oxidative modifications to key proteins involved in the excitation contraction process (16, 18), while increasing activity of the ubiquitin proteasome system, thus mediating fiber atrophy and reductions in contractile protein content (19, 2). Furthermore, endothelial dysfunction is also implicated in HF (21 23), which can directly reduce muscle blood flow and promote tissue hypoxia that potentiates ROS production and ultimately impairs diaphragm contractile function (24, 25). One particular type of oxidation reaction that occurs with increased ROS is the formation of protein carbonyls (26), which are known to be increased in the diaphragm of HF rats (14). This particular oxidative modification has also been reported to specifically target key contractile proteins in the diaphragm of HF animals, including myosin heavy chain (MyHC) and actin (16, 18). As hypertension is also characterized by oxidative tissue damage (27), ROSmediated diaphragm contractile dysfunction may also occur in this condition; however, at present, current evidence on this topic remains elusive. Exercise training is a well-established therapeutic intervention that attenuates skeletal muscle dysfunction in cardiovascular disease (28), which can specifically reduce oxidative stress (29). In support, recent data from animals demonstrated that high-intensity interval training (HIIT) prevents the diaphragm dysfunction induced by HFpEF (8), with additional studies showing general antioxidant administration (14) or genetic deletion of the NADPH oxidase subunit p47 phox can prevent ROS-mediated diaphragm impairments in HF (15). Evidence also indicates that inspiratory muscle training can improve respiratory muscle function in patients with hypertension (3), and it can be collectively inferred that diaphragm dysfunction may be an existing problem in hypertension that would be attenuated by exercise training, which may reduce oxidant-mediated muscle weakness. In particular, HIIT has been touted as the most beneficial exercise intervention for patients with cardiovascular disease (31 33), providing superior benefits in the skeletal muscle compared with traditional endurance training (34). In contrast, however, recent evidence has suggested that HIIT induces a pathologic state in hypertension (35, 36) that also has deleterious effects on skeletal muscle (37). Therefore, it remains controversial whether HIIT is conducive or detrimental in the setting of hypertension, with more evidence needed to support or reject either of these hypotheses. In the present study, therefore, we used a hypertensive deoxycorticosterone acetate (DOCA)-salt mouse model (27, 38 4) to determine whether hypertension independently induces diaphragm contractile dysfunction, while further investigating the effects of HIIT. To provide mechanistic insight, we made numerous molecular and cellular measures in the diaphragm with a specific focus on whether oxidative stress may mediate muscle impairmentsinhypertensionthatcould be subsequently attenuated by HIIT. Based on current evidence showing that diaphragm dysfunction in HF is mediated, in part, by increased ROS which can be attenuated by exercise training, we hypothesized that HIIT would prevent oxidant-mediated diaphragm dysfunction induced by hypertension. MATERIALS AND METHODS Ethics approval This study was approved by the Landesdirektion Sachsen and the University of Leipzig Animal Research Council (TVV 1/15). Experimental design Hypertension was induced over 4 wk in 8-wk-old C57BL/6 male mice by unilateral nephrectomy, with subcutaneous implantation of a controlled-release DOCA pellet (.7 mg/d; Innovative Research of America, Sarasota, FL, USA), with the addition of 5% NaCl to drinking water (27, 38, 39). Sham-treated mice underwent the same operation but without pellet implantation and received normal drinking water. At d 14, a time point where hypertension has already developed in this animal model (27, 38 4), acohortofdoca-saltmicewasrandomizedtoperformhiitfor the remaining 2 wk. Overall, 3 groups of mice were therefore studied: sham (n=11), DOCA-salt (n=11), and DOCA-salt+HIIT (n=15). Animals were exposed to identical conditions, in a h light/dark cycle, with food and water provided ad libitum. The mice were euthanized after deep anesthetization with intraperitoneal administration of fentanyl (5 mg/kg), medetomidine ( mg/kg), midazolam (5 mg/kg), and ketamine (1 mg/kg). Treadmill exercise At 2 and 4 wk after surgery, all mice performed a maximum exercise test on a treadmill at 25 incline for determination of exercise intolerance. All mice underwent a familiarization process in the 5 d leading up to each test, which consisted of 1 min treadmill running per session at 15 m/min. The maximum exertion test consisted of increasing the running speed by 1.8 m/min every2minuntilthemicewereunabletorun(41).thetimeof exhaustion and maximum running speed were recorded. In addition, at d 14, a cohort of DOCA-salt mice was randomized to 8 sessions of HIIT over the remaining 2 wk, with each session consisting of 4 min of 4 intervals at 9% peak running speed, separated by 3 min recovery at 6% peak running speed at a 25 incline. Each session was preceded and followed by a 1 min warm-up and cool-down at 4% peak running speed, respectively. Cardiovascular measurements Hypertension was confirmed by measurement of systolic blood pressure in conscious, nonanesthetized mice on heated platforms (35 C), by using the Tail-Cuff Method (TSE Systems GmbH, Bad Homburg, Germany) (42). Mice underwent 7 d of familiarization with procedures (e.g., restraint, heating, and cuff inflation) before final assessment, where 5 measures were first performed before the final 3 measurements, which provided a mean systolic blood pressure and heart rate value. Before euthanasia, the mice were HIIT AND DIAPHRAGM FUNCTION IN HYPERTENSION 61

3 maintained in an ;% isoflurane oxygen balanced mixture, and echocardiography was performed in M-mode with a 3 MHz transducer (Vevo 77; Visual Sonics Europe, Amsterdam, The Netherlands), with LV fractional shortening and ejection fraction, calculated from internal chamber diameters (16). At euthanasia, the heart was dissected, blotted dry, and weighed, with a medial section fixed in 4% PBS-buffered formalin. Serial cross sections (3 mm) mounted on glass slides were subsequently stained with Sirius Red to assess fibrosis with image analysis software (Analysis 3.; Olympus Soft Imaging Solutions, GmbH, Münster, Germany). Diaphragm contractile function A laparotomy and thoracotomy were performed to allow complete excision of the diaphragm. The right costal diaphragm muscle was immediately snap frozen in liquid N 2 for subsequent molecular analyses, with a small part also pinned to cork and fixed in 4% PBS-buffered formalin for further histologic analyses. The left costal diaphragm muscle was prepared in a Krebs- Henseleit buffer solution [(mm) 12 NaCl, 4.8 KCl, 1.2 MgSO 4, 1.2 NaH 2 PO 4,2.4NaHCO 3,1.6CaCl 2, 1 dextrose, and 1 pyruvate (ph ;7.4)] at room temperature equilibrated with 95% O 2, 5% CO 2 for contractile measurements (n =9 1 per group). In brief, a muscle bundle connected from rib to central tendon was dissected, attached to silk sutures (4-) at either end, and mounted vertically in a buffer-filled organ bath. The suture connected to the rib was secured to a hook at the bottom of the bath, and the tendon was tied to a length-controlled lever system (31B; Aurora Scientific, Inc., Aurora, ON, Canada). In vitro muscle function was assessed by using platinum electrodes to stimulate the muscle with a supramaximum current(6 ma; 5 ms train duration;.25 ms pulse width) via a high-power bipolar stimulator (71B; Aurora Scientific Inc.). The muscle bundle was setatanoptimallength(l o ) equivalent to the maximum twitch force produced, after which the bath temperature was increased to 37 C, followed by a 15 min thermoequilibration period. A force frequency protocol was then performed at 1, 15, 3, 5, 8, 12, 15, and 3 Hz, separated by 1 min rest intervals. After a 5 min period in which muscle length was measured with digital calipers, the muscle underwent a fatigue protocol over 5 min (4 Hz every 2 s). The muscle was subsequently detached, trimmed free from rib and tendon, blotted dry, and weighed. Force (N) was normalized to muscle cross-sectional area (CSA; cm 2 ) by dividing muscle mass (g) by the product of L o (cm) and estimated muscle density (6) (43), which allowed specific force (N/cm 2 )to be calculated. Fiber histology Paraffin-embedded diaphragm sections (3 mm) were incubated overnight at 4 C with primary antibodies for slow- and fasttwitch fibers (1:4; Sigma-Aldrich, Taufkirchen, Germany), which were subsequently detected by a CSA-II kit (Dako, Hamburg, Germany). Fiber type and CSA were then evaluated by image analysis software (Analysis Five; Olympus Soft Imaging Solutions, GmbH). Western blot analysis Frozen muscle samples were homogenized in relaxing buffer [(mm) 9 HEPES, 126 KCl, 36 NaCl, 1 MgCl, 5 EGTA, 8 ATP, and 1 creatine phosphate (ph 7.4)], containing a protease inhibitor mix (inhibitor mix M; Serva, Heidelberg, Germany), and sonicated for 1 cycles (Sonoplus GM7; Bandelin Electronics, Berlin, Germany), with protein content of the homogenate subsequently determined (bicinchoninic acid assay; Thermo Fisher Scientific, Bonn, Germany). Diaphragm homogenates (5 2 mg) mixed with loading buffer [126 mm Tris-HCl, 2% glycerol, 4% SDS, % 2-ME, and 5% bromophenol blue (ph 6.8)] were separated by SDS-PAGE. Proteins were transferred to a PVDF membrane and incubated overnight at 4 C with the following primary antibodies: MyHC (1:1; Sigma-Aldrich); troponin T, C, and I (all 1:2; Santa Cruz Biotechnology, Santa Cruz, CA, USA); MAFbx (1:2; generated in rabbits against the following peptide sequence CYPRKEQYGDTLQL; Eurogentec, Seraing, Belgium); and MuRF1 (1:1; Abcam, Cambridge, United Kingdom). Membranes were subsequently incubated with a horseradish peroxidase conjugated secondary antibody, and specific bands were visualized by enzymatic chemiluminescence (Super Signal West Pico; Thermo Fisher Scientific) and densitometry quantified with a 1D scan software package (Scanalytics, Inc., Rockville, MD, USA). Blots were then normalized to the loading control GAPDH (1:3,; HyTest Ltd., Turku, Finland). Carbonylation of MyHC was also determined with an OxyBlot kit, in line with the company s instruction manual (Millipore, Darmstadt, Germany). In brief, the samples (15 mg of total protein) were incubated with 2,4-dinitrophenylhydrazine (DNPH) for 2 min at room temperature to derivatize the carbonyl groups in the protein side chains to DNP-hydrazone. The reaction was stopped by the addition of neutralization solution and followed by 8% SDS-PAGE and subsequent transfer to a PVDF membrane. Membranes were incubated with antibody against DNP (1:15; 1 h at room temperature) followed by a horseradish peroxidase conjugated secondary antibody (1:3; 1 h at room temperature). Specific bands were then visualized as previously described, and stained with Ponceau S to normalize for protein loading. All data are presented as fold change relative to sham-treatment group. Enzyme activities Frozen muscle samples were homogenized and protein content determined as previously described. Activities of pro- and antioxidative enzymes, including NADPH oxidase (44), catalase (45), glutathione peroxidase (GPX) (46), superoxide dismutase (SOD) (47), and the mitochondrial enzymes citrate synthase (48) and lactate dehydrogenase (49), were measured photometrically according to standard protocols and recorded as units per milligram protein. Data are presented as fold change relative to sham-treatment group. Effects of H 2 O 2 on diaphragm function In a further set of experiments, the direct effects of the oxidant H 2 O 2 on diaphragm function were tested. Ten-week-old male C57BL/6 mice (n =28) were randomized to 2 wk of sedentary behavior (n =2) or HIIT (n =8; same protocol as described above). At euthanasia, diaphragm fiber bundles from sedentary (n=1) and HIIT (n=8) mice were incubated in Krebs-Henseleit buffer containing 1 mm H 2 O 2 (35% solution; Roth, Karlsruhe, Germany), whereas fiber bundles from the remaining sedentary mice (n=1) incubated in buffer alone served as the control. After 3 min, fiber bundles were mounted in an organ bath in fresh buffer and diaphragm function was assessed (as described above). H 2 O 2 concentration and incubation time were based on previous experiments showing that contractile function is altered in mouse skeletal muscle under these conditions (5). Endothelial function To provide an index of vascular function, aortic rings (n=4) from each animal were mounted in a buffer-filled (in mm: 118 NaCl, 62 Vol. 31 January 217 The FASEB Journal x BOWEN ET AL.

4 25 NaHCO 3,4.7KCl,1.2KH 2 PO 4,1.2MgSO 4, 2.5 CaCl 2,5.5 glucose) organ bath between a hook and a force transducer (21). After 3 min of equilibrium, the maximum constriction was achieved by adding KCl (final concentration, 1 mm) to the buffer. After several rinses, the aortic rings were preconstricted by adding increasing concentrations of phenylephrine (Sigma-Aldrich; M) to ;7% of maximum KCl constriction. Relaxation was then recorded to increasing concentrations of acetylcholine (Sigma-Aldrich, M). Statistical analyses Data are presented as means 6 SEM. One-way ANOVA, followed by Tukey s post hoc test was used to compare groups, whereas 2-way repeated-measures ANOVA followed by Tukey s post hoc test was used to assess diaphragm and endothelial function (Prism; GraphPad, La Jolla, CA, USA). Significance was accepted as P, 5. RESULTS Animal model Physical, cardiovascular, and exercise measures of all mice are presented in Table 1. Numerous cardiovascular measures, including elevated systolic blood pressure, increased heart weight, and LV hypertrophy, confirmed the development of hypertension in DOCAsalt mice independent of HIIT. In contrast, whereas DOCA-salt mice demonstrated significantly impaired exercise tolerance, this effect was reversed after 2 wk of HIIT. In addition, hypertensive DOCA-salt mice presented with normal systolic function independent of HIIT, with no evidence of pulmonary congestion or cardiac fibrosis. Overall, therefore, these data support that our intervention induced hypertension in the presence of no secondary cardiovascular disease. Diaphragm contractile function In vitro assessment of muscle fiber bundles revealed significant muscle weakness in DOCA-salt compared to sham-treated mice, with maximum force impaired by ;2% (P, 5; Fig. 1A). In contrast, however, 2 wk of HIIT in DOCA-salt mice normalized diaphragm contractile function (P. 5), with submaximum and maximum forces comparable to those of shamtreated mice. In addition, diaphragm fiber bundles generated lower specific forces (P, 5) at an earlier time point during the fatigue protocol from DOCAsalt compared to sham-treated and DOCA-salt+HIIT mice by ;15% (Fig. 1B); however, no differences were discerned in relative fatigue after force was normalized to initial contraction (P. 5). TABLE 1. Physical, cardiovascular, echocardiography, and exercise intolerance measures of all mice Parameter Sham treatment, n=11 DOCA-salt, n=11 DOCA-salt + HIIT, n=15 Physical characteristic Body weight (g) * Heart weight (mg) Heart-to-body weight (mg/g) Cardiac fibrosis (%) Lung wet/dry ratio Diaphragm-to-body weight (mg/g) Cardiovascular and echocardiography Systolic blood pressure (mmhg) Heart rate (bpm) LVFS (%) LVEF (%) LVAWd (mm) LVAWs (mm) LVPWd (mm) * LVPWs (mm) * LVEDD (mm) LVESD (mm) LV mass (mg) Exercise intolerance Time to exhaustion (min) 2 wk wk Maximum speed (m/min) 2wk wk Data are means 6 SEM. LV, left ventricular; LVFS, left ventricular fractional shortening; LVEF, left ventricular ejection fraction; LVAWd and LVAWs, left ventricular anterior wall during diastole and systole, respectively; LVPWd and LVPWs, left ventricular posterior wall during diastole and systole, respectively; LVEDV, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter. *P, 5 vs. sham treatment; P, 5 vs. other groups. HIIT AND DIAPHRAGM FUNCTION IN HYPERTENSION 63

5 A Specific force (N/cm 2 ) B Specific force (N/cm 2 ) * * * Sham DOCA HIIT Frequency (Hz) Sham DOCA HIIT activity of the antioxidative enzyme SOD was significantly reduced (;4%) in DOCA-salt mice compared to that in those that performed HIIT (Fig. 4B). In contrast, no significant differences were found between groups for the other antioxidative enzymes of catalase (Fig. 4C) and GPX (Fig. 4D). As an increased NADPH oxidase and lower SOD enzyme activity indicates greater ROS load, we subsequently probed for post-translational oxidative protein modifications in carbonyl formation. In accordance, we also found carbonylation of MyHC significantly increased (46%) in DOCA-salt mice compared to sham-treated mice, but this effect was prevented by HIIT (Fig. 5A). Overall, A Fibre CSA (µm2) 3 Sham DOCA 25 HIIT All fibres Type I Type II Time (s) Figure 1. In vitro contractile function of diaphragm fiber bundles from each experimental group: sham treatment (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT), stimulated across a range of frequencies (A) and during repetitive contractions at 4 Hz (B). Data are means 6 SEM. P, 5, DOCA-salt vs. DOCA-salt+HIIT; *P, 5, sham treatment and DOCA-salt+HIIT vs. DOCA-salt. B Fibre proportion (%) 1 Sham DOCA 75 HIIT 5 25 Muscle fiber atrophy and contractile protein expression Fiber CSA was not different between groups (P. 5; Fig. 2A), nor was fiber type proportion (P. 5; Fig. 2B). In accordance, therefore, we found no differences between groups in the protein expression of 2 key markers of muscle atrophy: MAFbx and MuRF1 (P. 5; Fig. 2C). Compared to sham-treated mice, however, the protein expression of the main contractile protein MyHC was reduced by 35% in DOCA-salt mice (P = 6), but this effect was prevented after HIIT (Fig. 3A). No further changes (P. 5) were found in the protein expression of other contractile proteins, including troponin T, I (Fig. 3C), and C (Fig. 3D) and actin (Fig. 3E). C 2. Type I Type II S S D D H H 4 kda MuRF-1 42 kda MafBx 38 kda GAPDH Oxidative stress The enzyme activity of NADPH oxidase, a key source of ROS, was elevated in DOCA-salt mice by 77% compared to sham-treated mice (P, 5); however, the increase was attenuated after HIIT (P. 5: Fig. 4A). Furthermore, the MAFbx MuRF1 Figure 2. Diaphragm fiber CSA (A), fiber type proportion (B), and protein expression of key markers of muscle atrophy MAFbx and MuRF1 (C) in sham-treated (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT) mice. Data are means 6 SEM. 64 Vol. 31 January 217 The FASEB Journal x BOWEN ET AL.

6 A MyHC B Troponin T * Sham DOCA HIIT Sham DOCA HIIT C Troponin I D Troponin C Sham DOCA HIIT Sham DOCA HIIT E Actin F S S D D H H 25 kda MyHC 45 kda Actin 36 kda 25 kda Trop-T Trop-I 18 kda Trop-C Sham DOCA HIIT 38 kda GAPDH Figure 3. Proteinexpressioninthediaphragmofkeycontractileproteins,includingMyHC(A); troponin T (B), I (C), and C (D); and actin (E), as well as representative examples of 2 per group (F) in sham-treated (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT) mice. Values were normalized to GAPDH, which served as the loading control. Data are means 6 SEM.*P = 6 vs. other groups. these data indicate that HIIT protected against elevated oxidative stress and post-translational oxidative protein modifications that likely impaired force generation. Therefore, in an additional set of experiments, we directly confirmed this notion, by incubating diaphragm fiber bundles in the oxidant H 2 O 2 for 3 min. Fiber bundles incubated in H 2 O 2 were significantly weaker by ;3%incomparisontobundlesincubated in buffer alone (Fig. 5B). In contrast, however, diaphragm fiber bundles from mice that had performed 2 wk of HIIT did not demonstrate impairments of force generation after H 2 O 2 incubation, with the force frequency relationship essentially unchanged from control values (P. 5). Mitochondrial and vascular function There was a strong trend for HIIT to benefit both oxidative and glycolytic metabolism, with citrate synthase activity (a marker of mitochondrial density) tending to be 24 and HIIT AND DIAPHRAGM FUNCTION IN HYPERTENSION 65

7 A NADPH oxidase B SOD 2. * Sham DOCA HIIT Sham DOCA HIIT C Catalase D GPX Sham DOCA HIIT Sham DOCA HIIT Figure 4. Diaphragm enzyme activity of NADPH oxidase, a key source of ROS (A), as well as radical scavenging enzymes, including SOD (B), catalase (C), and GPX (D) measured in sham-treated (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT) mice. Data are means 6 SEM. *P, 5 vs. sham treatment; P, 5 vs. other groups. 46% higher after HIIT compared to the sham-treatment and DOCA-salt groups, respectively (P = 7; vs and fold sham), whereas lactate dehydrogenase activity (a marker of glycolysis) also tended to be increased by ;2% after HIIT compared with sham-treated and DOCA-salt mice, respectively (P = 7; vs. 62- and fold sham). In addition, endothelium-dependent function measured in aortic rings (an index of vascular function) was severely impaired in DOCA-salt mice compared with that in sham-treated mice at most acetylcholine concentrations (P, 5: Fig. 6). However, the endothelial response was fully normalized in DOCA-salt+HIIT mice relative to sham-treated mice (P. 5), suggesting that impairmentsinvascularfunctionwerepreventedafterhiit. DISCUSSION In the present study, we used a DOCA-salt mouse model to induce hypertension over 4 wk, while further combining 2 wk of HIIT, to investigate the subsequent functional and molecular effects on the diaphragm muscle. The main findings of this study were that hypertension induced diaphragm contractile dysfunction, and this effect was associated with: 1) higher activity in the key ROS source NADPH oxidase; 2) increased MyHC oxidation; 3)astrongtrendtowardlower MyHC protein expression; 4) endothelial dysfunction; and 5) significantly impaired exercise tolerance. In contrast, however, HIIT in hypertensive mice prevented diaphragm contractile dysfunction concomitant with a significant attenuation of oxidative stress (i.e., lower pro- but higherantioxidative enzyme activities), with no increase in MyHC oxidation and also unchanged levels of MyHC protein expression, whereas endothelial function and exercise tolerance were normalized. Moreover, in an additional set of experiments, diaphragm fiber bundles excised from healthy mice after 2 wk of HIIT were protected from oxidant-mediated diaphragm dysfunction after prolonged exogenous exposure to H 2 O 2, which lends direct support to the notion that HIIT protects against ROS-induced contractile dysfunction. Overall, therefore, our data provided novel evidence suggesting that HIIT can prevent the oxidant-mediated diaphragm muscle weakness in hypertension that likely contributes to reversing exercise intolerance. Our study therefore provides initial evidence that hypertension may induce diaphragm impairments before the onset of subsequent chronic diseases such as HFpEF, whereas the intervention of HIIT can prevent or even reverse many of these detrimental alterations. Diaphragm muscle weakness in hypertension Hypertension is a main risk factor associated with the development of HF, particularly in patients with HFpEF (2, 3). Diaphragm muscle weakness is a common problem in patients with HF with reduced left ventricular ejection fraction (HFrEF), but more recent data have suggested that 66 Vol. 31 January 217 The FASEB Journal x BOWEN ET AL.

8 A 2. B Specific force (N/cm 2 ) Sham MyHC Carbonylation * DOCA HIIT S * * * D Buffer H 2 O 2 H Carbonylated MyHC Total MyHC Figure 5. A) Oxidation of MyHC, as assessed by carbonylation, measured from the diaphragm in sham-treated (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT) mice, with a representative example also presented. Carbonylated MyHC was normalized to total MyHC. Data are means 6 SEM. *P, 5 vs. other groups. B) Contractile function of diaphragm fiber bundles stimulated across a range of frequencies, as assessed after incubation for 3 min in either normalbufferorwiththeoxidanth 2 O 2 at a concentration of 1 mm from healthy sedentary mice (H 2 O 2 ) or mice that had performed 2 wk of HIIT (HIIT+H 2 O 2 ). Data are means 6 SEM. P, 5, buffer vs. H 2 O 2 ;*P, 5, buffer and HIIT+H 2 O 2. vs. H 2 O Frequency (Hz) HIIT+H 2 O 2 those with HFpEF are also appreciably susceptible (6, 8, 9). To our knowledge, however, the evidence concerning whether hypertension can independently induce diaphragm muscle weakness remains unclear. Given the prevalence of hypertension in patients with HFpEF, we argued that this risk factor alone might initiate diaphragm weakness a suggestion supported by evidence showing that abnormalities are also developed in limb skeletal muscle in this condition (1 12). That respiratory muscle has been suggested to be more sensitive to functional and molecular alterations in HF compared to that of limb muscle (8, 13) provides additive support for our rationale to investigate diaphragm function in hypertension. In support of our hypothesis, we provide novel evidence demonstrating that diaphragm fiber bundles from hypertensive mice had impaired forces of ;15 2%, which is similar to that reported in HFpEF (16) and also in HFrEF induced by myocardial infarction (15 17, 51). Our data suggest that a significant proportion of patients with hypertension may be predisposed to development of diaphragm muscle weakness, which could occur during the preclinical stage of subsequent chronic diseases such as HFpEF. The onset of diaphragm dysfunction is predicted to exacerbate symptoms of dyspnea, fatigue, and exercise intolerance in hypertensive patients all of which are also typical of the HF syndrome. Although patient studies investigating diaphragm (or inspiratory) muscle weakness in hypertension remain sparse, our data provide a strong rationale for focus to be directed toward this topic. However, 2 wk of HIIT in our hypertensive DOCA-salt mice prevented diaphragm contractile dysfunction, providing evidence that this muscle has a certain degree of plasticity in hypertension. Indeed, our data complement previous findings that showed that hypertensive patients are able to improve inspiratory muscle strength, an index of diaphragm function, after inspiratory muscle training (3). Because the present study confirmed functional deficits to diaphragm fiber bundles in hypertension, we next investigated the potential underlying cellular and molecular mechanisms. Relaxation (%) * * * * * sham DOCA HIIT * * * * * * Acetylcholine concentration (mol/l) Figure 6. Endothelial function in aortic rings excised from shamtreated (sham), DOCA-salt (DOCA), and DOCA-salt+HIIT (HIIT) mice, as assessed by the relaxation (i.e., vasodilation) in response to various concentrations of acetylcholine. Data are means 6 SEM. P, 5, DOCA-salt+HIIT vs. DOCA-salt; *P, 5, DOCA-salt vs. sham treatment and DOCA-salt+HIIT. HIIT AND DIAPHRAGM FUNCTION IN HYPERTENSION 67

9 Mechanisms of diaphragm muscle weakness in hypertension That diaphragm function was reduced in our hypertensive mice to levels often observed in HF (8, 15 17, 51) indicates that similar mechanisms probably act in both conditions. Indeed, akin to HF, hypertension is characterized by a prooxidative state (27), which we reasoned would contribute to the potential diaphragm impairments developed in our DOCA-salt mice. Evidence suggests that diaphragm weakness in HF is caused by both muscle atrophy and contractile dysfunction, with the former mediated by an upregulation of catabolic factors (e.g., the E3 ligases MuRF1 and MAFbx and also the ubiquitin proteasome, calpain, and caspase systems) and the latter mediated by post-translational oxidative modifications to intracellular proteins involved in the excitation contraction process (16, 18) and reductions in contractile protein content (19, 2). Muscle atrophy and contractile protein content Although atrophy of the diaphragm is common in HF, we found no change in fiber CSA in both the type I and II isoforms. The finding of unchanged protein expression levels in the 2 key E3 ubiquitin ligases involved in skeletal muscle atrophy, MuRF1 and MAFbx (52), supports the notion that diaphragm muscle weakness in the present study was mediated by intracellular impairments, rather than by a general loss of muscle mass. A key determinant underpinning skeletal muscle force generation is MyHC, which determines the number of functional myosin actin cross bridges developed. Indeed, previous studies have documented a lower MyHC content in single diaphragm fibers taken from HF rats, which was associated with impaired forces but in the absence of fiber atrophy (19, 2). Similarly, in the present study there was a strong trend (P = 6) for MyHC protein content to be reduced (by ;3%) in hypertensive DOCA-salt mice compared to sham-treated mice. This result may reveal, at least in part, a potential mechanism for the diaphragm muscle weakness observed in hypertension. Further, our data also showed no significant differences (or statistical trends) in levels of other key contractile proteins (i.e., actin or troponin) between sham-treated and DOCA-salt mice, which suggests that hypertension induces a preferential loss of MyHC content. This supposition is in line with data taken from single vastus lateralis fibers in patients with HF, where a preferential loss in MyHC content was observed that was associated with no fiber atrophy (53). Therefore, the finding of lower MyHC content in the absence of fiber atrophy seems consistent in HF (19, 2, 53), with the present data suggesting that this lower MyHC level is also prevalent in hypertension. MyHC protein expression tended to be lower in our hypertensive mice, but key molecular markers (i.e., MuRF1 and MAFbx) characterizing the ubiquitin proteasome system (the main pathway of protein degradation) were unchanged. Although it remains unclear what pathway is responsible for increasing MyHC degradation, it may be related to activation of other degradation pathways (54) or the temporal profile of the measurement(i.e., elevated E3 ubiquitin ligase levels may have already returned to baseline) (55), or alternatively, a decrease in protein synthesis may have occurred (rather than an increase in protein degradation) (54). Taken together, therefore, our findings of no fiber atrophy or isoform shift, but lower MyHC content in hypertension, support the argument that contractile dysfunction underpins diaphragm muscle weakness in hypertension. ROS and post-translational oxidative modifications Current data collected from the diaphragm tissue in HF animal models provide strong support that alterations are mediated upstream in response to an increased production of ROS (14 16, 18), with one key source recently confirmed as NADPH oxidase (15). In support, our hypertensive DOCA-salt mice also demonstrated a significant increase in NADPH oxidase enzyme activity, suggesting increased ROS formation. To probe further for oxidative modifications, we also quantified protein carbonylation specific to the contractile protein MyHC. This protein is a key determinant of force generation that is appreciably susceptible to oxidation, as previously reported in a study of diaphragm fibers from a rat HF model (18). Compared to sham-treated mice, we found that hypertensive DOCAsalt mice had a significant increase in the carbonylation of MyHC, which provides strong evidence that diaphragm muscle weakness in hypertension is mediated, in part, by increased ROS formation via the NADPH oxidase that, in turn, elevates post-translational oxidative modifications of MyHC. Such a notion is reinforced by isolated myofilament data, showing that increased carbonylation of MyHC correlates significantly with impaired sliding velocity (18). Our collective findings therefore suggest that both quantitative and qualitative changes in MyHC contribute to inducing diaphragm muscle weakness in hypertension. Another mechanism that can increase ROS and directly induce diaphragm contractile dysfunction is hypoxia (24, 56). Hypoxia can develop after impairments of vasodilation and thus blood flow (relative to metabolic rate), which in cardiovascular diseases is often underpinned by endothelial dysfunction (21 23). To provide an index of systemic vascular function that we also assumed reflected (at leasttosomedegree)thatofthediaphragm,wemeasured in vitro endothelial function from aortic rings. As expected, endothelial function was severely impaired in hypertensive DOCA-salt mice compared to shams, which likely predisposed to a lower vasodilatory response and a failure of blood flow to match local metabolic demands. Impaired blood flow has been shown to cause diaphragm muscle weakness in a canine model (57), with hypoxia further demonstrated to directly increase ROS formation and induce contractile dysfunction in diaphragm fiber bundles excised from rats (56). These data, along with ours, indicate that diaphragm dysfunction in hypertension may also be mediated by impaired blood flow, which exacerbates ROS formation and directly inhibits force production. Relative contribution of MyHC loss vs. oxidation to diaphragm dysfunction To our knowledge, the present data are the first to show that diaphragm dysfunction induced by hypertension is 68 Vol. 31 January 217 The FASEB Journal x BOWEN ET AL.

10 associated with both a loss of content and an increased oxidation of MyHC (i.e., both quantitative and qualitative changes). These findings raise the intriguing question of what relative importance each of these mechanisms has in contributing to diaphragm muscle weakness in hypertension, as well as in other cardiovascular diseases. Although the current study did not provide a definitive answer, our data reinforce results of previous studies that suggest that both mechanisms are associated with impaired muscle force generation; earlier studies of a rat HF models have demonstrated that diaphragm dysfunction is indeed parallel with both a loss of MyHC content (19, 2, 53) and an increase in MyHC oxidation (18). However, whether these two processes are interrelated where, for example, an increase in the rate of MyHC oxidation consequently stimulates an increase in MyHC degradation remains unclear. A causal relationship for this notion is still lacking (26), but available data have demonstrated that diaphragm muscle weakness is initiated only hours after acute cardiac failure and is associated with elevated oxidation of key contractile proteins (16), whereas the reduction in MyHC content seems to occur some weeks later (19, 2, 53). It therefore remains a possibility that an increase in MyHC oxidation causes the initial decrement in diaphragm function in hypertension and subsequently triggers or exacerbates the subsequent degradation of MyHC. Overall, therefore, attenuating MyHC oxidation may provide the most effective treatment for diaphragm dysfunction observed in hypertension and other cardiovascular diseases. Effects of HIIT on diaphragm dysfunction in hypertension In confirming contractile dysfunction in diaphragm fiber bundles taken from hypertensive DOCA-salt mice, we next assessed the therapeutic benefit associated with HIIT. At present, whether HIIT in the setting of hypertension provides a beneficial or detrimental intervention is highly controversial (32, 35). Indeed, although significant evidence suggests that HIIT is beneficial in a wide range of cardiovascular disorders (31 33), others have posited that it may, in fact, provide a detrimental stimulus that promotes a pathologic state specific to hypertension (35 37). Although we cannot resolutely refute the latter argument, our current data provide strong evidence that HIIT is associated with significant beneficial adaptations in hypertension, as related to improvements in diaphragm function, endothelial function, and exercise tolerance. In addition, variables and parameters of the heart were relatively similar between the groups of hypertensivedoca-saltmicethatdidordidnotperformhiit. Therefore, our data, albeit from a small but well-controlled animal cohort, provide sound evidence supporting the notion that HIIT is beneficial in the setting of hypertension. One of the most important findings in the present study was that diaphragm contractile dysfunction induced by hypertension was prevented after HIIT and was associated with not only unchanged MyHC protein expression levels, but perhaps more important, no change in levels of markers of oxidative stress (e.g., NADPH oxidase, SOD, and MyHC oxidation). These data therefore predict a shift toward a lower ROS load in the hypertensive-hiit mice, which is reinforced by our finding that protein oxidation was not significantly increased. Further, HIIT normalized endothelial dysfunction, which supports our earlier hypothesis that tissue hypoxia elevated ROS formation to induce contractile dysfunction. Overall, therefore, our data collectively suggest that HIIT attenuates oxidative stress, by limiting ROS formation in parallel with preventing a decline in radical scavenging enzymes, which inhibits subsequent post-translational oxidative modifications in MyHC. Available data support that exercise training can reduce oxidative stress in patients with HF, an effect that is associated with improved skeletal muscle and endothelial function (21, 29, 58), but direct support has been provided in animal models where administration of antioxidants has been shown to attenuate ROS and diaphragm muscle weakness in both HF (14, 59) and hypoxic conditions (56). To support our hypothesis that HIIT prevents oxidantmediated diaphragm dysfunction in hypertension, we next attempted to provide causal evidence, by incubating diaphragm muscle fibers from healthy mice in the oxidant H 2 O 2 from either sedentary mice or those that had performed 2 wk of HIIT. Although diaphragm function in sedentary mice (relative to fibers incubated in buffer) was significantly impaired (by ;3%) after H 2 O 2 incubation, this deficit was prevented in mice that had performed 2 wk of HIIT. Our finding of oxidant-induced muscle dysfunction is in line with previous studies in mice, where similar time and concentration impairments have been observed, but in the limbs (5, 6). We speculate that HIIT directly protected against oxidant-mediated diaphragm dysfunction in our hypertensive mice. The underlying mechanisms of this effect are not fully clear, but it is probably related to reduction of the pro-oxidative state and the increase in antioxidant enzyme capacity caused by exercise training and thus results in a reduction in the detrimental chronic effects of ROS on contractile function. Such a suggestion is supported by a recent study that demonstrated that 2 wk of endurance exercise training similarly prevented diaphragm muscle weakness but in mechanically ventilated rats, which was also associated with lower markers of ROS formation and a higher antioxidant enzyme capacity (61). Furthermore, it is likely that HIIT specifically reduces oxidantinduced dysfunction of the contractile apparatus it being well known that oxidants can damage thiol groups on MyHC (18). Taken together, therefore, our data and those of others provide support that the diaphragm dysfunction we observed in our hypertensive DOCA-salt mice was likely related to impairments at the contractile protein level, but HIIT acted to prevent both quantitative and qualitative changes specifically related to MyHC. Limitations Because of technical limitations, we could not make any measurements related to calcium handling in the diaphragm. It remains possible that impairments to sarcoplasmic reticulum Ca 2+ release (62) or myofibrillar Ca 2+ sensitivity (6) contributed to the diaphragm dysfunction that we observed in hypertension. In addition, the mitochondria have recently been shown to be a key source of increased ROS production that mediates diaphragm dysfunction in HF (17). Although HIIT AND DIAPHRAGM FUNCTION IN HYPERTENSION 69

11 we did not measure mitochondrial ROS production in the present study, we acknowledge that it may have contributed to the elevation of protein oxidation and contractile dysfunction. Further studies are therefore warranted to shed light on these issues. CONCLUSIONS The present study showed in a mouse model that hypertension can independently induce diaphragm muscle weakness, and the effect is likely mediated in part by elevated ROS. In contrast, however, 2 wk of HIIT prevented such contractile dysfunction while reducing oxidative stress, with our data further demonstrating HIIT directly prevented oxidant-mediated diaphragm dysfunction. Overall, therefore, our findings suggest that hypertensive patients may be predisposed to oxidant-mediated diaphragm dysfunction, but HIIT could provide a potential therapeutic treatment. ACKNOWLEDGMENTS The authors thank Angela Kricke (Leipzig University Heart Center) for excellent technical assistance. This work was supported by the European Commission: Framework Programme 7 (EU ). The authors declare no conflicts of interest. AUTHOR CONTRIBUTIONS T. S. Bowen, A. Linke, N. Mangner, G. Schuler, and V. Adams designed the research; T. S. Bowen, S. Eisenkolb, J. Drobner, T. Fischer, S. Werner, and V. Adams analyzed the data;t.s.bowen,s.eisenkolb,j.drobner,t.fischer,s. Werner, and V. Adams performed the research; T. S. Bowen andv.adamswrotethepaper;anda.linke,g.schuler, and V. Adams contributed reagents or analytical tools. REFERENCES 1. Dickstein,K.,Cohen-Solal,A.,Filippatos,G.,McMurray,J.J.,Ponikowski, P., Poole-Wilson, P. A., Strömberg, A., van Veldhuisen, D. J., Atar, D., Hoes, A.W.,Keren,A.,Mebazaa,A.,Nieminen,M.,Priori,S.G.,andSwedberg, K.; ESC Committee for Practice Guidelines (CPG). (28) ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 28: the TaskForce forthe diagnosisandtreatment of acute and chronic heart failure 28 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM) Eur. J. Heart Fail. 1, ; correction, 12, Borlaug, B. A., and Paulus, W. J. (211) Heart failure with preserved ejection fraction: pathophysiology, diagnosis, and treatment. Eur. Heart J. 32, Sharma, K., and Kass, D. A. (214) Heart failure with preserved ejection fraction: mechanisms, clinical features, and therapies. Circ. Res. 115, Mancini, D. M., Ferraro, N., Nazzaro, D., Chance, B., and Wilson, J. R. (1991) Respiratory muscle deoxygenation during exercise in patients with heart failure demonstrated with near-infrared spectroscopy. J. Am. Coll. Cardiol. 18, Mancini, D. M., Henson, D., LaManca, J., and Levine, S. (1994) Evidence of reduced respiratory muscle endurance in patients with heart failure. J. Am. Coll. Cardiol. 24, Yamada, K., Kinugasa, Y., Sota, T., Miyagi, M., Sugihara, S., Kato, M., and Yamamoto, K. (216) Inspiratory muscle weakness is associated with exercise intolerance in patients with heart failure with preserved ejection fraction: a preliminary study. J. Card. Fail. 22, Meyer, F. J., Borst, M. M., Zugck, C., Kirschke, A., Schellberg, D., Kübler, W., and Haass, M. (21) Respiratory muscle dysfunction in congestive heart failure: clinical correlation and prognostic significance. Circulation 13, Bowen,T.S.,Rolim,N.P.,Fischer,T.,Baekkerud,F.H.,Medeiros,A., Werner, S., Bronstad, E., Rognmo, O., Mangner, N., Linke, A., Schuler, G., Silva, G. J., Wisloff, U., Adams, V., and Optimex Study Group. (215) Heart failure with preserved ejection fraction induces molecular, mitochondrial, histological, and functional alterations in rat respiratory and limb skeletal muscle. Eur. J. Heart Fail. 17, Lavietes,M.H.,Gerula,C.M.,Fless,K.G.,Cherniack,N.S.,andArora, R. R. (24) Inspiratory muscle weakness in diastolic dysfunction. Chest 126, Bloemberg, D., McDonald, E., Dulay, D., and Quadrilatero, J. (214) Autophagy is altered in skeletal and cardiac muscle of spontaneously hypertensive rats. Acta Physiol. (Oxf.) 21, Gray, S. D., Carlsen, R. C., and Deng, J. (1994) Soleus muscle contractile properties in hypertensive rats. Am. J. Physiol. 267, R735 R Ben Bachir-Lamrini, L., Sempore, B., Mayet, M. H., and Favier, R. J. (199) Evidence of a slow-to-fast fiber type transition in skeletal muscle from spontaneously hypertensive rats. Am. J. Physiol. 258, R352 R VanHees,H.W.,Ottenheijm,C.A.,Granzier,H.L.,Dekhuijzen, P. N., and Heunks, L. M. (21) Heart failure decreases passive tension generation of rat diaphragm fibers. Int. J. Cardiol. 141, Supinski, G. S., and Callahan, L. A. (25) Diaphragmatic free radical generation increases in an animal model of heart failure. J. Appl. Physiol. 99, Ahn, B., Beharry, A. W., Frye, G. S., Judge, A. R., and Ferreira, L. F. (215) NAD(P)H oxidase subunit p47phox is elevated, and p47phox knockout prevents diaphragm contractile dysfunction in heart failure. Am.J.Physiol.LungCell.Mol.Physiol.39, L497 L Bowen, T. S., Mangner, N., Werner, S., Glaser, S., Kullnick, Y., Schrepper, A., Doenst, T., Oberbach, A., Linke, A., Steil, L., Schuler, G., and Adams, V. (215) Diaphragm muscle weakness in mice is early-onset post-myocardial infarction and associated with elevated protein oxidation. J. Appl. Physiol. 118, Laitano, O., Ahn, B. S., Patel, N., Coblentz, P. D., Smuder, A. J., Yoo, J.K.,Christou,D.D.,Adhihetty,P.J.,andFerreira,L.F.(216) Pharmacological targeting of mitochondrial reactive oxygen species counteracts diaphragm weakness in chronic heart failure. J. Appl. Physiol. 12, Coirault, C., Guellich, A., Barbry, T., Samuel, J. L., Riou, B., and Lecarpentier, Y. (27) Oxidative stress of myosin contributes to skeletal muscle dysfunction in rats with chronic heart failure. Am. J. Physiol. Heart Circ. Physiol. 292,H19 H Van Hees, H. W., Li, Y. P., Ottenheijm, C. A., Jin, B., Pigmans, C. J., Linkels, M., Dekhuijzen, P. N., and Heunks, L. M. (28) Proteasome inhibition improves diaphragm function in congestive heart failure rats. Am.J.Physiol.LungCell.Mol.Physiol.294, L126 L Van Hees, H. W., van der Heijden, H. F., Ottenheijm, C. A., Heunks, L. M., Pigmans, C. J., Verheugt, F. W., Brouwer, R. M., and Dekhuijzen, P. N. (27) Diaphragm single-fiber weakness and loss of myosin in congestive heart failure rats. Am. J. Physiol. Heart Circ. Physiol. 293, H819 H Adams,V.,Alves,M.,Fischer,T.,Rolim,N.P.,Werner,S.,Schutt,N., Bowen, T. S., Linke, A., Schuler, G., and Wisloff, U. (215) High intensityinterval training attenuates endothelial dysfunction in a Dahl salt-sensitive rat model of heart failure with preserved ejection fraction (HFpEF). J. Appl. Physiol. 119, Hambrecht, R., Adams, V., Gielen, S., Linke, A., Möbius-Winkler, S., Yu, J., Niebauer, J., Jiang, H., Fiehn, E., and Schuler, G. (1999) Exercise intolerance in patients with chronic heart failure and increased expression of inducible nitric oxide synthase in the skeletal muscle. J. Am. Coll. Cardiol. 33, Kubo, S. H., Rector, T. S., Bank, A. J., Williams, R. E., and Heifetz, S. M. (1991) Endothelium-dependent vasodilation is attenuated in patients with heart failure. Circulation 84, Clanton, T. L. (27) Hypoxia-induced reactive oxygen species formation in skeletal muscle. J. Appl. Physiol. 12, Supinski, G., Stofan, D., and DiMarco, A. (1993) Effect of ischemiareperfusion on diaphragm strength and fatigability. J. Appl. Physiol. 75, Barreiro, E., and Hussain, S. N. (21) Protein carbonylation in skeletal muscles: impact on function. Antioxid. Redox Signal. 12, Vol. 31 January 217 The FASEB Journal x BOWEN ET AL.