LDL receptor knockout mice (LDLr^-/-) founders were purchased from Jackson Laboratory (Bar Harbor, ME) and the breeding colony was maintained at the Universidade Estadual de Campinas (CEMIB-Unicamp), Campinas, Brazil. LDLr^-/- mice had access to standard laboratory rodent chow (AIN 93M, PragSoluções, SP, Brazil), and water ad libitum and were housed at 22 ± 2°C on a 12h light-dark cycle. This study was performed in accordance with the Guide for the Care and Use of Laboratory Animals published by National Academy of Sciences and with the approval of University Committee for Ethics in Animal Experimentation (protocol # 3401-1). Chemicals were purchased from Sigma (St. Louis, MO, USA).

Thirty-day-old male LDLr^-/- mice received pravastatin sodium (Medley) diluted in the drinking water (400 mg/L) during 2 or 3 months according to Lorza-Gil et al. (2016). The estimated pravastatin dose of 40 mg/Kg body weight per day was based on average consumption rate measurements (3.5 mL/day). Controls received filtered tap water without pravastatin. Additional groups of mice were treated with 2% creatine supplemented into standard diet (AIN 93M, PragSoluções, SP, Brazil) without alteration of total calories during the last 15 days of pravastatin treatment.

Blood samples were collected with heparin from LDLr^-/- mice tail between 8 and 9 am after a 12-h fasting. Samples were centrifuged and plasma was utilized for cholesterol measurement using a standard commercial kit (Roche Diagnostics) according to the manufacturer’s instructions. Plasma cholesterol levels in pravastatin treated LDLr^-/- mice were significantly reduced compared to untreated mice (437 ± 57 vs. 390 ± 36, respectively, P < 0.05).

Plantaris and soleus muscles were harvested from LDLr^-/- mice and placed on ice-cold buffer containing 10 mM Ca-EGTA buffer (2.77 mM of CaK₂EGTA + 7.23 mM of K₂EGTA, free concentration of calcium 0.1 mmol/L), 20 mmol/L imidazole, 50 mmol/L K⁺/ 4-morpholinoethanesulfonic acid, 0.5 mmol/L dithiothreitol, 7 mmol/L MgCl₂, 5 mmol/L ATP, 15 mmol/L phosphocreatine, pH 7.1. Individual fiber bundles from three to 5 mg of soleus or plantaris skeletal muscle were separated with forceps. Samples were permeabilized in ice-cold buffer containing saponin (50 μg/mL) during 30 min, gently stirred and washed three times with MiR05 medium (60 mmol/L potassium lactobionate, 0.5 mmol/L EGTA, 3 mmol/L MgCl₂, 20 mmol/L taurine, 10 mmol/L KH₂PO₄, 20 mmol/L HEPES, 110 mmol/L sucrose, 1 g/L BSA, pH 7.1) at 4°C. Samples were dried with filter paper and weighted (Kuznetsov et al., 2008; La Guardia et al., 2013).

Oxygen consumption was evaluated in permeabilized skeletal muscle according to Kuznetsov et al. (2008) and La Guardia et al. (2013) with slight modifications. Permeabilized tissues were added to MiR05 medium without EGTA containing Ca²⁺ (4.4 μM) at 37°C supported with 10 mM glutamate plus 5 mM malate in a high-resolution oxygraph OROBOROS (Innsbruck, Austria). ADP (400 μM), oligomycin (0.63 μM), and FCCP (0.6 μM) were added during the experiments. Some analyses were evaluated in the presence of EGTA (500 μM), CsA (0.83 mM), ruthenium red (1 μM) or coenzyme Q₁₀ (10 μM). Figure 1A shows the typical experimental respiratory profile.

Plantaris and soleus muscles were harvested from LDLr^-/- mice and homogenized in 9 volumes (1:10, w/v) of 20 mM sodium phosphate buffer, pH 7.4 containing 140 mM KCl. Homogenates were centrifuged at 1000 × g for 10 min at 4°C for nuclei and cell debris removal (Evelson et al., 2001). The pellet was discarded and the total supernatant was used for enzymatic activity determination.

Glutathione peroxidase, glutathione reductase, superoxide dismutase and peroxiredoxin were determined according to Wendel (1981), Carlberg and Mannervik (1985), Marklund (1985), and Kim et al. (2005), respectively. Catalase activity was analyzed by measuring the absorbance decrease at 240 nm according to Aebi (1984) and one unit (U) of the enzyme is defined as the metabolization of 1 μmol of H₂O₂ per min. The specific activity was calculated and expressed as U/mg protein. The activity of aconitase was measured according to Morrison (1954), following the reduction of NADP⁺ at wavelengths of excitation and emission of 340 and 466 nm, respectively. Aconitase activity was expressed as nmol NADPH/min/mg protein. Protein content was measured according to Lowry et al. (1951) using bovine serum albumin as standard.

Catalase mRNA expression was quantified by RT-qPCR using GAPDH housekeeping gene to normalize each sample. Plantaris muscles were harvested from LDLr^-/- mice, total RNA was extracted using TRIzol (Thermo Fisher Scientific) following manufacturer’s instructions. Total RNA was used as template for cDNA synthesis in a reaction with oligo(dT)₁₈ primer (Exxtend Biotecnologia) and SuperScript III Reverse Transcriptase (Thermo Fisher Scientific) at 50°C for 60 min. The enzyme was then inactivated at 70°C for 15 min. Real-time PCR was performed on a Rotor Gene system (Qiagen, Hilden, Germany) using Rotor Gene SYBR Green PCR kit (Qiagen) and the following cycling conditions: 95°C for 5 s and 60°C for 10 s. Data acquisition was performed during the annealing step at 60°C. Primers used in qPCR were as follows: CAT (98 bp), 5′ GTTGAACGAGGAGGAGAGG 3′ (forward) and 3′ GTGAAATTCTTGACCGCTTTC 5′ (reverse); GAPDH (175 bp), 5′ GCACCACCAACTGCTTAGC 3′ (forward) and 3′ ATGCAGGGATGATGTTCTGG 5′ (reverse). CAT and GAPDH mRNA quantification was performed twice in N = 5 animals from each group. Data were analyzed using the Delta CT method of Rotor Gene Q series Software and catalase relative mRNA expression levels were obtained by normalizing against the level of GAPDH from the same sample and conditions. Efficiencies of CAT and GAPDH qPCRs were 1.00 and 0.90, respectively. Standard curves were prepared for each run using known quantities of pGEM-T-easy plasmids (Promega) containing CAT and GAPDH genes.

Protein oxidative damage was evaluated by sulfhydryl content measurement according to Aksenov and Markesbery (2001). The reduction of 5,5′-dithio-bis (2-nitrobenzoic acid (DTNB) by thiols present in the sample generates a yellow compound (TNB) whose absorption is measured spectrophotometrically at 412 nm. Briefly, 30 μL of 10 mM DTNB and 980 μL of PBS were added to 50 μL muscle supernatant followed by a 30 min incubation at room temperature in the dark. The absorption measured was proportional to the amount of thiol groups present in the sample. Results were calculated as nmol TNB/mg of protein.

Plantaris muscles were removed from LDLr^-/- mice and rapidly homogenized with a methanol:H₂O (50:50) solution under sonication. Resulting homogenates were filtered through a 0.22 μm nylon membrane; 10 μL of the filtrate were further diluted in methanol:H₂O (50:50) solution containing 0.1% formic acid to a final volume of 1 mL. Samples were directly infused in an ESI-LTQ-XL Orbitrap Discovery instrument (Thermo Scientific, Bremen, Germany). Typical operating conditions were as follows: sheath gas at 10 arbitrary units, 4.5 kV and m/z range of 50–1000 in the positive ion mode. Structural elucidation was carried out using mass accuracy as the main parameter, with a mass shift (error) less than 2 ppm. Spectral data were submitted to a partial least squares discriminant analysis (PLS-DA) using MetaboAnalyst 3.0 (Xia et al., 2015) to identify markers for each condition. Data normalization was performed using log transformation and range scaling. The selected ions were then researched in the Lipid Maps database, where oxidized species were identified.

Results are presented as mean ± standard deviation of at least eight mice Data were analyzed using one-way analysis of variance (ANOVA) followed by the post hoc Tukey’s multiple comparison test when F was significant. The Student’s t-test for unpaired samples was also used for two-means comparisons. Differences between groups were rated significant at P < 0.05. All analyses were carried out using the GraphPad software.

In order to investigate the effects of pravastatin chronic treatment on mitochondrial respiration of soleus and plantaris muscle, LDLr^-/- mice received pravastatin (40 mg/kg/day) added to the drinking water during 3 months. Oxygen consumption supported by 10 mM glutamate plus 5 mM malate was evaluated in the presence of Ca²⁺ (4.4 μM) with the addition of ADP (400 μM), oligomycin (0.63 μM), and FCCP (0.6 μM) during the experiments. Figure 1A shows typical traces of mitochondrial respiration rates in all conditions. Figure 1B shows that pravastatin treatment promoted significant inhibition of mitochondrial respiration in all states: phosphorylating (ADP), resting (oligomycin) and maximal (FCCP) respiration rates of plantaris muscle in the presence of Ca²⁺. The inhibitions were 14, 24, and 40% for ADP-, oligomycin- and FCCP- stimulated respiration, respectively [n = 8; P = 0.0103; P = 0.0442; P = 0.0004]. The lower rate of FCCP-induced respiration compared to ADP-induced respiration is in agreement with recent data (Ruas et al., 2016) showing that oligomycin treatment previous to FCCP addition leads to an underestimation of maximal respiratory capacity induced by FCCP. In contrast to plantaris, no significant alterations of oxygen consumption rates were observed in soleus muscle (Figure 1C). Furthermore, no differences in oxygen consumption were observed in plantaris after only 2 months of pravastatin treatment (data not shown).

Considering that Ca²⁺ is essential for PTP opening (Hunter et al., 1976; Kowaltowski et al., 2001), and that toxic effects of statins have been associated with alterations in calcium homeostasis (Sirvent et al., 2005b; Sirvent et al., 2012), our next step was to investigate the role of Ca²⁺ on oxygen consumption of plantaris muscle of LDLr^-/- mice. For this purpose, the Ca²⁺ chelator EGTA, ruthenium red (a mitochondrial Ca²⁺ uptake inhibitor) or cyclosporin A (CsA, a permeability transition inhibitor) were added in the reaction medium before oxygen consumption measurements. Figure 2 shows that all these compounds fully reversed the mitochondrial respiration inhibition in plantaris muscle of pravastatin treated LDLr^-/- mice.

Creatine acts directly as antioxidant (Lawler et al., 2002). In addition, creatine supplementation acts on ATP/ADP ratio maintenance due to creatine kinase (CK) activation and CK is part of the protein complex that is involved in MPT regulation (Kowaltowski et al., 2001; Dolder et al., 2003; Meyer et al., 2006). Therefore, we supplemented LDLr^-/- mouse chow diet with 2% of creatine during the last 15 days of pravastatin treatment. Figure 3A shows that creatine diet supplementation prevented the inhibitory action of pravastatin on ADP- and FCCP-stimulated oxygen consumption in the presence of Ca²⁺ in plantaris muscle of LDLr^-/- mice [n = 10; P < 0.05].

CoQ₁₀, which was previously reported by us to protect against mitochondrial dysfunction caused by simvastatin in rat soleus muscle (La Guardia et al., 2013), also showed the same protective effect under the experimental in vitro conditions of mouse mitochondrial phosphorylation (ADP) and maximal respiration (FCCP) rates [n = 10; P < 0.05] (Figure 3B).

Considering that several studies claim an antioxidant activity of statins due to upregulation of antioxidant defenses (Carneado et al., 2002; Wassmann et al., 2002; Manfredini et al., 2010; Zhou and Liao, 2010), we investigated the activity of antioxidant enzymes in muscle of LDLr^-/- mice. Figure 4A shows that pravastatin treatment increased catalase activity up to 30% in plantaris muscle homogenates. In addition, creatine diet supplementation abolished the differences in catalase activity between control and pravastatin treated mice [n = 5, P < 0.05]. This increase is probably the consequence of a pravastatin effect at a post-transcriptional step or on the enzymatic catalysis, since catalase mRNA expression levels were not altered in LDLr^-/- mice muscle (Figure 4B). On the other hand, pravastatin treatment caused no differences in superoxide dismutase, glutathione reductase, glutathione peroxidase, peroxiredoxin and glucose-6-phosphate dehydrogenase activities either in plantaris or in soleus muscle (data not shown).

To verify that ROS production may occur due to pravastatin treatment, we investigated the presence of oxidized lipids in LDLr^-/- mice. Using electrospray ionization high-resolution mass spectrometry analysis and a lipidomics approach, we identified oxidized lipid markers, especially phosphatidic acid and derivatives of arachidonic acid in plantaris muscle of LDLr^-/- mice under pravastatin treatment (Table 1).

To further investigate possible oxidative damage on other cellular components, we evaluated aconitase activity, a ROS-susceptible enzyme (Tretter and Adam-Vizi, 2000), and total sulfhydryl content, a protein oxidative damage marker in plantaris muscle of LDLr^-/- mice. Both oxidative stress markers were not altered, suggesting that oxidative damage to proteins is probably not occurring in plantaris muscle of LDLr^-/- mice under pravastatin treatment (Supplementary Figure S1).