Simvastatin Effects on Skeletal Muscle

Relation to Decreased Mitochondrial Function and Glucose Intolerance

Participants

Twenty male subjects were recruited to participate in the study: 10 patients with hypercholesterolemia, in medical treatment with simvastatin (10 to 40 mg/day: 10 mg/day, n = 1; 20 mg/day, n = 4; and 40 mg/day, n = 5) for at least 12 months (average: 5 years) (recruited via the Unit for Inherited Cardiac Diseases, Department of Cardiology, Rigshospitalet), and 10 healthy control subjects. The groups were matched for age, weight, body mass index, fat percentage (total and abdominal), and maximal oxygen uptake (Vo_2MAX) (Table 1). The patients with hypercholesterolemia took 1 dose of medication per day in the evening and did not take any other medication, and control subjects did not take any medication. Medication was not withdrawn prior to the test days. Subjects with a family history of type 2 diabetes were excluded from the study. Muscle pain was evaluated with the visual analog scale. After the subjects were recruited and the blood and skeletal muscle were analyzed, 1 of the control subjects turned out to have impaired glucose tolerance; therefore, data from this subject were omitted, leaving 9 in the control group.

The ethics committee of the municipality of Copenhagen and Frederiksberg in Denmark approved the study protocol (H-4-2009-095), and oral and written consent were obtained from each participant in accordance with the Helsinki Declaration.

Experimental protocol

Subjects reported to the laboratory at 8:00 am after an overnight fast (10 to 12 h) on 2 separate days. The subjects were instructed to abstain from alcohol and strenuous exercise the day before each visit. On the first day, a medical history was obtained, followed by a standard 2-h OGTT with blood samples drawn at baseline (fasting) and every 30 min until 2 h for glucose and insulin measurements. A resting electrocardiogram was then recorded to exclude individuals with signs of coronary ischemia, followed by measurement of body composition using a dual-energy X-ray absorptiometry scan (Lunar Prodigy Advance, Lunar, Madison, Wisconsin). Finally, an incremental cycling test was performed to determine Vo_2MAX (Jaeger ER 800, Erich Jaeger, Würzburg, Germany). Achievement of Vo_2MAX was accepted when a respiratory exchange ratio >1.15 and/or maximal heart rate (220 − age) was present, or a leveling off or decline in Vo₂ was reached. For further characterization of fitness, the daily physical activity was evaluated with the International Physical Activity Questionnaire (24). On day 2, subjects were placed in a supine position; after 30 min, a blood sample was drawn, and a muscle biopsy was obtained from the m. vastus lateralis.

Analytical procedures

Mitochondrial Respiration

Mitochondrial respiratory capacity was measured in permeabilized skeletal muscle fibers. The details of the procedure have been described previously (30). Briefly, all respiratory measurements were carried out in duplicate after hyperoxygenation to avoid any potential oxygen limitation to respiration. Five different substrate and inhibitor protocols were employed simultaneously (Oxygraph-2k, Oroboros Instruments, Innsbruck, Austria).

Protocol 1 (evaluating complex I and complex II OXPHOS capacity): State 2 respiration was assessed with malate (2 mmol/l), glutamate (10 mmol/l), and pyruvate (5 mmol/l), followed by state 3 respiration (5 mmol/l, adenosine diphosphate [ADP]). Rotenone (0.5 μmol/l) was added to inhibit complex I, followed by succinate (10 mmol/l) as a substrate for complex II, then cytochrome c (10 μmol/l) was added to control for outer mitochondrial membrane integrity. Malonate (5 mmol/l) was used to inhibit complex II, and finally, antimycin A (2.5 μmol/l) was added to inhibit complex III.

Protocol 2 (evaluating complex I and complex I + II OXPHOS capacity): State 2 respiration was assessed with malate (2 mmol/l), glutamate (10 mmol/l), and pyruvate (5 mmol/l), followed by state 3 respiration (5 mmol/l, ADP). Simultaneous electron input into complex I + II (maximal OXPHOS capacity) was assessed with succinate (10 mmol/l), followed by cytochrome c (10 μmol/l). Finally, oligomycin (2 μg/ml) was added to inhibit complex V.

Protocol 3 (evaluating complex IV OXPHOS capacity): Complex IV (cytochrome c oxidase) activity was measured as follows: ADP (5 mmol/l) was added followed by cytochrome c (10 μmol/l) and antimycin A (2.5 μmol/l) inhibiting complex III, finally ascorbate and N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) was added (2 mmol/l and 0.5 mmol/l, respectively).

Protocol 4 (evaluating complex I substrate sensitivity): State 2 respiration was assessed with malate (2 mmol/l), followed by state 3 respiration (5 mmol/l, ADP). Subsequently, glutamate (complex I linked substrate) was titrated in 11 steps (0.1 to 96.0 mmol/l), followed by cytochrome c (10 μmol/l).

Protocol 5 (evaluating complex II substrate sensitivity): State 2 respiration was assessed with malate (2 mmol/l) followed by state 3 respiration (5 mmol/l, ADP). Glutamate (10 mmol/l) (complex I substrate) was added, and succinate (complex II linked substrate) was titrated in 10 steps (0.1 to 48.0 mmol/l), followed by cytochrome c ((10 μmol/l). See Figure 1 for an overview of the used substrates and inhibitors in relation to the mitochondrial electron transport chain.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Simplified View of the Electron Transport Chain Illustrating the Electron Carrier Function of Q₁₀

Reactive oxygen species–producing complexes are emphasized. Below each complex is listed the substrates and inhibitors that were used. ADP = adenosine diphosphate; Asc = ascorbate; ATP = adenosine triphosphate; Pi = phosphate; Q₁₀ = coenzyme Q₁₀; TMPD = N,N,N′,N′-tetramethyl-p-phenylenediamine.

Blood samples

Blood samples were collected in chilled tubes and immediately centrifuged (2,500 g; 4°C) for 10 min. Plasma was stored at −20°C or −80°C before analysis. The analysis of plasma glucose, insulin, nonesterified fatty acids, cholesterol, low-density lipoproteins, high-density lipoproteins, and glycated hemoglobin (HbA_1c) was done as previously described (31). Glucose and insulin concentrations during the OGTT were used to calculate the Cederholm index of peripheral insulin sensitivity (32), which has been validated against the hyperinsulinemic–euglycemic clamp (33). Lactate and CK levels were analyzed enzymatically on an automated clinical chemical analyzer (COBAS 501, Roche, Mannheim, Germany).

Statistics and calculations

Data are presented as mean ± SD in the text and in the tables, and presented in all figures as mean ± SE. p < 0.05 was considered significant. Statistical analysis of differences in mitochondrial OXPHOS capacity between the groups was carried out with a 2-way analysis of variance for repeated measures. The restrictive assumptions, normality and equal variance, were checked before the statistical analysis was conducted. Significant main effects or interactions were further analyzed by the Holm-Sidak post hoc test. If normality and equal variance test failed, the data were transformed and reanalyzed; this has been stated in the figure legend if this was the case. Differences between the 2 groups were evaluated using a 1-way analysis of variance. If normality of the data was not present, the data were analyzed using a nonparametric test, and further analyzed by the Dunn, Tukey, or Kruskal-Wallis test. In this situation, data are summarized with medians and quartiles (median [interquartile range]) (Tables 1 and 2,⇓ and in the Results section). Pearson′s correlation analysis was performed to establish the presence of correlations. All statistical analysis was performed using the software program SigmaStat 3.1 (Systat Software, San Jose, California).

The calculation of substrate sensitivity (C₅₀) and OXPHOS capacity has been described previously (31).

Muscle characteristics

Q₁₀ protein content was reduced (p = 0.05) in patients compared with control subjects (Table 2). The antioxidant (catalase [p < 0.05], MnSOD [p < 0.01], and glutathione peroxidase [p < 0.05]) as well as UCP3 (p < 0.01) protein content was significantly reduced in patients compared with control subjects (Table 2). Complex IV content was significantly reduced (p < 0.05), whereas complex V showed a tendency (p = 0.07) to be reduced in patients compared with control subjects, but complex I to III protein content was similar between the groups (Table 2). CS activity, VDAC protein content, and cardiolipin content were similar between groups. When VDAC protein content was expressed per CS activity (VDAC/CS activity) or cardiolipin content (VDAC/cardiolipin content), patients had a significantly (p < 0.05) higher ratio compared with control subjects (data not shown); this could indicate that VDAC protein content is increased due to simvastatin treatment. A tendency (p = 0.06) towards lower HAD activity was seen in patients compared with control subjects (Table 2); the same was seen in HAD protein content (p = 0.07) in patients compared with control subjects. Skeletal muscle free cholesterol was similar in the 2 groups (Table 2). Skeletal muscle fiber type composition was different between the groups (Table 2).

Mitochondrial respiration

OXPHOS capacity with complex I–linked substrates was not different between control subjects and patients treated with simvastatin (26 ± 6 pmol/mg/s vs. 24 ± 8 pmol/mg/s, respectively) (Figs. 3 and 4A).⇓⇓ However, substrate sensitivity was significantly higher (p < 0.05) in the patients compared with control subjects when glutamate (complex I–linked substrate) was used as substrate (Fig. 4B). Thus, C₅₀ was lower in the patients (0.98 ± 0.23 mmol/l) compared with the control subjects (1.37 ± 0.51 mmol/l) (p < 0.05) (Fig. 4B). A tendency (p = 0.076) for a drug dose effect was seen in substrate sensitivity when glutamate was used as substrate; the patients that were treated with 40 mg of simvastatin per day (n = 5) had a higher mitochondrial substrate sensitivity (complex I–linked substrate) compared with patients treated with 10 to 20 mg of simvastatin per day (n = 5) and with the control subjects (0.84 ± 0.06 mmol/l vs. 1.12 ± 0.26 mmol/l vs. 1.37 ± 0.51 mmol/l). When succinate (complex II–linked substrate) was used as a substrate, no differences were seen in OXPHOS capacity (Fig. 3) or in substrate sensitivity (Fig. 5B) between the groups. However, maximal OXPHOS capacity with dual electron input to complex I and II (feeding electrons into the Q intersection) was significantly higher (p < 0.01) in control subjects compared with patients treated with simvastatin (64 ± 10 pmol/mg/s vs. 54 ± 8 pmol/mg/s, respectively) (Figs. 3 and 5A). When complex I was inhibited (protocol 1, inhibition with rotenone), no difference was seen in complex II OXPHOS capacity (Fig. 3) between control subjects and patients treated with simvastatin (60 ± 9 pmol/mg/s vs. 57 ± 15 pmol/mg/s, respectively). No difference was seen in complex IV OXPHOS capacity between control subjects and patients treated with simvastatin (198 ± 56 pmol/mg/s vs. 164 ± 49 pmol/mg/s, respectively) (Fig. 3). No cytochrome c response was observed in any of the protocols (data not shown), implying that the outer mitochondrial membrane was intact after the experimental procedure.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Mitochondrial OXPHOS Capacity for Complexes I, II, I + II, and IV

Oxidative phosphorylation (OXPHOS) capacity (measured as O₂ flux rates per mg tissue) in skeletal muscle from patients with hypercholesterolemia (solid bars) and control subjects (open bars) is shown. The following substrates were used: complex I: malate, glutamate, pyruvate, ADP; complex II: malate, glutamate, pyruvate, ADP, rotenone (inhibiting complex I), succinate; complex I + II: malate, glutamate, pyruvate, ADP, succinate; and complex IV: ADP, cytochrome c, antimycin A, TMPD + Asc. The y-axis to the right refers to the O₂ flux rates for complex IV. Data are mean ± SE. *p < 0.05 for patients versus control subjects. Abbreviations as in Figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Mitochondrial OXPHOS Capacity With Glutamate

(A and B) O₂ flux rates per mg tissue in skeletal muscle from patients with hypercholesterolemia (solid circles) and control subjects (open circles) are shown for the following protocol: malate, ADP, and glutamate titration (0.1 to 96.0 mmol/l). (A) Glutamate titration where Michaelis-Menten kinetics can be seen (entire titration protocol). (B) Extract of the respiration curve highlighting respiration for glutamate concentration (0.1 to 2.5 mmol/l), which makes it possible to appreciate the substrate kinetics (C₅₀). Drop lines (B) indicate C₅₀ (x-axis) corresponding to half maximum OXPHOS capacity (y-axis), respectively. In order to perform the appropriate statistical testing, the data were transformed before they were analyzed. Data are mean ± SE. *p < 0.05 for patients versus control subjects. Abbreviations as in Figures 1 and 3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Mitochondrial OXPHOS Capacity With Succinate

(A and B) O₂ flux rates per mg tissue in skeletal muscle from patients with hypercholesterolemia (solid circles) and control subjects (open circles) are shown for the following protocol: malate, ADP, glutamate, and succinate titration (0.1 to 48.0 mmol/l). (A) Succinate titration where Michaelis-Menten kinetics can be seen (entire titration protocol). (B) Extract of the respiration curve highlighting respiration for succinate concentration (0.1 to 8.0 mmol/l), which makes it possible to appreciate the substrate kinetics (C₅₀). Drop lines (B) indicate C₅₀ (x-axis) corresponding to half maximum OXPHOS capacity (y-axis). In order to perform the appropriate statistical testing, the data were transformed before they were analyzed. Data are mean ± SE. *p < 0.05 for patients versus control subjects. Abbreviations as in Figures 1 and 3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Representative Western Blots Are Presented

The bands for the different proteins are from the same 2 subjects (a patient and a control). GPX = glutathione peroxidase; HAD = beta-hydroxyacyl-CoA dehydrogenase; MHC = myosin heavy chain; MnSOD = manganese superoxide dismutase; Q₁₀ = coenzyme Q₁₀; UCP3 = uncoupling protein 3; VDAC = voltage-dependent anion channel.

A weak positive correlation (r = 0.46, p < 0.05) was observed between maximal OXPHOS capacity (complex I + II–linked substrates) and Q₁₀ content. This correlation was not observed between OXPHOS capacity with complex I or complex II–linked substrates alone and Q₁₀ content.

Q₁₀ content and mitochondrial OXPHOS capacity

Other studies with high simvastatin doses (80 mg/day) have reported decreased Q₁₀ protein content in muscle from simvastatin-treated patients (12,13), whereas studies using lower (20 mg/day) doses have been unable to demonstrate decreased muscle Q₁₀ protein content (14,15), but no studies have so far combined this measurement of a key protein in the mitochondrial respiratory chain with a functional (high-resolution respirometry) mitochondrial test. In doing so, we can now show that OXPHOS capacity is not different between patients and control subjects, when substrates for complex I and complex II are used separately. Only when complex I and II substrates were combined, thus maximally taxing the electron transport system, did the effect of statin treatment become apparent (Figs. 1 and 3), indicating a defect in the Q intersection when the electron transport system is challenged maximally. The fact that we were able to demonstrate this impairment in OXPHOS capacity (complex I and II substrates combined) in the patients even at these relatively low flux rates, emphasizes the potential of our finding (in vivo flux rates are much higher than these ex vivo flux rates [38]). It may be anticipated that the OXPHOS limitation observed ex vivo in the present study serves as a mechanism for the impaired exercise tolerance and fatigability seen in some patients using statins. The acute effect of statins has, however, been studied in isolated mitochondria from L6 cells (a rat skeletal muscle cell line) (39) and in skinned human and rat muscle (16) in which a toxic effect of simvastatin and depressed complex I respiration was found. Recently Kwak et al. (40) reported impaired ADP-stimulated respiration in primary human skeletal myotubes after 48 h of simvastatin treatment, and this was accompanied by increased mitochondrial oxidative stress (hydrogen peroxide emission). It may not be that surprising that a statin can produce an acute toxic effect in cells, but the fact that an impairment of function can be observed in human muscle fibers, obtained via a biopsy that is mechanically and chemically dissected and permeabilized, washed, and finally transferred to the respiration chamber, is truly remarkable. It shows that the impact of long-term treatment with the lipophilic statin simvastatin has profound functional (limitation of maximal OXPHOS capacity) and structural (decreased content of Q₁₀, complex IV, antioxidant proteins, and UCP3) impact. The present study differs from the aforementioned studies in that the patients are chronically treated with simvastatin, whereas the other studies are conducted in cell lines or the skeletal muscle is acutely treated with very high (thus not clinically relevant) concentrations of statins. Therefore, the clinical applicability of the present study is more obvious.

The decreased Q₁₀ protein content in patients in long-term treatment with simvastatin offers a mechanistic explanation for the limitation in maximal OXPHOS capacity (Figs. 1 and 3). This hypothesis requires further testing, but is to some extent supported by the correlation between Q₁₀ content and maximal OXPHOS capacity, even though this correlation was not strong and has to be confirmed in other studies.

Mitochondrial substrate sensitivity

Skeletal muscle from simvastatin-treated patients had increased mitochondrial substrate sensitivity (lower C₅₀ value) with complex I–linked substrate (glutamate) compared with muscle from the control subjects, but no difference was seen with complex II–linked substrate (succinate). The reason for this disparity may be explained by the difference in muscle fiber type composition and/or mitochondrial membrane potential. First, higher substrate sensitivity for complex I–linked substrates has previously been reported in patients with type 2 diabetes compared with healthy control subjects (31). Second, differences between type I and type II muscle fibers have been reported previously in rats, with type II fibers being more sensitive than type I (41). Accordingly, the significantly lower proportion of type I fibers in the patients compared with the control subjects in the present study may contribute to the observed difference in substrate sensitivity between the groups. Third, it has previously been reported that statins, in a dose-dependent manner, decrease the mitochondrial membrane potential (16,39). The membrane potential is mainly created by virtue of the activity of complexes I, III, and IV, and a compromised function of 1 or more of these may lower the membrane potential. Therefore, a higher complex I substrate sensitivity may be regarded as a compensatory mechanism for upholding the mitochondrial membrane potential. In support of this notion is our observation that the patients treated with the highest simvastatin dose (40 mg/day) also showed a nonsignificant tendency to have the highest substrate sensitivity compared with both patients treated with the lowest simvastatin dose (10 to 20 mg/day) and the control subjects.

The difference between simvastatin-treated patients and the control subjects in complex I–linked substrate sensitivity is a true intrinsic event in the mitochondrion, because the derived sensitivity parameter (C₅₀) of the titration studies is independent of mitochondrial density.

Muscle characteristics

In the present study, ROS production was not measured, but other studies have shown increased ROS production (3,42) and a decreased antioxidant capacity (3) with statin treatment. The latter finding is confirmed by the present study in which we found decreased protein content of catalase, manganese superoxide dismutase, and glutathione peroxidase. Mitochondrial content evaluated by CS activity, cardiolipin content, or VDAC protein content was similar between the groups (Table 2). The ratio between VDAC and CS activity or cardiolipin was significantly higher in the patients compared with the control subjects, indicating that simvastatin-treated patients have a higher abundance of VDAC content per mitochondrion. It has been reported that VDAC located in the outer mitochondrial membrane is important for the in- and outflux of Ca²⁺ of the mitochondria (43), and the up-regulated VDAC protein content in the simvastatin-treated patients could be because of the known statin-induced elevation of cytoplasmic Ca²⁺ concentration (44). In turn, an increased Ca²⁺ load in the mitochondrial matrix has been linked to increased ROS production and increased permeability transition pore opening, leading to apoptosis (40,45). Our findings of significantly reduced antioxidant capacity in combination with the known (3,40) elevated ROS production in the simvastatin-treated patients make it plausible that increased apoptosis could contribute to muscle pain in patients treated with statins. Furthermore, we observed a decreased protein content of complex IV in the patients compared with the control subjects; this was a somewhat unexpected finding.