Mitochondrial complex III activity: from invasive muscle biopsies to patient-friendly buccal swab analysis

Abstract

Drug-induced mitochondrial dysfunction is a common adverse effect, particularly in case of statins-the most prescribed drugs worldwide. These drugs have been shown to inhibit complex III (CIII) of the mitochondrial oxidative phosphorylation process, which is related to muscle pain. As muscle pain is the most common complaint of statin users, it is crucial to distinguish it from other causes of myalgia to prevent unnecessary cessation of drug therapy. However, diagnosing CIII inhibition currently requires muscle biopsies, which are invasive and not practical for routine testing. Less invasive alternatives for measurement of mitochondrial complex activities are only available yet for complex I and IV. Here, we describe a non-invasive spectrophotometric method to determine CIII catalytic activities using buccal swabs, which we validated in a cohort of statin and non-statin users. Our data indicate that CIII can be reliably measured in buccal swabs, as evidenced by reproducible results above the detection limit. Further validation on a large-scale clinical setting is recommended.

Figure 1

CIII activity in dilution series of HeLa cells compared to buccal swab samples from a non-CIII deficient person based on lack of symptoms. The CIII activity expressed in mU mL⁻¹; mean ± SEM with individual data points plotted. The dilution series for HeLa cells are from one vial, the number of cells (in millions) depicted on the x-axis. One sample of four swabs was used as mentioned before. AA (antimycin A, light blue) was the inhibited or basal cytochrome C reduction capacity for each condition. If the concentration-dependent curve reached plateau, measurements were considered equal to background (= around 62,500 cells).

Figure 2

Fluorescent and spectrophotometric based CIII activities in HeLa dilution series. Fluorescence measurement in 96-well plate (A) and 384-well plate (B,C) and spectrometric measurements in 384-well plate (D) of HeLa cell dilution series (in millions of cells) and swab sample of four pooled swabs. Results are presented as mean ± SEM. Black lines are the linear regressions curves drawn through all points. Dilutions (in millions of HeLa cells) below the intersection of this regression line and the x-axis represent the minimal working range. The red dotted line is the level of CIII activity for a swab sample of a non-CIII deficient person. If this line intersects with the regression line above the intersection with the x-axis, the results were considered to fall above the detection limit of the assay.

Figure 3

CIII activities in HeLa dilution series in 384-well plate measured with spectrophotometry. Spectrophotometric measurements of HeLa cell dilution (in millions of cells) and four swabs pooled as swab sample in a 384-well plate. Pooled results of three biologically independent experiments. Results are presented as mean ± SEM with individual data points plotted. The swab sample falls within the working range of the assay if the bar is above the lowest values of HeLa cells that showed no further decline in CIII activity after increased dilutions (background from dilution 0.333 onwards).

Figure 4

Coefficient of variation for spectrophotometric CIII activities in HeLa cells and swabs. Spectrophotometric measurements for CIII activity of HeLa cells and swab samples (a uncorrected) and corrected for citrate synthase activity in swab samples (b). Results are from three individual experiments (N = 1–3) and pooled (mean). Results are presented as mean ± SEM with individual data points plotted. Inter-assay coefficient of variation (CV) is shown. CS citrate synthase.

Figure 5

CS activity corrected CIII activity and CS activity in buccal swab samples from participants of the Nijmegen Four Days Marches. (a) CIII and (c) CS in patients treated with statins (n = 60) and controls (n = 27) and (b) CIII and (d) CS in statin users with muscle complaints (n = 32), without complaints (n = 28), and non-statin users (n = 27). Activity was measured as the absorbance from the cytochrome c reduction reaction and normalised by citrate synthase (CS) activity or as CS activity alone. No statistical differences were found. Data is presented as mean ± SEM with individual data points plotted.