The effect of differentiation and TGFβ on mitochondrial respiration and mitochondrial enzyme abundance in cultured primary human skeletal muscle cells

Abstract

Measuring mitochondrial respiration in cultured cells is a valuable tool to investigate the influence of physiological and disease-related factors on cellular metabolism; however, the details of the experimental workflow greatly influence the informative value of the results. Working with primary cells and cell types capable of differentiation can be particularly challenging. We present a streamlined workflow optimised for investigation of primary human skeletal muscle cells. We applied the workflow to differentiated and undifferentiated cells and we investigated the effect of TGFβ1 treatment. Differentiation of myoblasts to myotubes increased mitochondrial respiration and abundance of mitochondrial enzymes and mitochondrial marker proteins. Differentiation also induced qualitative changes in mitochondrial protein composition and respiration. TGFβ1 reduced complex IV protein MTCO1 abundance in both myoblasts and myotubes. In myoblasts, spare electron transport system (ETS) capacity was reduced due to a reduction in maximal oxygen consumption. In TGFβ1-treated myotubes, the reduction in spare ETS capacity is mainly a consequence of increased oxidative phosphorylation capacity and complex III protein UQCRC2. Taken together, our data shows that it is important to monitor muscle cell differentiation when mitochondrial function is studied. Our workflow is not only sensitive enough to detect physiological-sized differences, but also adequate to form mechanistic hypotheses.

Figure 1

Omitting serum during differentiation reduces the number of non-fused cells. Myoblasts were grown to subconfluence and subsequently differentiated for five days in medium with 2% FBS or without FBS. (a) Cells were subsequently stained for skeletal muscle myosin (green, MYH7 or MYH1 + 2) or CD56 (green) and TOPRO (red, nuclei) and images were obtained at 100x magnification. (b–d) mRNA abundance of PPARGC1A, MYH1 and MYH7 related to RPS13 as measured by qPCR. Lines indicate points belonging to the same biopsy. Two-sided, paired t-test. n = 4; **p < 0.01.

Figure 2

Workflow for characterisation of mitochondrial function using high resolution respirometry and immunoblot. (a) Myoblasts were seeded on 15 cm dishes and after 80–90% subconfluence differentiated for five days to myotubes. Stimulation with TGFβ1 was performed two days before subconfluence or the last two days of differentiation. (b) Cells were collected by scraping in 1 ml MirO5. Myoblasts were counted and 400,000 cells were collected for protein extraction while 800,000 cells were used for respirometry. For myotubes, 400 µl scraped lysate was pelleted and protein was extracted and quantified. A volume containing 200 µg cellular protein was then used for respirometry. Protein extracts were used for immunoblot analysis. (c) High respiration respirometry of myoblasts and myotubes was carried out using the shown injection order. A representative respiration measurement of myotubes, normalised to mg protein, is shown. Different states of respiration are calculated. The states indicated at “data collection” denote where the measurements reported in this manuscript were obtained. LEAK: respiration corresponding to proton leak. ME: malic-enzyme dependent respiration. FAO: fatty acid oxidation, increase after addition of octanoylcarnitine. TCA: respiration, after addition of all substrates that need to pass through the tricarboxylic acid cycle. OX: oxidative phosphorylation (OXPHOS) after addition of succinate and all TCA substrates. ETS: maximal capacity of the electron transport system after uncoupling. CII: complex II respiration after blocking complex I by rotenone addition. ROX: non-mitochondrial/residual respiration after blocking complex III with antimycin A.

Figure 3

Mitochondria in myotubes are more abundant and qualitatively different from mitochondria in myoblasts. Oxygen consumption of myoblasts and myotubes was measured according to the scheme in Fig. 2. Shown are only cells without TGFβ1 treatment. (a) Respiration normalised to total protein content post measurement. (b) Respiration normalised to total CS content post measurement. (c) CS content normalised to total protein content, both after measurement, and standardised to myoblasts. (d) Relative increase in respiration by succinate addition (e) β-oxidation respiration relative to OXPHOS capacity. (f) Respiration after addition of pyruvate relative to OXPHOS capacity. (g) Respiration after addition of rotenone relative to OXPHOS capacity. (h) Maximal FCCP respiration relative to OXPHOS capacity. (a,b) Repeated measures two-way ANOVA. (c to h) Paired, two-sided t-tests. *p < 0.05, **p < 0.01, ***p < 0.01. n = 4, mean ± SD.

Figure 4

Myotubes contain more mitochondria-specific proteins than myoblasts. Proteins were extracted from myoblasts and myotubes and characterised by immunoblot according to Fig. 2. Only pre-respirometry samples are shown. (a) Abundance normalised to protein loaded of mitochondrial abundance markers (CS and HSP60), respiratory chain complex enzymes (UQCRC2, MTCO1 and ATP5), β-oxidation enzymes (HADHA and MCAD), and RPS6 as a housekeeping protein (b) Abundance normalised to CS. Only vehicle-treated cells are shown. All values were standardised to myoblasts protein abundances. Two-sided t-tests, *p < 0.05, **p < 0.01, ***p < 0.001. n = 3–4, mean ± SD. Due to missing CS values, less data points could be included for (b) than for (a).

Figure 5

TGFβ1 influences respiratory chain activity and spare capacity in myoblasts and myotubes. Primary human myoblasts and myotubes were treated with TGFβ1 (0.25 or 0.5 ng/ml) or SB431542 (10 μM) for 48 hours and processed according to Fig. 2 and high resolution respirometry was carried out. (a) Cytochrome C effect calculated by dividing respiration after addition of cytochrome C by respiration after succinate addition. (b) Respiration rates after the indicated additions in myoblasts and (c) myotubes. (d) Spare respiratory capacity calculated by dividing the difference between ETS and OXPHOS capacity by OXPHOS capacity. For (a and d) one-way ANOVA was performed for each differentiation state. For (b and c) two-way ANOVA was performed. *p < 0.05, **p < 0.01. ***p < 0.001. n = 3–4, mean ± SD.

Figure 6

TGFβ1 modifies respiration in myoblasts by influencing MTCO1. Primary human myoblasts and myotubes were treated with TGFβ1 (0.25 or 0.5 ng/ml) or SB431542 (10 μM) for 48 hours and processed according to Fig. 2. Pre-respirometry protein samples were analysed by immunoblot. (a) MTCO1 abundance per mg protein loaded. (b) MTCO1 normalised to CS abundance. (c) Correlation between MTCO1/CS and ETS capacity (Fig. 4c) in myoblasts. (d) UQCRC2 abundance per µg protein loaded. (e) UQCRC2 normalised to CS abundance. (f) Correlation between UQCRC2/CS and ETS capacity (Fig. 4d) in myotubes. (a,b,d,e) Two-way ANOVA, *p < 0.05, ***p < 0.001, n = 3–4, mean ± SD. (c,f) Linear regression modelling using log-transformed data. Representative blot examples can be found in Figure S2.

Figure 7

TGFβ1 influences differentiation and MTCO1 abundance. Primary human myoblasts and myotubes were treated with TGFβ1 (0.25 or 0.5 ng/ml) or SB431542 (10 μM) for 48 hours and processed according to Fig. 2. Pre-measurement samples were analysed by immunoblot. (a and b) TGFBI, normalised to RPS6. (c) MYH7, normalised to RPS6 and standardised for donor. (d) Correlation of TGFBI (normalised, log-transformed) and MYH7 (normalised, standardised, log-transformed) in myotubes. (e,f) Correlation of TGFBI and MTCO1/CS in myoblasts and myotubes. For (a to c) one-way ANOVA and for (d to f) linear regression modelling was used. *p < 0.05, **p < 0.01. ***p < 0.001. n = 4 for TGFBI, n = 3 for MYH7, mean ± SD. For unknown reasons, the pre-measurement samples for one donor gave no MYH7 signal, despite the post-measurement samples doing so.