Chemotherapeutic drug screening in 3D-Bioengineered human myobundles provides insight into taxane-induced myotoxicities

Figures

• 

  Figure 1. Nanomolar concentrations of TAX alter α-tubulin acetylation, glucose homeostasis, and cellular respiration in human primary myotubes (A–D) Western blot analysis and representative blots of total α-tubulin (A), acetylated K40 α-tubulin (B), the calculated ratio (C), and γ-tubulin (D) were performed using lysates of primary myotubes exposed to a regimen of 3-h treatments with TAX concentrations of 0, 0.1, 1, 10, or 100 nM, for three consecutive days. Each 24-h period is considered as one cycle. (E) Representative blots showing TAX-induced acetylation of α-tubulin persists up to 7 days post-treatment. (F) Lactate efflux rates calculated from sampled media collected from well (each containing four bundles) after each 24-h TAX cycle. (G) Time course of oxygen consumption rates, OCR (top), and extracellular acidification rate, ECAR (bottom), during a mitochondrial stress test in primary myotubes 24 h after treatment. (H and I) (H) Basal mitochondrial OCR (basal OCR – OCR Rotenone/Antimycin A) and (I) maximal respiratory capacity (OCR FCCP – OCR Rotenone/Antimycin A) are expressed as the % relative to CTL owing to the variability between cell batches. (J) Reserve respiratory capacity…

• 

  Figure 2. Experimental workflow of 3D-myobundle generation and treatment with pharmacokinetic-mimicking regimens (A) Study design overview. Human myogenic cells isolated from biopsied muscle samples are expanded in 2D-culture (myoblasts), and utilized to generate 3D-myobundles using a hydrogel-molding technique (300,000 cells per myobundle) in PDMS molds as previously described (Madden et al., 2015). Shift days S6–9 refer to the number of days after switching from growth media to differentiation media. Once compacted, myobundles anchored to the frames are removed for free-floating culture and drug treatments. Following 5 days of differentiation, 3D-myobundles were exposed to either one of four treatment regimens, CTL, TAX alone, TAX + DOX, or DOX alone. The regimens consisted of three cycles with a 3 h peak-dose, followed by a 21 h low-dose treatment with TAX +/− DOX, for three consecutive days (S6–S8). TAX dose was (1 μM × 3 h + 10 nM × 21 h) × 3 days; and DOX dose (0.5 μM × 3 h + 50 nM × 21 h) × 1 day. The single dose of DOX was administered in the second cycle only (at S7). (B) Immuno-staining of 3D-myobundles: (1) fibroblasts expressing vimentin are shown in the periphery of the bundle…

• 

  Figure 3. TAX and DOX cause distinct abnormalities in the 3D-myobundle structure and expression of OXPHOS proteins Western Blot analysis of total α-tubulin (A) and acetylated K40 α-tubulin (B), their ratio (C), and cellular localization assessed by immuno-histochemistry confocal imaging (D). α-Tubulin (right) and acetylated-K40 α-tubulin (left) are both shown in red, and nuclei in blue. Scale bars represent 25 μm. Protein abundance of myosin heavy chain (E), γ-tubulin (F), and voltage-dependent anion channel (VDAC) (G) was determined by western blot analysis. Data are means ± SEM, n = 8–12 per group, where each data point is a homogenate of two myobundles. ∗p < 0.05 ∗∗∗p < 0.0005 vs CTL from one-way ANOVA analysis followed by Dunnett's test. The representative blots shown in panel (H) correspond to one batch of treated myobundles (n = 4 per group). The experiment was repeated with three independent batches of myobundles, all derived from a single donor, as detailed in the STAR Methods.

• 

  Figure 4. TAX disrupts glucose homeostasis in 3D-myobundles (A and B) Glucose consumption determined as 2-DOG uptake (A) and lactate efflux rates (B) were determined from media sampled every 24 h following each treatment cycle. (C) Linear relationship between glucose consumption and lactate efflux rates is shown at S9 after treatments. N = 5–6 per group, where each data point represents a sampled well containing four myobundles. (D) Western blot analysis of GLUT1 expression. Data are means ± SEM, n = 8 per group, where each data point is a homogenate of two myobundles. ∗p < 0.05, ∗∗p < 0.005 vs CTL from one-way/two-way ANOVA analysis followed by Dunnett's test. (E and F) Mass spectrometry-based metabolite profiling of myobundles at S9 after respective treatments, showing levels of acetyl-carnitine (E), and TCA cycle intermediates (F). (G) Western blot analysis of total and phosphorylated pyruvate dehydrogenase (PDH), expressed as the ratio of p-PDH/total PDH. N = 4 per group, with each data point being homogenates of two myobundles. ∗p < 0.05 vs CTL from one-way ANOVA followed by Dunnett's test for all panels, or t-test (F, G). Experiments A–D, G versus E–F were performed using myoblasts…

• 

  Figure 5. TAX and DOX disrupt mitochondrial function in 3D-myobundles via distinct mechanisms (A) Oxygen consumption (JO2) determined via Oroboros O2K respirometry after the addition of 10-mM pyruvate, 2.5-mM malate, sequential additions of ADP (50 μM–4 mM), 10-mM succinate, and 20 uM cytochrome c. (B) Respiratory response (% JO2) following cytochrome c addition. (C–G) (C) Analysis of ADP kinetics via ADP titration. Data fitted to the Michaelis–Menten function, and the corresponding double reciprocal plots are depicted in (D)–(G). TAX changes the KMADP, inducing a biphasic response, whereas DOX decreases maximal respiratory capacity (Vmax). Data are means ± SEM, n = 24 CTL, 24 TAX, and 14 TAX + DOX and DOX, with each sample representing two myobundles in the O2K chamber. ∗p < 0.05, ∗∗p < 0.005 vs CTL from one-way ANOVA analysis followed by Dunnett's test. (H) Protein abundance of mitochondrial respiratory complexes I-V measured by western blot analysis. Data are means ± SEM, n = 12 per group, where each data point is a homogenate of two myobundles. ∗p < 0.05 vs CTL from one-way ANOVA analysis followed by Dunnett's test. (I) Representative blot from western blot analysis in H, corresponding…

• 

  Figure 6. DOX but not TAX decreases tetanic force generation in 3D-myobundles Experiments were performed 24 h after drug treatments at S9, with electrical stimulation for 50 s at 20 Hz. (A) Myobundle cross-sectional area was calculated from the diameter measured from the top view during experiment. (B and C) (B) Tetanic force-time traces, and (C) force-time traces normalized to the peak tetanic force. (D and E) (D) Calculated tension time integral (area under the curve) from force-time plots, and (E) the respective values normalized to the myobundle cross-sectional area. (F and G) (F) Peak tetanic force (maximal force) and (G) peak force normalized to the myobundle cross-sectional area. (H–J) (H) Time to reach peak tetanic force. Analysis of fatigue included (I) T1/2 fatigue calculated as delta time to reach 50% of the peak force during a continuous tetanic contraction, and (J) fatigue slope from Figure 6C in the 30 s following peak force. Data are means ± SEM, n = 5 CTL, and n = 6 bundles per treatment group, with each data point specific to individual 3D-myobundles. ∗p < 0.05, ∗∗p < 0.005 vs CTL from one-way ANOVA analysis followed by Dunnett's test. Experiments were performed using…

• 

  Figure 7. Working model of TAX-induced myotoxicity TAX-mediated stabilization of the microtubule network leads to marked acetylation of α-tubulin, impaired glucose uptake and metabolism, along with dysregulated ATP/ADP transport, possibly via the blockage of VDAC.