Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Detyrosinated microtubules modulate mechanotransduction in heart and skeletal muscle

Abstract

In striated muscle, X-ROS is the mechanotransduction pathway by which mechanical stress transduced by the microtubule network elicits reactive oxygen species. X-ROS tunes Ca2+ signalling in healthy muscle, but in diseases such as Duchenne muscular dystrophy (DMD), microtubule alterations drive elevated X-ROS, disrupting Ca2+ homeostasis and impairing function. Here we show that detyrosination, a post-translational modification of α-tubulin, influences X-ROS signalling, contraction speed and cytoskeletal mechanics. In the mdx mouse model of DMD, the pharmacological reduction of detyrosination in vitro ablates aberrant X-ROS and Ca2+ signalling, and in vivo it protects against hallmarks of DMD, including workload-induced arrhythmias and contraction-induced injury in skeletal muscle. We conclude that detyrosinated microtubules increase cytoskeletal stiffness and mechanotransduction in striated muscle and that targeting this post-translational modification may have broad therapeutic potential in muscular dystrophies.

Introduction

The cytoskeleton is critical for the cellular response to the mechanical environment, as it integrates and transduces mechanical energy to mechanosensitive proteins that generate biological signals. Known collectively as mechanotransduction, these pathways regulate diverse cellular physiology1 and dysregulation of these pathways is linked to disease2.

In striated muscle, microtubule (MT)-dependent mechanotransduction of reactive oxygen species (ROS) and calcium (Ca2+) signals has been characterized3,4,5,6,7. Termed ‘X-ROS’ signalling, this mechanotransduction pathway links the mechanical stress of contraction or stretch to NADPH Oxidase 2 (NoX2) generated ROS signals that target Ca2+ channels3,4,6,7,8,9.

The MT contribution to X-ROS has gained considerable attention in the context of Duchenne muscular dystrophy (DMD), in which a dense and disorganized MT network accompanies severe striated muscle dysfunction5,10. In the mdx mouse model of DMD, we found that increased MT density4,6 and cytoskeletal stiffness4 correlated with excessive X-ROS that induced detrimental Ca2+ signals that are believed to underlie cardiac arrhythmia6,9 and contraction-induced damage in skeletal muscle4.

The MT network is a dynamic structure whose density is regulated by enzymes and binding proteins that affect the equilibrium between MT filament growth and disassembly11. In the mdx mouse, we demonstrated that an acute pharmacological strategy to promote MT disassembly (colchicine) reduced X-ROS and detrimental Ca2+ signalling in vitro and proffered therapeutic benefit in vivo, implicating increased MT density in the dysregulation of mechanotransduction in DMD. In light of transcriptional evidence from human DMD samples4 indicating an upregulation of the X-ROS pathway, as well as a link between MT density and X-ROS in additional dystrophy models3,7, we initially proposed that the increase in MT network density was a pathological factor in muscular dystrophy.

Assuming force transmission (F) through the cytoskeleton is a product of the stiffness (k) and deformation (x) of the cytoskeleton (F=kx), we reasoned that alterations in stiffness would regulate the magnitude of mechanotransduction. In this regard, we found detyrosinated α-tubulin12,13 a potentially attractive target for altering cytoskeletal stiffness. Detyrosination is the cleavage of a C-terminal tyrosine residue by an unidentified tubulin carboxypeptidase (TCP)14,15, a process readily reversed by tubulin tyrosine ligase (TTL)16,17. This modification stabilizes MT filaments by limiting disassembly18,19 and can increase the association of the MT network with other cytoskeletal elements20,21,22, which could increase cytoskeletal stiffness. Intriguingly, detyrosinated α-tubulin is abundant in disease and experimental conditions in which X-ROS is elevated3,4,6,7, yet its contribution to cytoskeletal stiffness and mechanotransduction remain undefined.

Here we report that MT detyrosination profoundly regulates mechanotransduction by modifying the mechanical properties of the cytoskeleton. By reducing detyrosination, we decreased cytoskeletal stiffness and increased the speed of muscle contraction and relaxation. Furthermore, these cytoskeletal changes dramatically reduced X-ROS signalling and activation of Ca2+ channels in both heart and skeletal muscle. These effects occurred without detectable changes in MT network density, indicating a specific role for detyrosination. Consistent with this reduction in X-ROS, we show that in vivo reduction of detyrosination in a murine model of DMD proffers significant protection from contraction-induced injury in skeletal muscle and workload-induced arrhythmia of the heart, both hallmarks of DMD.

We conclude that detyrosinated MT filaments affect the mechanosensitivity of striated muscle by altering the mechanical properties of the cytoskeleton. Given the limitations associated with colchicine treatment23, the identification of a specific subset of MTs that regulate cytoskeletal stiffness and mechanotransduction is a potentially viable therapeutic option. As this therapeutic strategy has already been established in cancer clinical trials24, targeting detyrosinated α-tubulin may have broad therapeutic potential for the treatment of muscular dystrophies.

Results

Striated muscles are enriched in detyrosinated MT filaments

Immunolabeling α-tubulin reveals a MT network favouring a fenestrated organization in skeletal flexor digitorum brevis (FDB) myofibres (Fig. 1a; control), and a longitudinal axis distribution in wild-type (WT) cardiomyocytes (Fig. 1d; control), consistent with previous reports10,25,26. Microtubule detyrosination is a post-translational modification (PTM) occurring post-polymerization, and estimates of total detyrosinated α-tubulin in various cell types range from 16-20% in epithelial cells, to approximately 60% in striated muscle, with a limited fraction (<10%) in the free tubulin pool27,28. Co-labelling cardiomyocytes and skeletal myofibres with an antibody against detyrosinated α-tubulin demonstrates that this PTM is abundant in WT cells and labels a subset of the total polymerized MT network (Fig. 1a,d).

Figure 1: Detyrosinated α-tubulin is decreased following parthenolide treatment.
figure1

(a) Skeletal FDB fibres isolated from WT rats were treated with PTL, taxol, or vehicle (control) before fixation, and immunostained for α-tubulin (green) and detyrosinated α-tubulin (red). (scale bar, 10 μm). (b) The density of the fraction of α-tubulin or detyrosinated α-tubulin was quantified (see Methods) following PTL (red) or taxol (blue) treatment and expressed as a per cent change from control (DMSO (dimethylsulphoxide)). PTL and taxol significantly decreased and increased detyrosinated tubulin, respectively. No significant differences in α-tubulin were identified with either treatment (n=7 fibres per group; *P<0.05). (c) Western blots of isolated FDB fibres showing that PTL significantly decreased, and taxol increased, the amount of detyrosinated tubulin, with neither treatment affecting the total α-tubulin content (n=5 mice; *P<0.05). Each treatment was normalized to DMSO control condition. (d,e) Cardiomyocytes isolated from WT rats treated with PTL, taxol, or vehicle control as above in A (scale bar, 10 μm) and quantified as in B. The detyrosinated MT fraction was significantly decreased with PTL and increased following taxol treatment, while the density of the total α-tubulin network was unchanged by PTL treatment and taxol treatment (n=8 cells per group; *P<0.05). (f) Western blots of cardiomyocytes revealed that PTL significantly decreased detyrosinated α-tubulin while taxol increased the amount of detyrosination. Neither treatment had an effect on total α-tubulin content (n=6 mice; *P<0.05). Blots quantitated as per cent change from controls. Differences between groups were determined by using a one-way analysis of variance followed by Dunnett’s post hoc test. All values are mean±s.e.m.

Parthenolide (PTL) is a sesquiterpene lactone that inhibits the activity of the TCP29,30,31 and decreases the fraction of detyrosinated α-tubulin in vitro. A 2 hr treatment of isolated FDB myofibres or cardiomyocytes (Fig. 1a,d; PTL) with PTL results in an approximate 60% reduction in the amount of detyrosinated microtubules with no appreciable alteration in the density of the total network (Fig. 1b,e). Western blot verified this decrease in detyrosination, and confirmed that PTL did not alter the total α-tubulin content (Fig. 1c,f). As western blot measures both free and polymerized tubulin, and the majority of tubulin in striated muscle is in the free pool10,32, immunofluorescence is a more accurate indicator of the status of the polymerized network.

We also examined the effects of PTL on other common MT PTMs, namely, acetylation, polyglutamylation and the formation of Δ2 tubulin, all of which are associated with stabilized MTs with each being attributed to functional specificity in MT network regulation33,34. Although we detected polyglutamylation in FDB fibres and cardiomyocytes, the levels were unaltered by PTL treatment, as shown by western blot and immunofluorescence (Supplementary Figs 1,2). Acetylation was found in FDB fibres and levels were not altered by PTL, whereas little acetylation was detected in cardiomyocytes. We found Δ2-tubulin nearly undetectable in either cell type (Supplementary Figs 1,2). These results are consistent with a previous study of healthy adult myocytes28 and is suggestive of a unique role for detyrosination. Together, these results suggest that PTL specifically reduces the fraction of detyrosinated MT filaments, independent of filament density, tubulin content or other tubulin PTMs.

In contrast to PTL, taxol stabilizes MTs, increasing both network density and the proportion of detyrosinated MT filaments27,35. Treatment of FDB myofibres and cardiomyocytes in vitro with taxol (Fig. 1a,d; Taxol) significantly increases MT density and the degree of detyrosination by immunofluorescence (Fig. 1b,e). Western blots (Fig. 1c,f) further show that taxol increased the fraction of detyrosinated α-tubulin without altering total α-tubulin content. Taxol treatment also led to a modest increase in acetylation in skeletal myofibers, no detectable acetylation in cardiomyocytes, and little alteration to other PTMs in either cell type (Supplementary Figs 1,2) suggesting detyrosination as a potential contributor to taxol-dependent effects of MT function.

Detyrosinated MTs regulate mechanical properties of muscle

MT detyrosination increases concurrently with the increased MT density that is suggested to suppress contractility in cardiac development36,37,38 and the failing heart39,40, and is dramatically increased in diseased skeletal muscle4,5,7. However, the distinction between MT network density and MT detyrosination per se, and how each contributes to cytoskeletal mechanics, remains unclear. We therefore examined sarcomere dynamics during paced contractions in cardiomyocytes and skeletal myofibres as a gross measure of the contribution of cytoskeletal properties to cellular mechanics. In cardiomyocytes, PTL treatment had no effect on the resting sarcomere length (Fig. 2a; SL) yet significantly increased the degree of sarcomere contraction (Fig. 2a; red) and the maximal velocity of contraction and relaxation (Fig. 2b; red), findings consistent with decreased cytoskeletal resistance. In contrast, taxol decreased the velocity of contraction and relaxation, consistent with increased cytoskeletal resistance (Fig. 2b; blue) as previously described41,42. In skeletal muscle fibres, neither resting SL nor the change in SL with contraction was significantly different following PTL or taxol treatment (Fig. 3a). The maximal rate of contraction during 1 Hz stimulation was increased with PTL, yet the relaxation rate was not significantly altered (Fig. 3a). However, the time to peak contraction was significantly increased with PTL and decreased with taxol (Fig. 3a), again consistent with the MT network contributing to cytoskeletal viscoelastic resistance. Measurements of calcium transients in PTL-treated cardiac and skeletal muscle cells indicate that changes in global calcium kinetics cannot explain the altered mechanics (Figs 2c and 3c), leaving a decrease in the viscoelastic properties of the cytoskeleton as the most plausible explanation for the increased contractility.

Figure 2: Contractile mechanics in cardiomyocytes are regulated by detyrosinated microtubules.
figure2

(a) Sarcomere shortening was assayed in single cardiomyocytes under single action potential stimulation (1 Hz; field stimulation), contracting under no-load conditions (see Methods). SL measures of a control (black) and PTL (red) treated fiber are shown. Neither taxol nor PTL treatment significantly altered resting SL compared with the control (ctrl) condition. The per cent change (contraction %) in SL during contraction was significantly increased with PTL treatment but not by taxol. *P<0.05. (b) First derivative calculations of the SL shortening of a control (black) and PTL (red) treated cell expressed as a % from control. PTL significantly increased the velocity of SL shortening and relaxation while taxol decreased the velocity of each. *P<0.05 (n=36 control cells, 30 PTL cells, 30 taxol cells). (c) Ca2+ transients were assayed simultaneously with sarcomere mechanics following PTL treatment in single cardiomyocytes (n=5 cells per treatment). PTL (red) did not affect peak amplitude (measured by Fluo-3; see Methods) or the rate of decay of the global Ca2+ transient, and slightly slowed the rate of rise compared with the control condition (black/grey). *P<0.05 (n=38 control cells, 27 PTL cells). Statistical significance determined by one-way analysis of variance followed by Tukey post hoc test. All values are mean±s.e.m.

Figure 3: Mechanical properties of skeletal muscle are regulated by detyrosinated MT filaments.
figure3

(a) Sarcomere shortening was assayed in single FDB fibres contracting with single action potential stimulation (1 Hz; field stimulation) under no-load conditions (see Methods). Representative raw SL measures of control (black) and PTL (red) treated fibres are shown. Neither taxol nor PTL treatment significantly altered resting SL or the per cent change (% change) in SL compared with the control condition. (b) PTL significantly increased the maximal velocity of contraction, although relaxation kinetics were unaltered. Taxol treatment did not significantly alter either contraction or relaxation velocity. However, PTL significantly decreased, while taxol significantly increased, the time to peak SL shortening. *P<0.05 (n=10 control, 18 PTL, 14 taxol). Statistical significance determined by using a Kruskal–Wallis with Dunn’s post hoc test. (c) Following PTL treatment, Ca2+ transients (displayed as the resulting ratio, F(a.u.)) were assayed using the ratiometric dye, Indo1-PE during unloaded shortening of single FDB fibres (n=5 cells per treatment). PTL did not affect peak calcium or the kinetics of rise or decay. Statistical significance determined by Student’s t-test. All values are mean±s.e.m.

We then directly tested the contribution of α-tubulin detyrosination to cytoskeletal stiffness by assaying near-membrane mechanical properties by atomic force microscopy (AFM)4. Differentiated skeletal C2C12 myotubes were used due to their robust adhesion to a substrate that permits paired pre- and post-treatment assessments of elastic moduli (Fig. 4d,e). We confirmed that the effects of both PTL and taxol on detyrosination in skeletal myotubes were similar to their respective effects on adult myofibres and myocytes, as shown via western blot (Fig. 4a,b) and immunofluorescence (Fig. 4c). In AFM studies, acute treatment with PTL significantly reduced, while taxol significantly increased, near-membrane transversal stiffness (Fig. 4f). Taken together, the differences in contractile mechanics and near-membrane stiffness support the conclusion that α-tubulin detyrosination can influence the overall mechanical properties of the cytoskeleton. The effects of PTL were independent of significant alterations in MT network density (see Figs 1b,e and 4a,b), suggesting that detyrosination can, by itself, modulate cytoskeletal mechanics.

Figure 4: Detyrosinated tubulin affects cytoskeletal stiffness
figure4

(a) Differentiated C2C12 cells were treated with either DMSO (dimethylsulphoxide), taxol (10 μM), or PTL (10 μM) before extraction for western blotting. Following PTL or taxol treatment, no change in α-tubulin was noted. In contrast, PTL significantly decreased and taxol significantly increased detyrosinated tubulin (normalized to control, as in (b)). *P<0.05 (n=6 per treatment) Statistical significance was determined by using a one-way analysis of variance (ANOVA) with a Tukey’s post hoc test. (c) Differentiated C2C12 cells treated with DMSO, taxol, or PTL were fixed and immunostained for α-tubulin and detyrosinated tubulin. PTL decreased the amount of detyrosinated tubulin detected, while taxol increased detyrosinated MTs. Scale bar, 10 μm. (d) α-actinin staining confirming sarcomere development in differentiated C2C12 cells used for AFM measurements (Scale bar, 10 μm.) (e) Phase contrast showing AFM tip and C2C12 cells (Scale bar, 10 μm). (f) PTL treatment significantly decreased while taxol treatment significantly increased cytoskeletal stiffness. Unhashed bars indicate pre-treatment measures; hashed bars indicate measurements taken post-treatment. Statistical significance determined by repeated measures ANOVA with post hoc Tukey, *P<0.05 (n=12 cells per treatment) All values are mean±s.e.m.

Detyrosinated MTs regulate X-ROS mechanotransduction

In our initial characterization of X-ROS, a small, brief stretch (10% SL, 5–10 s.) resulted in a burst of ROS in cardiomyocytes6, though this same protocol failed to evoke an X-ROS signal reliably in WT skeletal muscle fibres4. With refined methods and new fiber attachment technology (Fig. 5a,b), we now show that a repetitive sinusoidal length change (5 μm, 0.5 Hz; Fig. 5c) elicits a significant ROS signal in WT FDB myofibres (Fig. 5d,e; grey). The Nox2 inhibitory peptide gp91ds-TAT abolishes the generation of ROS during stretch (Fig. 5d,e; black), confirming that the ROS signal is indeed X-ROS.

Figure 5: Mechano-activated ROS production and Ca2+ influx are regulated by detyrosinated tubulin.
figure5

(a) Apparatus for the mechanical manipulation of single skeletal muscle fibres mounted over an inverted fluorescence microscope (left panel). Middle panel is close up view of myofibre holders mounted to the piezo length controller (left) and force transducer (right). (b) Microfabricated glass myofibre holders (kind gift of Ionoptix, Inc.; see Methods) have an etched channel (30 μm diameter; side view on top, bottom view below), that when coated with Myotak adhesive, are designed to receive and hold a single muscle fibre. The far right panel shows an isolated myofibre (pseudocolored pink) held by the microfabricated glass holders. (c) Single FDB myofibres were challenged with a protocol of dynamic sinusoidal stretch. Representative experimental records of piezo actuator length output (top) and SL(bottom) recorded in single control fibres. (d) Averaged fluorescence records of single FDB fibres loaded with the ROS indicator DFF at resting length (SL 1.83), during 20 s of dynamic sinusoidal stretch (8 μM length excursion at 1 Hz; from resting SL to 1.95 μm SL, approximately 8% SL excursion) and post stretch. (e) Dynamic stretch of control fibres (grey) elicited a significant increase in DFF fluorescence that returned to pre-stretch rates with the cessation of stretch (hashed bars). The Nox2 inhibitor peptide gp91ds-TAT ameliorated the mechano-activated ROS signal indicating X-ROS as the mechanism of ROS generation. Pre-treatment with PTL or overexpression of TTL significantly reduced the mechano-activated ROS signal. *P<0.05 (n=12 cells per treatment). (f) FDB fibres loaded with near-membrane Indo-1 were assayed for sarcolemmal Ca2+ permeability during mechanical stretch. The rate of decline in Indo1 fluorescence due to internalization of extracellular Mn and subsequent quenching of Indo1 fluorescence was taken as sarcolemmal Ca2+ permeability. (g) Compared with pre-stretch rates, dynamic sinusoidal stretch elicited a significant increase in Mn quench that was abrogated by PTL. The stretch-dependent Mn influx was abrogated by the mechanosensitive channel inhibitor GsMTx4. *P<0.05 (n=10 cells per treatment). Statistical significance determined by one-way analysis of variance with Tukey’s post hoc test. All values are mean±s.e.m.

Having established X-ROS in both skeletal muscle fibres and cardiomyocytes, we sought to determine the role of detyrosination in its generation. We found that skeletal muscle fibres treated with PTL had a significantly reduced X-ROS (Fig. 5d,e; red). To confirm this pharmacological approach, we overexpressed the highly-specific TTL (ref. 43), the enzyme responsible for the reversal of detyrosination by ligation of the C-terminal tyrosine residue on α-tubulin, by electroporation into FDB muscle. In fibres overexpressing TTL, we found a significant reduction in MT detyrosination (Supplementary Fig. 3) and stretch-dependent ROS generation in vitro (Fig. 5d,e; blue), firmly establishing that detyrosinated MTs promote the mechanotransduction of X-ROS.

In adult dystrophin-deficient skeletal muscle, excess X-ROS hyper-sensitized sarcolemmal Ca2+ influx channels to drive Ca2+ influx during stretch4. To test if X-ROS modulates Ca2+ influx in WT muscle, we used a modified manganese (Mn2+) quench assay and a near-membrane ratiometric calcium indicator (NM-Indo1) to determine sarcolemmal Ca2+ permeability during sinusoidal stretch. Following the replacement of extracellular Ca2+ with Mn2+ (Fig. 5f,g; grey), the stretch-dependent influx of Mn2+ through sarcolemmal cation channels quenched the NM-Indo1 signal. The mechanosensitive channel blocker GsMTx4 inhibited this Mn2+ quench (Fig. 5f,g blue), confirming the mechanosensitivity of this pathway. PTL treatment also significantly reduced stretch-dependent Mn2+ quench (Fig. 5f,g; red), supporting detyrosinated MTs as critical for the mechano-activation of X-ROS and subsequent calcium influx.

In cardiomyocytes, X-ROS is generated with both stretch and electrically evoked contractions6,9. PTL treatment significantly reduced contraction-induced ROS production in cardiomyocytes, and eliminated ROS production with sinusoidal stretch (1 Hz, 10% of cell length) (Fig. 6a-c). X-ROS was previously demonstrated to increase the sensitivity of ryanodine receptor (RyR) calcium release channels, resulting in an increased frequency of Ca2+ ‘sparks’6, elementary calcium release events from the sarcoplasmic reticulum (Fig. 6d,e; control). Consistent with the inhibition of X-ROS, we found that PTL treatment dramatically inhibited Ca2+ sparks induced by sinusoidal stretch (Fig. 6e).

Figure 6: Mechanical stress-induced ROS production and Ca2+ sparks in heart.
figure6

(a) Average traces (baseline subtracted, see Methods) of contraction-induced ROS production in isolated cardiomyocytes. (b) Parthenolide (PTL) reduces contraction-induced ROS production, quantified as per cent change in ROS production from baseline during (solid bars) and after (striped bars) 2 Hz field stimulation of isolated cardiomyocytes. *P<0.05 (n=49 control cells, 48 PTL cells). (c) PTL blocks stretch-dependent ROS production, quantified as per cent change in ROS production from baseline during (solid bars) and after (striped bars) 2 Hz 10% cyclic stretch of isolated cardiomyocytes. *P<0.05 (n=11 control cells, 12 PTL cells). (d) Representative fluorescence surface plots of calcium sparks in cyclically stretched myocytes with or without PTL treatment. PTL abrogated the stretch-dependent increase in calcium sparks seen in control cells. (e) Quantification of Ca2+ spark rate on first 10 s of initial cyclic stretch, after 1 min of sustained cyclical stretch, immediately following the release of stretch, and 30 s after the release of stretch, normalized to the pre-stretch rate *P<0.05 (n=22 control, 16 PTL cells). Statistical significance determined by repeated measures analysis of variance with post hoc Tukey. All values are mean±s.e.m.

To control for non-specific effects of PTL treatment on the ROS-dependent regulation of calcium channels independent of MT and mechanotransduction, we applied H2O2 to cardiomyocytes and FDB fibres. H2O2 elicited a rapid increase in Ca2+ spark activity and Ca2+influx (Supplementary Fig. 4), independent of PTL. These data indicate that PTL directly interferes with the mechanotransduction through X-ROS, rather than altering Nox2 or ROS-dependent regulation of downstream targets.

Reducing MT detyrosination protects DMD skeletal muscle

Growing evidence implicates a pathological role for altered MTs in cardiomyopathy44,45,46,47,48,49 and the muscular dystrophies4,6,10. Our group and others have demonstrated that acute pharmacologic disruption of the MT network with colchicine is effective at reducing disease-related signalling and functional pathologies in pre-clinical models4,6,40,50. However, the toxicity profile of this pharmacologic class23,51,52, underpinned by the interruption of essential MT network functions other than mechanotransduction (e.g., transport, signalling, metabolism), may limit the therapeutic potential of MT disruption.

Our previous work established that X-ROS drives detrimental Ca2+ signalling and pathology in heart and skeletal muscle from a mouse model of DMD (mdx mice), and that there is an increase in detyrosinated tubulin in both skeletal4 and cardiac6 mdx muscle. We confirm here, using our new sinusoidal stretch protocol, that mdx muscle fibres generate more X-ROS during stretch than WT controls (Supplementary Fig. 5). As in WT cells, acute treatment with PTL specifically reduces MT detyrosination and alleviates this excess X-ROS in mdx fibres (Supplementary Fig. 5). Based on this finding and the above in vitro evidence that reducing detyrosination inhibits X-ROS and its downstream consequences (Figs 5 and 6), we tested whether reducing detyrosination in vivo provided a significant benefit to the mdx mouse. We treated adult mdx mice (n=6) with LC-1 (or dimethylamino-parthenolide)53, a prodrug of PTL with enhanced solubility, or vehicle, once a day for three days. On the last day of treatment, mice from each group were evaluated for their resistance to either skeletal or cardiac stress-induced injury.

Resistance to contraction-induced skeletal muscle injury was assessed in vivo by challenging anesthetized mice with 20 eccentric contractions of the hind limb plantar flexors4. The isometric torque of the initial contraction revealed no difference between the vehicle and LC-1 treated group, but subsequent contractions revealed a significant protection against contraction-induced force loss in animals treated with LC-1 (Fig. 7a,b). This level of protection was similar to our previous findings with pharmacologic ablation of the MT network or inhibition of Nox activity4. Western blot analysis of the contralateral uninjured muscle revealed that LC-1 treatment had no significant effect on Nox2 (gp91phox), α-tubulin content or levels of acetylated, polyglutamylated, or Δ2-tubulin, while decreasing the amount of detyrosinated tubulin to approximately WT levels (Fig. 7c, Supplementary Fig. 1)4.

Figure 7: Targeting detyrosinated tubulin in vivo reduces contraction-induced force decline.
figure7

Mdx mice (12 +/− 3.2 months) were treated with the PTL prodrug, LC-1 (100 mg kg−1) or vehicle control by oral gavage for 3 days. (a) On the fourth day, a protocol of 20 eccentric contractions revealed a decline in isometric force before stretch in both vehicle and LC-1 treated mdx mice. (b) Repetitive eccentric contractions of the gastrocnemius muscle were elicited in vivo. The characteristic decline in isometric force was significantly inhibited with LC-1 treatment. *P<0.05 (n=5). (c) Western blot of gastrocnemius muscles following LC-1 treatment show significant decrease in detyrosinated α-tubulin, but no change in Nox2 (gp91phox) or total α-tubulin content. Blots quantitated as per cent change from vehicle treatment. *P<0.05 (n=5) (df) FDB fibres from vehicle control or LC-1 treated mice were enzymatically isolated, loaded with calcium indicator dye (Fluo-4-AM). (d) Aggregate normalized Fluo4 traces during a sustained train of action potential pulses (20 Hz for 3 min). Ten sec before the cessation of stimulation, FDB myofibres from LC-1 treated mdx mice exhibited a significantly reduced peak cytosolic Ca2+. *P<0.05 (n=15 cells per treatment). (e) Quantitation of D. (f) Fibres from LC-1 treated mdx mice exhibited a significant increase in the decay of cytosolic Ca2+ following stimulation, as the post-stimulation fluorescence (Fpost) declined lower than pre-stimulation levels (Fo) which yielded a negative Fpost-F0. In contrast, the Fpost in vehicle-treated FDBs were significantly greater than pre-stimulation values 60 s after the end of the stimulation. *P<0.05 (n=15 cells per treatment). Statistical significance determined by Student’s t-test. All values are mean±s.e.m.

We previously showed that ablation of the MT network can reduce detrimental mechano-activated [Ca2+]i signalling in mdx fibres4, a hallmark of DMD. As a direct test of the effect of LC-1 treatment on contraction-induced Ca2+ signalling dysfunction, single mdx FDB myofibres were isolated from the foot of the uninjured leg of each treatment group, loaded with the fluorescent Ca2+ dye Fluo-4-acetoxymethyl (AM) and challenged with a 2 min, 5 Hz train of twitch contractions. Before stimulation, both groups exhibited a similar level of cytosolic fluorescence, however, during the 2 min test period, FDBs from LC-1 treated mdx animals maintained significantly lower cytosolic [Ca2+] compared with vehicle-treated mdx fibres (Fig. 7f,g). At the end of the pulse train, the return to resting [Ca2+] was significantly faster in LC-1 treated fibres, as [Ca2+]i remained elevated above pre-stimulation values for upwards of 100 s following stimulation in untreated mdx fibres (Fig. 7h). Taken together, these results suggest that reducing detyrosination protects against contraction-induced Ca2+ signalling dysfunction and injury in mdx muscle.

Reducing MT detyrosination protects DMD heart

We and others have previously shown that mechanical stress drives detrimental ROS and Ca2+ signalling in the mdx heart6,54 that can trigger arrhythmias55,56. We next assayed the contribution of the detyrosinated MT network to this pathology in the dystrophic heart. Mdx mice treated with either LC-1 or vehicle for three days were anesthetized, instrumented to record electrocardiogram (ECG) and challenged with the β-adrenergic agonist isoproterenol (0.5 mg kg−1) to increase cardiac work. Immediately following isoproterenol challenge, the resting heart rate increased two- to three-fold in both treatment groups. In vehicle-treated mdx mice (Fig. 8a), arrhythmias became apparent within 10 min and progressed in frequency until the mice succumbed to arrhythmia-related cardiac failure and death (mortality in 12 of 13 animals tested). In contrast, no mice in the LC-1 treatment group (n=6) succumbed to arrhythmia-related death (Fig. 8a). Western blot analysis of the vehicle versus LC-1 treated hearts revealed that LC-1 treatment had no significant effect on α-tubulin, acetylated tubulin, or Nox2 (gp91phox) content, but significantly decreased the amount of detyrosinated α-tubulin (Fig. 8c, Supplementary Fig. 1e). Consistent with LC-1 acting through X-ROS to exert its effect, protection from arrhythmia-associated death was also conferred by the in vivo inhibition of Nox2 with gp91ds-TAT (n=3/3 survival) or by pharmacologic disruption of the MT network with colchicine (n=5/5 survival).

Figure 8: In vivo targeting of detyrosination in mdx heart rescues isoproterenol induced arrhythmia and lethality.
figure8

(a) Representative ECG recordings from mdx mice given vehicle treatment or the parthenolide prodrug, LC-1, and exposed to an isoproterenol challenge to increase cardiac work. Off-scale readings in mdx trace represent agonal breathing shortly preceding death. (LC-1: n=6; vehicle: n=13) Scale bar, 1 s. (b) Per cent survival of mice following isoproterenol challenge. All 6 LC-1 treated mice survived (100% survival rate); 12 of 13 vehicle-treated mice were anesthetized due to fatal arrhythmia (7.6% survival rate). (c) Western blotting of isolated hearts following LC-1 or vehicle treatment shows a significant decrease in detyrosinated tubulin without significantly altering α-tubulin or Nox2 content. Blots quantitated as per cent change from vehicle-treated. Statistical significance determined by Student’s t-test. *P<0.05 (n=3 per group). All values are mean±s.e.m.

To directly test the role of MT detyrosination in generating arrhythmogenic calcium signals, we used an established high-frequency stimulation protocol to evaluate Ca2+-dependent arrhythmogenic potential in single mdx cardiomyocytes loaded with the Ca2+ indicator Fluo-3 (ref. 56). Following a 30 s burst of 3Hz electrical pacing, untreated mdx cardiomyocytes exhibited spontaneous, propagating Ca2+ waves linked to arrhythmia formation, as well as a significant increase in resting Fluo-3 fluorescence (Fig. 9a–c). These findings are consistent with RyR-mediated calcium ‘leak’ that underscores arrhythmogenesis in mdx myocytes6,55,56. PTL treatment significantly reduced the frequency of Ca2+ waves and prevented the elevation in resting calcium concentration following high-frequency stimulation (Fig. 9a–c), suggesting that targeting MT detyrosination reduces detrimental calcium signalling and stress-dependent arrhythmogenesis in the DMD heart.

Figure 9: Parthenolide treatment ablates stress-induced Ca2+ waves and leak in mdx cardiomyocytes.
figure9

(a) Representative traces of calcium transients and calcium waves during and after high-frequency (3 Hz) stimulation of mdx myocytes. (b) Parthenolide treatment significantly reduced the frequency of Ca2+ waves following high-frequency stimulation of mdx myocytes. (c) Parthenolide treatment prevented an elevation in resting calcium concentration following high-frequency stimulation of mdx myocytes *P< 0.01 versus mdx untreated, #P<0.05 versus mdx immediately post stim. (n=36 control cells, 36 PTL cells). Statistical significance determined by Student’s t-test. All values are mean±s.e.m.

Discussion

We report the novel discovery that detyrosination is a potent regulator of MT-dependent mechanotransduction in striated muscle. Using both pharmacologic and genetic strategies, we show that MT detyrosination correlates with the magnitude of X-ROS signalling in heart and skeletal muscle. As the level of detyrosination affects both the velocity of contraction and cytoskeletal stiffness, we conclude that detyrosination regulates mechanotransduction by altering the mechanical properties of the cytoskeleton.

Iribe et al57 first identified the role of MTs in the mechanotransduction of cardiac Ca2+ signalling. Subsequent studies identified X-ROS as the proximate activator of these Ca2+ signals in both cardiac6 and skeletal muscle4 by showing the blockage of X-ROS with pharmacologic ablation of the MT network. Here we report that a selective reduction in detyrosinated MTs yields a similar effect on mechanotransduction, while preserving the overall MT network. This was at first a surprising finding, particularly as there is no evidence that detyrosination contributes to the intrinsic stiffness of the individual MT filament per se, and historically, some have viewed this PTM (along with acetylation, polyglutamylation and Δ2-tubulin) as simply a marker of stable microtubules.

However, this view is being challenged, and our study adds to an emerging body of work demonstrating specific roles for detyrosination in the regulation of MT function14, including the organization of chromosomes in mitosis31 and the regulation of Na+/K+ ATPase activity58. We found detyrosination uniquely abundant in healthy muscle compared with other common PTMs (Fig. 1, Supplementary Figs 1,2)28, and PTL and LC-1 selectively reduced this PTM without detectably altering acetylation, polyglutamylation or Δ2-tubulin, allowing the attribution of functional consequences to detyrosination. On the other hand, taxol has complex effects35 as it increases the amount of polymerized MTs, and several tubulin modifying enzymes favour the polymerized form14. Here taxol increased MT network density, dramatically increased detyrosination, and modestly affected other PTMs, and thus results should be interpreted with this in mind. Separating the contribution of these individual PTMs to specific biological functions will be enabled by the continued identification of tubulin modifying enzymes and molecular tools to manipulate them, including the elusive TCP14.

How can reducing the fraction of detyrosinated MTs have such significant effects on mechanotransduction? While detyrosination may not alter intrinsic MT stiffness in vitro, it does increase MT stability by inhibiting disassembly via motor-proteins18. In addition, while MTs can resist compressive loads, the mechanical properties of the MTs alone are insufficient59, and likely require cross-linking interactions with other proteins and cytoskeletal elements to confer the requisite mechanical stiffness needed to resist cytoskeletal compression60. Recent studies linking functional deficits in heart failure and muscular dystrophy to the overexpression of Microtubule-Actin Cross-Linking Factor 1 (ref. 61) or the cytoskeletal cross-linker plectin62 offer support for this idea in striated muscle. Further evidence comes from non-muscle cells, where increased cytoskeletal stiffness is driven by interactions between detyrosinated MTs and intermediate filaments (IFs)21,22,29. In this context, it is notable that the expression of desmin, the major IF in striated muscle, is elevated in the mdx mouse63, and its deletion provides significant functional benefit63. While the authors did not attribute a rescue of dysregulated mechanotransduction to their results, our current data support this possibility, as increased MT:IF cross-linking offers a parsimonious explanation for the detyrosination-dependent elevations in X-ROS signalling and cytoskeletal stiffness. Finally, detyrosination also regulates the interaction of MTs with CAP-Gly proteins, which may have important consequences for both cytoskeletal organization and cell signalling64.

DMD provides an ideal model to test the contribution of detyrosinated MTs in mechanotransduction-dependent dysfunction, as an increase in MT network density and detyrosination correlates with pathology in DMD4,5,6,10. While previous studies have demonstrated that acute pharmacologic disruption of MT structure can be beneficial in rodent models of DMD and pathologic cardiac hypertrophy4,32, this strategy is limited by the potential for depressing other central MT functions23,51,52. Targeting detyrosination is an attractive strategy as it exerts significant effects while sparing the MT network. Here we report that acute in vivo pharmacologic reduction of detyrosination protects the DMD heart from workload-induced cardiac death and skeletal muscle from contraction-induced injury. Our in vitro studies confirm that reducing detyrosination inhibited the detrimental Ca2+ signals that underlie these DMD pathologies.

Although we limited our pre-clinical effort in this study to DMD, elevated MT density and detyrosination have been identified in multiple models of muscular dystrophy4,5,7,8 and pathologic cardiac hypertrophy6,36, suggesting that these MT alterations are conserved maladaptive responses in disease. Despite the growing evidence for a MT-targeted strategy to slow or halt disease progression in DMD and the failing heart, the move to clinical translation has been slowed by the unavailability of promising therapeutic targets and effective pharmacologic agents. PTL is a sesquiterpene lactone that has demonstrated promising utility as an anti-proliferative agent for treating human cancers24. Given that PTL (as feverfew extract) and LC-1 have advanced to Phase I human clinical trials65, compounds such as PTL represent novel lead candidates for MT-centric therapy in striated muscle disease. Beneficially, significant clinical data generated by cancer therapeutics65 will be available to guide the translation of SL therapy to striated muscle diseases.

In summary, we present evidence that MT detyrosination regulates mechanotransduction through ROS and Ca2+ signals. In diseased muscle sensitive to mechanical stress, inhibiting detyrosination blunts the stress-induced ROS and Ca2+ signalling dysfunction that drives pathogenic progression. We conclude that MT detyrosination regulates mechanotransduction in striated muscle and that targeting this PTM may have broad therapeutic potential in the muscular dystrophies.

Methods

All reagents and drugs were purchased from Sigma-Aldrich unless otherwise noted. Gp91ds-TAT peptide was purchased from Anaspec. GsMTx4 was kindly provided by F. Sachs.

Murine models

All animal care and experimental procedures were approved and performed in accordance with the ethical standards set by the Institutional Animal Care and Use Committees of both the University of Maryland and the University of Pennsylvania and by the National Institutes of Health (NIH). Mice (stock numbers 001801 and 000664) were acquired from Jackson Laboratories. Rats (Hsd:Sprague Dawley SD) were purchased from Harlan Labs. Animal cohorts within each genotype were randomly assigned to experimental groups.

Preparation of flexor digitorum brevis (FDB) muscle fibres

FDB muscles from male control rat (WT), as well as male control mice (C57/Bl6; WT) and dystrophin-deficient (C57/10ScSn-Dmdmdx/J; mdx) mice at 9–11 months of age were isolated as follows: after euthanasia, FDB muscles were harvested bilaterally and placed in 4 mg ml−1 collagenase A. Following incubation at 37 °C for 2 h, the muscles were gently triturated to release individual fibres. Fibres were imaged or fixed within 24 h of isolation.

Cardiomyocyte isolation

Adult C57/Bl6 or mdx mice and Sprague–Dawley rats were terminally anesthetized by injection of pentobarbital, followed by excision of the heart and enzymatic isolation of ventricular myocytes66. Hearts were suspended from a Langendorff perfusion column via cannulation of the aorta and perfused with cell isolation buffer containing in mM: NaCl 130, HEPES 25, Glucose 22, KCl 5.4, lactic acid 1, MgCl2 0.5, NaH2PO4 0.33, CaCl2 0.3, to which insulin (100 micro-units per millimetre), collagenase II (0.66 mg ml−1 Worthington LS004176), trypsin (0.033 mg ml−1 Sigma T0303) and proteinase (0.033 mg ml−1 Sigma P8038) were added. Tissue was then minced and subjected to additional digestion with 0.7 mM CaCl2 and bovine serum albumin (2 mg ml−1) at 37 °C until myocytes were easily liberated by trituration with a Pasteur pipette. Myocytes were then removed from enzyme and then slowly reintroduced to physiological CaCl2 in cell isolation buffer without enzymes. Cardiomyocytes were stored in a normal Tyrode solution containing (in mM): NaCl 140, KCl 5, CaCl2 1.8, MgCl2 0.5, HEPES 5, Glucose 5 and NaH2PO4 0.33. Experiments were performed at room temperature, 22 °C. Due to the relatively greater ease and reproducibility of highest quality cell preparations, rat myocytes were used for all experiments except when mdx myocytes and appropriate murine controls were needed.

Immunofluorescence

Isolated FDB fibres were adhered to glass coverslips coated with Geltrex ECM (Life Technologies). Fibres were fixed at room temperature for 20 min using 4% paraformaldehyde in PBS, followed by permeabilization at room temperature for 15 min with 0.1% Triton X-100 in PBS (ref. 67). Non-specific staining was limited by adding SuperBlock PBS (Life Technologies) for 1 h before the addition of primary antibodies. Primary antibodies (α-tubulin clone DM1A, Sigma; rabbit detyrosinated tubulin, AbCam #48389; acetylated tubulin clone 6-11-B, Sigma; rabbit Δ2tubulin, Novus #NB100-57397; mouse polyglutamylation clone GT335, Adipogen) were diluted in SuperBlock PBS and added to the coverslips for an overnight incubation at 4 °C. Secondary antibodies (Goat anti-mouse Alexa 488, 647 and goat anti-rabbit Alexa 488, 568; Life Technologies; 1:100 dilution) were diluted in SuperBlock PBS and incubated at room temperature for 6 h. Coverslips were mounted on glass slides using ProLong Diamond (Life Technologies) and imaged using a Zeiss 510Live laser confocal microscope (× 63 1.4 numerical aperature plan-apo oil objective).

Isolated cardiac myocytes were fixed and permeablized in solution using 4% paraformaldehyde and 0.1% Triton X-100, both in PBS. Myocytes were then transferred to SuperBlock PBS and allowed to settle on coated coverslips for 1 h. The coverslips were then treated with antibodies, mounted and imaged as above.

Quantitation of α- and detyrosinated tubulin was performed with Volocity (Perkin Elmer). Total pixel density above threshold within a region of interest was calculated for each channel. The average density for each treatment condition and channel were calculated and normalized to the control untreated condition.

Skeletal fibre mechanics and [Ca2+]i transients

Isolated FDB fibres treated (2 h) with PTL (10 μM), taxol (10 μM), at room temperature, suspended in Hepes-buffered Ringer solution containing (in mM): 140 NaCl, 4 KCl, 1 MgSO4, 5 NaHCO3, 10 glucose and 10 HEPES (pH 7.3). Following treatment, fibres were placed in a custom rotating glass-bottomed perfusion chamber (Four-Hour Day Foundation) mounted over an inverted microscope (Olympus IX-70, × 40 objective). Contractions were elicited under no-load conditions with field stimulation (0.2-ms square pulse, 1 Hz elicited), while SL was monitored and recorded using a high-speed video SL system (HSVL 901B, Aurora Scientific). Across five successive contractions, velocity of shortening and relaxation as well as the time to peak shortening were calculated. Similarly, PTL or untreated FDBs equilibrated with the Ca2+ indicator Indo-1PE (5 μM, 0.1% Pluronic F127, 30 min) were electrically stimulated (1 Hz) and the peak, Vmin and Vmax of the Indo-1 ratio (405 nm/475 nm) were analysed (950 A, Aurora Scientific).

Myocyte contraction and [Ca2+]i transients

Experiments were performed in custom fabricated cell chambers (Four-hour Day Foundation, Towson, MD) mounted on a ZeissLSM 780 inverted confocal microscope with a × 40 Oil 1.4 NA objective. For measurements of contractility, SL was monitored with a high-speed video camera and Fourier transform analysis. The [Ca2+]i transient was evaluated with Fluo-3 by 15 min incubation with 2 μM Fluo-3-AM ester and 0.01% Pluronic F127. Cells were allowed an additional 10 min for de-esterification and then scanned using a 488 nm argon ion laser in confocal line-scan mode at 1.92 ms per line. Cells were paced to steady state using 1 Hz field stimulation through platinum electrodes. After 20 s of 1 Hz stimulation, all cells showed steady state transients and contractions. At this point, five Ca2+ transients and five contractions were assayed. This protocol was repeated 2 × in each myocyte, and data was pooled for analysis.

Myofibre attachment and stretch

All experiments were performed in a custom rotating glass-bottomed chamber (Four-Hour Day Foundation) equipped with bath perfusion and mounted on an inverted microscope (Olympus IX-70, 40 objective) fitted with a DG4 (Sutter Instruments), an IonOptix Myostretcher (IonOptix) system with a 403 A-CF force transducer (Aurora Scientific), and PMT Sub-System (IonOptix). Isolated FDB fibres were suspended in Hepes-buffered ringer as above, supplemented with PTL (10 μM) or left untreated (2 h). Individual fibres were attached between microfabricated glass fibre holders (IonOptix), coated with MyoTak biological adhesive (IonOptix), that were mounted to a force transducer (403A-CF, Aurora Scientific) and a Piezo actuator (PI). The force and positional outputs were recorded at 1kHz on a 600A data acquisition system from Aurora Scientific.

Stretch-induced ROS production was assayed in H2DFFDA (5-(and-6)-carboxy-2′,7′-dihydrofluorescein diacetate; 5 μM, 30 min; Life Technologies) loaded FDB fibres with a sinusoidal waveform (8% of resting SL) imposed for 20 s. Data were collected offline (500 Hz; 950A, Aurora Scientific) and the rate of ROS generation analysed using the Origin Pro 9.1 (Origin Lab). The Nox2-specific inhibitor gp91ds-TAT (2 μM) was used to verify the source of ROS.

Stretch-induced Ca2+ influx was assayed with a modified manganese quench (MnQ) assay. Briefly, cells were pre-loaded with Indo-1 Near-Membrane (NearMem, formerly FIP-18, 5 μM, Teflabs) and transferred and attached to the stretch apparatus described above. Perfusion with normal Ringer was switched to a Ringer where Ca2+ was substituted with manganese. Fibres were excited (340 nm), stretched with a sinusoidal waveform as above, and sarcolemmal permeability to Ca2+ estimated by the manganese influx and quench of the total emission fluorescence (FTOT=405+475 nm). The rate of fluorescence decrease was calculated and analysed as above. GsMTx4, a specific inhibitor of stretch-activated channels, was used to confirm the source of influx. For certain experiments, fibres were perfused with 200 μm H2O2 to elicit sarcolemmal calcium influx, independent of sinusoidal stretch (Supplementary Fig. 4).

Cardiomyocyte attachment and stretch

Glass microrods were coated with a biological adhesive, MyoTak (IonOptix, Milton, MA). One glass microrod was connected to a force transducer (403A-CF, Aurora Scientific), and the other to a length controller (World Precision Instruments). Myocytes were attached at both ends by gently pressing down with the MyoTak-coated microrod and then lifting the cell from the chamber bottom. Axial stretch was applied by movement of the length controller in response to variable voltage output. SL was monitored with a high-speed video camera (Aurora Scientific). Average SL before stretch was 1.8 μm. Myocytes were subjected to three stretch-release trials, with 30–60 s rest allowed between trials. As each stretch produced a similar change in SL, DFF fluorescence, and Ca2+ spark rate, trials were pooled for analysis. WT cells did not commonly display Ca2+ waves, and cells with Ca2+ waves were discarded from analysis.

Ca2+ spark measurements

Cells were loaded with Fluo-4 by 10 min incubation with 3 μM Fluo-4-AM ester and 0.01% Pluronic F127 and allowed an additional 10 min for de-esterification. Cells were scanned using a 488 nm argon ion laser in confocal line-scan mode at 1.92 ms per line. Automated analysis of line-scan images for Ca2+ spark location and properties was performed using SparkMaster (ImageJ68). For H2O2 studies, spark measurements were taken in the same cell before and after acute (<1 min) H2O2 (200 μM) exposure. Some cells were pretreated with PTL (2 h, 10 μM).

DFF measurements in cardiomyocytes

Cells were loaded with the cell permeant ROS indicator DFF (a more photostable analogue to the commonly used DCF) by a 20 min exposure to a Tyrode’s solution containing 2.5 μM H2DFFDA. The cells were imaged using confocal microscopy at 1 frame per second. Cells were imaged at very low laser intensity and with a sampling rest interval of 2 s to avoid artifactual signal amplification on continuous light exposure. The slope of the DFF signal is indicative of ROS production; therefore the slope before stretch has been subtracted from the entire trace so that the initial slope is zero and changes in slope can be easily visualized4,6. The slope is expressed normalized to the slope before stretch in that same cell as this comparison minimizes differences in DFF loading and resting signal between cells. As the slope was linear for the duration of cyclic stretch, the slope was calculated from the entire stretch duration for all cyclic stretch experiments.

Atomic force microscopy

The transversal stiffness of differentiated C2C12 myotubes was quantitatively measured by means of nanoindentation measurements based on atomic force microscopy (AFM). Murine C2C12 myoblast cell line provides a well-established and reproducible model system for skeletal myogenesis69. When 80% cell confluence was reached, the medium containing 10% foetal calf serum was substituted with 2% foetal calf serum to induce myogenic differentiation, a process lasting a few days which can be recognized by a progressive elongation in cell shape and an assembling of structures involved in contraction. Differentiated C2C12 cells (American Type Culture Collection, ATCC) cultured on coverglass were treated for 2 h with PTL, Taxol or dimethylsulphoxide and then fixed and immunostained for MT as described above.

A commercial AFM (model 5500, Keysight Technologies) coupled with an inverted optical microscope was employed to record force-versus-distance curves, taken at a constant speed of 3 μm s−1 on differentiated C2C12 myotubes. At least three 500 × 500 nm2 areas, selected over the cell body to avoid nuclei regions, were sampled with a regular grid of 16 × 16 force curves each, for a total number of force curves ≥768 for each considered cell. The same commercial silicon microcantilever with a conical tip (model CSG11, NT-MDT, nominal spring constant k=0.03 N m−1) was used for all measurements. The region after contact of each force curve was transformed into a load-versus-indentation curve and the corresponding contact stiffness could be calculated at a certain maximum indentation. To maximize the sensitivity of the measurement towards the microtubule network structure and organization the reported values were calculated at a maximum indentation of 100 nm. Acute treatments were performed by incubating the already differentiated C2C12 culture for 2 in the medium containing either parthenolide or taxol at a final concentration of 10 μM. Before starting the AFM measurements the solution was replaced with fresh medium. Following AFM indentation measurement, cultures were labelled by using a primary antibody against α-actinin (clone EA-53, Sigma; goat anti-mouse Alexa 549 secondary) and fluorescent phalloidin (Alexa 488) to confirm the periodic sarcomeric organization of striated muscle fibres.

Formulation of LC-1

The parthenolide prodrug LC-1 (ref. 53) was formulated for oral dosing in 2-hydroxypropyl-β-cyclodextrin (HP-β-CD)70. In brief, 160 mg of LC-1 were dissolved in 3.75 ml of an aqueous 0.3 M HP-β-CD solution and mixed well. The LC-1 and vehicle only solutions were stored at 4 °C before administration.

In vivo injury

Mdx mice (9–11 months) were treated daily via gavage with a bioavailable parthenolide analogue (LC-1) or vehicle (HP-β-CD) for 3 days (100 mg kg−1). On the following day, mice were anesthetized and eccentric injury of the gastrocnemius was performed using a 305B muscle lever system (Aurora Scientific Inc.)4. The knee was stabilized and the foot firmly affixed to a footplate on the torque sensor. Contraction was elicited by percutaneous electrical stimulation of the sciatic nerve. Optimal isometric tetanic torque was determined with increasing the current, with a minimum of 30 s between each stimulation to prevent fatigue. Eccentric contractions were created by translating the footplate 30° backward at a velocity of 40 mm s−1 after the first 200 ms of the isometric contraction. Eccentric contractions at maximal isometric torque (0.2-ms pulse train at 100 Hz) were assayed 20 times with 1 min pauses between each, for the resistance to muscle damage. The decrease in the peak isometric force before the eccentric phase was used as a measure of damage. Following injury, the gastrocnemius was excised and flash frozen for western blotting.

In vivo manipulation of cardiac workload

Mice (9–11 months) were treated with either LC-1 or vehicle (HP-β-CD) for 3 days (100 mg kg−1). On the subsequent day, mice were lightly anesthetized and instrumented for non-invasive ECG (Biopaq; three lead). ECG was monitored and digitized (1KHz) for 5 min before a challenge with isoproterenol (ISO; 0.5 mg kg−1) and then for 45 min or until the heart rate was significantly depressed, at which time the mouse was euthanized under anaesthesia. Hearts were excised and flash frozen for western blotting.

Calcium stress assay

FDB muscle fibres from LC-1 and vehicle-treated mice were isolated as described above, placed into Hepes-buffered Ringer’s solution and plated into ECM-coated wells of a clear-bottom 96-well plate. Cells were loaded with Fluo-4 (5 μM) for 30 min and rinsed with Ringer’s solution. The cells of individual wells were electrically stimulated using a custom perfusion and stimulation apparatus (Four-Hour Day Foundation) for 5 min at 20 Hz, during which time Ca2+ was recorded.

Western blotting

For western blotting, 20 mg of clarified muscle or C2C12 extract was subjected to SDS-polyacrylamide gel electrophoresis, transferred on nitrocellulose membranes, and washed/blocked in 5% milk solution in PBS for 1 h. The membrane was probed overnight with primary antibody at room temperature. The primary antibodies were: anti-α-tubulin (Sigma DM1A, 1:1,000), anti-detyrosinated tubulin (AbCam #ab48389, 1:1,000), anti-acetylated tubulin (Sigma 6-11-B, 1:200), anti-Δ2-tubulin (Novus ##NB100-57397, 1:1,000), anti-polyglutamylation (Adipogene GT335, 1:1,000) and anti-gp91phox (BD Transduction Labs #611415, 1:1,000). Membranes were washed two times for 10 min in 5% milk solution at room temperature and incubated with appropriate secondary antibody (1:10,000) for 1 h at room temperature, and the membranes were washed in 0.5% Tween solution in PBS two times for 10 min. Membranes were then exposed to enhanced chemiluminescence with SuperSignalWest Pico Chemiluminescent Substrate to develop immunoblots. Blots were imaged and quantified with an imaging system (Syngene, G:Box; GeneTools software), normalized to total protein (via PonceauS) or GAPDH (Fig. 1, cardiomyocytes) and analysed using ImageJ. Images were cropped for presentation. Uncropped scans of western blots are shown in Supplementary Fig. 6.

Statistical analysis

The sample number for each experiment was determined by a power analysis (http://www.power-analysis.com/) assuming a Cohen’s maximum variability of the population mean (determined by variance in published experiments) and a type 1 error of 0.5%. For in vivo studies, animals were randomly assigned to treatment groups and blinded for experimentation and analysis. All other experiments were performed and analysed unblinded. Normality of data sets was ensured for each statistical calculation. Significance was set a priori at P<0.05.

Additional information

How to cite this article: Kerr, J. P. et al. Detyrosinated microtubules modulate mechanotransduction in heart and skeletal muscle. Nat. Commun. 6:8526 doi: 10.1038/ncomms9526 (2015).

References

  1. 1

    Nat. Rev. Mol. Cell Biol. 10, 1–1 (2009).

  2. 2

    Jaalouk, D. E. & Lammerding, J. Mechanotransduction gone awry. Nat. Rev. Mol. Cell Biol. 10, 63–73 (2009).

    CAS  Article  Google Scholar 

  3. 3

    Kombairaju, P. et al. Genetic silencing of Nrf2 enhances X-ROS in dysferlin-deficient muscle. Front. Physiol. 5, 1–7 (2014).

    Article  Google Scholar 

  4. 4

    Khairallah, R. J. et al. Microtubules underlie dysfunction in Duchenne muscular dystrophy. Signal 5, ra56 (2012).

    Google Scholar 

  5. 5

    Belanto, J. J. et al. Microtubule binding distinguishes dystrophin from utrophin. Proc. Natl Acad. Sci. USA 111, 5723–5728 (2014).

    CAS  ADS  Article  Google Scholar 

  6. 6

    Prosser, B. L., Ward, C. W. & Lederer, W. J. X-ROS signaling: rapid mechano-chemo transduction in heart. Science 333, 1440–1445 (2011).

    CAS  ADS  Article  Google Scholar 

  7. 7

    Prosser, B. L., Khairallah, R. J., Ziman, A. P., Ward, C. W. & Lederer, W. J. X-ROS signaling in the heart and skeletal muscle: stretch-dependent local ROS regulates [Ca2+]i. J. Mol. Cell Cardiol. 58, 172–181 (2013).

    CAS  Article  Google Scholar 

  8. 8

    Ward, C. W., Prosser, B. L. & Lederer, W. J. Mechanical stretch-induced activation of ROS/RNS signaling in striated muscle. Antioxid. Redox Signal. 20, 929–936 (2014).

    CAS  Article  Google Scholar 

  9. 9

    Prosser, B. L., Ward, C. W. & Lederer, W. J. X-ROS signaling is enhanced and graded by cyclic cardiomyocyte stretch. Cardiovasc. Res. 98, 307–314 (2013).

    CAS  Article  Google Scholar 

  10. 10

    Prins, K. W. et al. Dystrophin is a microtubule-associated protein. J. Cell Biol. 186, 363–369 (2009).

    CAS  Article  Google Scholar 

  11. 11

    Akhmanova, A. & Steinmetz, M. O. Tracking the ends: a dynamic protein network controls the fate of microtubule tips. Nat. Rev. Mol. Cell Biol. 9, 309–322 (2008).

    CAS  Article  Google Scholar 

  12. 12

    Webster, D. R., Modesti, N. M. & Bulinski, J. C. Regulation of cytoplasmic tubulin carboxypeptidase activity during neural and muscle differentiation: characterization using a microtubule-based assay. Biochemistry 31, 5849–5856 (1992).

    CAS  Article  Google Scholar 

  13. 13

    Kreis, T. E. Microtubules containing detyrosinated tubulin are less dynamic. EMBO J. 6, 2597–2606 (1987).

    CAS  Article  Google Scholar 

  14. 14

    Yu, I., Garnham, C. P. & Roll-Mecak, A. Writing and reading the tubulin code. J. Biol. Chem. R115, 637447 (2015).

    Google Scholar 

  15. 15

    Argarana, C., Barra, H. & Caputto, R. Release of [14C]tyrosine from tubulinyl-[14C]tyrosine by brain extract. Separation of a carboxypeptidase from tubulin-tyrosine ligase. Mol. Cell Biochem. 19, 17–21 (1978).

    CAS  Article  Google Scholar 

  16. 16

    Raybin, D. & Flavin, M. Enzyme which specifically adds tyrosine to the alpha chain of tubulin. Biochemistry 16, 2189–2194 (1977).

    CAS  Article  Google Scholar 

  17. 17

    Ersfeld, K. et al. Characterization of the tubulin-tyrosine ligase. J. Cell Biol. 120, 725–732 (1993).

    CAS  Article  Google Scholar 

  18. 18

    Peris, L. et al. Motor-dependent microtubule disassembly driven by tubulin tyrosination. J. Cell Biol. 185, 1159–1166 (2009).

    CAS  Article  Google Scholar 

  19. 19

    Gundersen, G. G., Kalnoski, M. H. & Bulinski, J. C. Distinct populations of microtubules: tyrosinated and nontyrosinated alpha tubulin are distributed differently in vivo. Cell 38, 779–789 (1984).

    CAS  Article  Google Scholar 

  20. 20

    Whipple, R. A. et al. Vimentin filaments support extension of tubulin-based microtentacles in detached breast tumor cells microtentacles in detached breast tumor cells. Cancer Res. 5678–5688 (2008).

  21. 21

    Kreitzer, G., Liao, G. & Gundersen, G. G. Detyrosination of tubulin regulates the interaction of intermediate filaments with microtubules in vivo via a kinesin-dependent mechanism. Mol. Biol. Cell 10, 1105–1118 (1999).

    CAS  Article  Google Scholar 

  22. 22

    Gurland, G. & Gundersen, G. G. Stable, detyrosinated microtubules function to localize vimentin intermediate filaments in fibroblasts. J. Cell Biol. 131, 1275–1290 (1995).

    CAS  Article  Google Scholar 

  23. 23

    Stamp, L. K. Safety profile of anti-gout agents: an update. Curr. Opin. Rheumatol. 26, 162–168 (2014).

    CAS  Article  Google Scholar 

  24. 24

    Kreuger, M. R. O., Grootjans, S., Biavatti, M. W., Vandenabeele, P. & D’Herde, K. Sesquiterpene lactones as drugs with multiple targets in cancer treatment: focus on parthenolide. Anticancer Drugs 23, 883–896 (2012).

    PubMed  Google Scholar 

  25. 25

    Oddoux, S. et al. Microtubules that form the stationary lattice of muscle fibres are dynamic and nucleated at Golgi elements. J. Cell Biol. 203, 205–213 (2013).

    CAS  Article  Google Scholar 

  26. 26

    Ralston, E., Lu, Z. & Ploug, T. The organization of the golgi complex and microtubules in skeletal muscle is fiber type-dependent. J. Neurosci. 19, 10694–10705 (1999).

    CAS  Article  Google Scholar 

  27. 27

    Gundersen, G. G., Khawaja, S. & Bulinski, J. C. Postpolymerization detyrosination of a-tubulin: a mechanism for subcellular differentiation of microtubules. J. Cell Biol. 105, 251–264 (1987).

    CAS  Article  Google Scholar 

  28. 28

    Belmadani, S., Poüs, C., Fischmeister, R. & Méry, P.-F. Post-translational modifications of tubulin and microtubule stability in adult rat ventricular myocytes and immortalized HL-1 cardiomyocytes. Mol. Cell Biochem. 258, 35–48 (2004).

    CAS  Article  Google Scholar 

  29. 29

    Whipple, R. A. et al. Parthenolide and costunolide reduce microtentacles and tumor cell attachment by selectively targeting detyrosinated tubulin independent from NF-κB inhibition. Breast Cancer Res. 15, R83 (2013).

    Article  Google Scholar 

  30. 30

    Fonrose, X. et al. Parthenolide inhibits tubulin carboxypeptidase activity. Cancer Res. 67, 3371–3378 (2007).

    CAS  Article  Google Scholar 

  31. 31

    Barisic, M. et al. Microtubule detyrosination guides chromosomes during mitosis. Science 348, 799–803 (2015).

    CAS  ADS  Article  Google Scholar 

  32. 32

    Ishibashi, Y. et al. Role of microtubules in myocyte contractile dysfunction during cardiac hypertrophy in the rat. Am. J. Physiol. 271, H1978–H1987 (1996).

    CAS  PubMed  Google Scholar 

  33. 33

    Song, Y. & Brady, S. T. Post-translational modifications of tubulin: pathways to functional diversity of microtubules. Trends Cell Biol. 25, 125–136 (2014).

    CAS  Article  Google Scholar 

  34. 34

    Janke, C. & Bulinski, J. C. Post-translational regulation of the microtubule cytoskeleton: mechanisms and functions. Nat. Rev. Mol. Cell Biol. 12, 773–786 (2011).

    CAS  Article  Google Scholar 

  35. 35

    Hammond, J. W. et al. Posttranslational modifications of tubulin and the polarized transport of kinesin-1 in neurons. Mol. Biol. Cell 21, 572–583 (2010).

    CAS  Article  Google Scholar 

  36. 36

    Webster, D. R. & Patrick, D. L. Beating rate of isolated neonatal cardiomyocytes is regulated by the stable microtubule subset. Am. J. Physiol. Heart Circ. Physiol. 278, H1653–H1661 (2000).

    CAS  Article  Google Scholar 

  37. 37

    Webster, D. R. Regulation of post-translationally modified microtubule populations during neonatal cardiac development. J. Mol. Cell Cardiol. 29, 1747–1761 (1997).

    CAS  Article  Google Scholar 

  38. 38

    Webster, D. R. Neonatal rat cardiomyocytes possess a large population of stable microtubules that is enriched in post-translationally modified subunits. J. Mol. Cell Cardiol. 29, 2813–2824 (1997).

    CAS  Article  Google Scholar 

  39. 39

    Belmadani, S., Poüs, C., Clapier, R. V., Fischmeister, R. & Méry, P. Post-translational modifications of cardiac tubulin during chronic heart failure in the rat. Mol. Cell Biochem. 237, 39–46 (2002).

    CAS  Article  Google Scholar 

  40. 40

    Sato, H. et al. Microtubule stabilization in pressure overload cardiac hypertrophy. J. Cell Biol. 139, 963–973 (1997).

    CAS  Article  Google Scholar 

  41. 41

    Yamamoto, S. et al. Role of microtubules in the viscoelastic properties of isolated cardiac muscle. J. Mol. Cell Cardiol. 30, 1841–1853 (1998).

    CAS  Article  Google Scholar 

  42. 42

    Tsutsui, H., Ishihara, K. & Cooper, G. Cytoskeletal role in the contractile dysfunction of hypertrophied myocardium. Science 260, 682–687 (1993).

    CAS  ADS  Article  Google Scholar 

  43. 43

    Prota, A. E. et al. Structural basis of tubulin tyrosination by tubulin tyrosine ligase. J. Cell Biol. 200, 259–270 (2013).

    CAS  Article  Google Scholar 

  44. 44

    Cheng, G. et al. Cytoskeletal role in protection of the failing heart by B-adrenergic blockade. Am. J. Physiol. Heart Circ. Physiol. 302, H675–H687 (2012).

    CAS  Article  Google Scholar 

  45. 45

    Koide, M. et al. Microtubule depolymerization normalizes in vivo myocardial contractile function in dogs with pressure-overload left ventricular hypertrophy. Circulation 102, 1045–1052 (2000).

    CAS  Article  Google Scholar 

  46. 46

    Ishibashi, Y. et al. Role of microtubules versus myosin heavy chain isoforms in contractile dysfunction of hypertrophied murine cardiocytes. Am. J. Physiol. Heart Circ. Physiol. 285, H1270–H1285 (2003).

    CAS  Article  Google Scholar 

  47. 47

    Zile, M. R. et al. Cardiocyte cytoskeleton in patients with left ventricular pressure overload hypertrophy. J. Am. Coll. Cardiol. 37, 1080–1084 (2001).

    CAS  Article  Google Scholar 

  48. 48

    Cooper, G. Cytoskeletal networks and the regulation of cardiac contractility: microtubules, hypertrophy, and cardiac dysfunction. Am. J. Physiol. Heart Circ. Physiol. 291, H1003–H1014 (2006).

    CAS  Article  Google Scholar 

  49. 49

    Stones, R., Benoist, D., Peckham, M. & White, E. Microtubule proliferation in right ventricular myocytes of rats with monocrotaline-induced pulmonary hypertension. J. Mol. Cell Cardiol. 56, 91–96 (2013).

    CAS  Article  Google Scholar 

  50. 50

    Zhang, C. et al. Microtubule-mediated defects in junctophilin-2 trafficking contribute to myocyte transverse-tubule remodeling and Ca2+ handling dysfunction in heart failure. Circulation 129, 1742–1750 (2014).

    CAS  Article  Google Scholar 

  51. 51

    Imazio, M. et al. Colchicine for prevention of postpericardiotomy syndrome and postoperative atrial fibrillation: the COPPS-2 randomized clinical trial. JAMA. 312, 1016–1023 (2014).

    CAS  Article  Google Scholar 

  52. 52

    Imazio, M. et al. Colchicine reduces postoperative atrial fibrillation: results of the Colchicine for the Prevention of the Postpericardiotomy Syndrome (COPPS) atrial fibrillation substudy. Circulation 124, 2290–2295 (2011).

    CAS  Article  Google Scholar 

  53. 53

    Neelakantan, S., Nasim, S., Guzman, M. L., Jordan, C. T. & Crooks, P. a. Aminoparthenolides as novel anti-leukemic agents: Discovery of the NF-κB inhibitor, DMAPT (LC-1). Bioorg. Med. Chem. Lett. 19, 4346–4349 (2009).

    CAS  Article  Google Scholar 

  54. 54

    Jung, C., Martins, A. S., Niggli, E. & Shirokova, N. Dystrophic cardiomyopathy: amplification of cellular damage by Ca2+ signalling and reactive oxygen species-generating pathways. Cardiovasc. Res. 77, 766–773 (2008).

    CAS  Article  Google Scholar 

  55. 55

    Fauconnier, J. et al. Leaky RyR2 trigger ventricular arrhythmias in Duchenne muscular dystrophy. Proc. Natl Acad. Sci. USA 107, 1559–1564 (2010).

    CAS  ADS  Article  Google Scholar 

  56. 56

    Ather, S. et al. Inhibition of CaMKII phosphorylation of RyR2 prevents inducible ventricular arrhythmias in mice with Duchenne muscular dystrophy. Heart. Rhythm. 10, 592–599 (2013).

    Article  Google Scholar 

  57. 57

    Iribe, G. et al. Axial stretch of rat single ventricular cardiomyocytes causes an acute and transient increase in Ca2+ spark rate. Circ. Res. 104, 787–795 (2009).

    CAS  Article  Google Scholar 

  58. 58

    Amaiden, M. R. et al. Effects of detyrosinated tubulin on Na+,K+-ATPase activity and erythrocyte function in hypertensive subjects. FEBS Lett. 589, 364–373 (2015).

    CAS  Article  Google Scholar 

  59. 59

    Parker, K. K. & Ingber, D. E. Extracellular matrix, mechanotransduction and structural hierarchies in heart tissue engineering. Philos. Trans. R Soc. Lond. B Biol. Sci. 362, 1267–1279 (2007).

    CAS  Article  Google Scholar 

  60. 60

    Brangwynne, C. P. et al. Microtubules can bear enhanced compressive loads in living cells because of lateral reinforcement. J. Cell Biol. 173, 733–741 (2006).

    CAS  Article  Google Scholar 

  61. 61

    Fassett, J. T. et al. Microtubule Actin Cross-linking Factor 1 regulates cardiomyocyte microtubule distribution and adaptation to hemodynamic overload. PLoS ONE 8, e73887 (2013).

    CAS  ADS  Article  Google Scholar 

  62. 62

    Raith, M. et al. Linking cytoarchitecture to metabolism: sarcolemma-associated plectin affects glucose uptake by destabilizing microtubule networks in mdx myofibers. Skelet. Muscle 3, 14 (2013).

    CAS  Article  Google Scholar 

  63. 63

    Banks, G. B., Combs, A. C., Odom, G. L., Bloch, R. J. & Chamberlain, J. S. Muscle structure influences utrophin expression in mdx mice. PLOS Genet. 10, e1004431 (2014).

    Article  Google Scholar 

  64. 64

    Peris, L. et al. Tubulin tyrosination is a major factor affecting the recruitment of CAP-Gly proteins at microtubule plus ends. J. Cell Biol. 174, 839–849 (2006).

    CAS  Article  Google Scholar 

  65. 65

    Curry, E. A. et al. Phase I dose escalation trial of feverfew with standarized doses of parthenolide in patients with cancer. Invest. New Drugs 22, 299–0305 (2004).

    CAS  Article  Google Scholar 

  66. 66

    Mitra, R. & Morad, M. A uniform enzymatic method for dissociation of myocytes from hearts and stomachs of vertebrates. Am. J. Physiol. 249, H1056–H1060 (1985).

    CAS  PubMed  Google Scholar 

  67. 67

    Kerr, J. P. et al. Dysferlin stabilizes stress-induced Ca2+ signaling in the transverse tubule membrane. Proc. Natl Acad. Sci. USA 110, 20831–20836 (2013).

    CAS  ADS  Article  Google Scholar 

  68. 68

    Picht, E., Zima, A. V., Blatter, L. A. & Bers, D. M. SparkMaster: automated calcium spark analysis with ImageJ. Am. J. Physiol. Cell Physiol. 293, C1073–C1081 (2007).

    CAS  Article  Google Scholar 

  69. 69

    Blau, H. M., Chiu, C.-P. & Webster, C. Cytoplasmic activation of human nuclear genes in stable heterocaryons. Cell 32, 1171–1180 (1983).

    CAS  Article  Google Scholar 

  70. 70

    Hexum, J. K. et al. Parthenolide prodrug LC-1 slows growth of intracranial glioma. Bioorg. Med. Chem. Lett. 25, 2493–2495 (2015).

    CAS  Article  Google Scholar 

Download references

Acknowledgements

Support was provided by the NIH (R01-AR062554 to CWW; R00 HL114879 to BLP; T32-AR07592 to J.P.K.; T32-AR053461 to PR), Hyundai Hope on Wheels (Hope Grant award to D.A.H.) and The V Foundation for Cancer Research (V Scholar award to D.A.H.). We thank Brunella Tedesco for her assistance with the AFM measurements. Assistance with the instrumentation and reagents for the myofibre stretch studies was kindly provided by IonOptix LLC and Aurora Scientific Inc. We thank E. Michael Ostap for careful review of the manuscript.

Author information

Affiliations

Authors

Contributions

J.P.K., B.L.P., P.R., G.S., R.R., N.B. and C.W.W. performed experiments; A.M.K., J.K.H., S.S.M. and D.A.H. contributed reagents; J.P.K., B.L.P. and C.W.W. wrote the manuscript; all authors interpreted data and edited the manuscript.

Corresponding authors

Correspondence to Benjamin L. Prosser or Christopher W. Ward.

Ethics declarations

Competing interests

B.L.P. and C.W.W. hold a US patent for MyoTak adhesive (patent number US 20120034620 A1), licensed to IonOptix LLC. C.W.W. holds a US patent for microtubule-targeted interventions for the muscular dystrophies (patent number US 2014015664 A1). The remaining authors declare no competing financial interests.

Supplementary information

Supplementary Information

Supplementary Figures 1-6 (PDF 953 kb)

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kerr, J., Robison, P., Shi, G. et al. Detyrosinated microtubules modulate mechanotransduction in heart and skeletal muscle. Nat Commun 6, 8526 (2015). https://doi.org/10.1038/ncomms9526

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing