Decreasing microtubule detyrosination modulates Na_v1.5 subcellular distribution and restores sodium current in mdx cardiomyocytes

Abstract

Aims

The microtubule (MT) network plays a major role in the transport of the cardiac sodium channel Na_v1.5 to the membrane, where the latter associates with interacting proteins such as dystrophin. Alterations in MT dynamics are known to impact on ion channel trafficking. Duchenne muscular dystrophy (DMD), caused by dystrophin deficiency, is associated with an increase in MT detyrosination, decreased sodium current (I_Na), and arrhythmias. Parthenolide (PTL), a compound that decreases MT detyrosination, has shown beneficial effects on cardiac function in DMD. We here investigated its impact on I_Na and Na_v1.5 subcellular distribution.

Methods and results

Ventricular cardiomyocytes (CMs) from wild-type (WT) and mdx (DMD) mice were incubated with either 10 µM PTL, 20 µM EpoY, or dimethylsulfoxide (DMSO) for 3–5 h, followed by patch-clamp analysis to assess I_Na and action potential (AP) characteristics in addition to immunofluorescence and stochastic optical reconstruction microscopy (STORM) to investigate MT detyrosination and Na_v1.5 cluster size and density, respectively. In accordance with previous studies, we observed increased MT detyrosination, decreased I_Na and reduced AP upstroke velocity (V_max) in mdx CMs compared to WT. PTL decreased MT detyrosination and significantly increased I_Na magnitude (without affecting I_Na gating properties) and AP V_max in mdx CMs, but had no effect in WT CMs. Moreover, STORM analysis showed that in mdx CMs, Na_v1.5 clusters were decreased not only in the grooves of the lateral membrane (LM; where dystrophin is localized) but also at the LM crests. PTL restored Na_v1.5 clusters at the LM crests (but not at the grooves), indicating a dystrophin-independent trafficking route to this subcellular domain. Interestingly, Na_v1.5 cluster density was also reduced at the intercalated disc (ID) region of mdx CMs, which was restored to WT levels by PTL. Treatment of mdx CMs with EpoY, a specific MT detyrosination inhibitor, also increased I_Na density, while decreasing the amount of detyrosinated MTs, confirming a direct mechanistic link.

Conclusion

Attenuating MT detyrosination in mdx CMs restored I_Na and enhanced Na_v1.5 localization at the LM crest and ID. Hence, the reduced whole-cell I_Na density characteristic of mdx CMs is not only the consequence of the lack of dystrophin within the LM grooves but is also due to reduced Na_v1.5 at the LM crest and ID secondary to increased baseline MT detyrosination. Overall, our findings identify MT detyrosination as a potential therapeutic target for modulating I_Na and subcellular Na_v1.5 distribution in pathophysiological conditions.

Graphical Abstract

Open in new tabDownload slide

This manuscript was handled by Consulting Editor David Eisner. Time of primary review: 29 days See the editorial comment for this article ‘Microtubules: highway to…arrhythmia?’, by A. Liutkute et al., https://doi.org/10.1093/cvr/cvae072.

1. Introduction

Microtubules (MTs) are part of the cytoskeletal network, consist of dynamic polymers of α- and β-tubulin, and mediate cell shape, polarization, mitosis, axonemal-based motility, and morphogenesis.¹ MTs furthermore constitute transportation routes within the cell, trafficking proteins such as ion channels to the plasma membrane. MT function is dynamically regulated by post-translational modifications such as tyrosination/detyrosination, acetylation, and glycosylation.² While MTs are mostly tyrosinated during control conditions,³ increased MT detyrosination is observed in the setting of heart failure (HF),⁴ hypertrophic cardiomyopathy (HCM),⁵ and Duchenne muscular dystrophy (DMD),^6–11 together with an increase in density and disorganization of the MT network.^6,8,10–15 These pathological conditions are also associated with a pro-arrhythmic reduction in cardiac sodium current (I_Na).^4,16–18 We have previously shown that taxol, a compound with anti-cancer properties that increases the amount of detyrosinated MTs, decreases I_Na and membrane localization of the cardiac sodium channel (Na_v1.5) in cardiomyocytes (CMs).¹⁹ Hence, reduction of MT detyrosination may have beneficial electrophysiological effects in the setting of pathological conditions by restoring I_Na.

Parthenolide (PTL) is a sesquiterpene lactone found in feverfew plants that has been used for centuries to treat fever, headache, pain, and migraine²⁰ and currently is predominantly investigated for its anti-cancer and anti-inflammatory properties.^21–25 More recently, PTL has been shown to decrease MT detyrosination, reduce membrane stiffness and improve CM contraction in models of HF and HCM.^5,26,27 In addition, PTL demonstrated beneficial effects in mdx (DMD) mice, improving calcium handling, attenuating reactive oxygen species production and reducing arrhythmia occurrence.⁷ These mdx mice lack dystrophin, a protein localized exclusively at the lateral membrane (LM) of CMs where it interacts with Na_v1.5 and is essential for its localization and function^{17,18,28–31} while binding directly to MTs.^12,13 Previous studies have indicated that loss of dystrophin in mdx CMs decreases Na_v1.5 localization predominantly at the LM,^29,32 particularly within the grooves (invaginations) of the LM where dystrophin is normally located.²⁹ However, the impact of MT remodelling on I_Na and subcellular Na_v1.5 distribution in mdx CMs has not been investigated. In addition, it is as yet unknown whether restoring MT dynamics in mdx CMs has beneficial effects on Na_v1.5/I_Na.

Using super-resolution microscopy (STORM), we demonstrate that in mdx CMs, Na_v1.5 clusters are decreased not only in the grooves of the LM but also at the LM crests and intercalated disc (ID) region. Attenuating MT detyrosination in mdx CMs restored I_Na and enhanced Na_v1.5 localization at the LM crest and ID. Hence, the reduced whole-cell I_Na density characteristic of mdx CMs is not only caused by loss of dystrophin-binding Na_v1.5 within the LM grooves but is also due to reduced Na_v1.5 at the LM crest and ID secondary to increased baseline MT detyrosination. A more specific inhibitor of MT detyrosination, EpoY, also increased I_Na density, confirming a direct mechanistic link. Overall, our findings identify regulation of MT dynamics as an anti-arrhythmic target in the setting of pathophysiological conditions associated with increased MT detyrosination and decreased I_Na.

2. Methods

2.1 Mouse model

Male control mice (C57BL/10ScSnJ) and dystrophin-deficient mice (C57/10ScSn-Dmd^mdx/J; mdx) were acquired from Jackson Laboratories (Bar Harbor) and used at 4–7 months of age. All methods were carried out in accordance with relevant guidelines (including ARRIVE guidelines), and the study design and all animal handling and experiments were approved by governmental and Institutional Animal Care and Use Committees of the University of Amsterdam (license 18–4986).

2.2 Cardiomyocyte isolation and drug incubation

After receiving an intraperitoneal injection of 20 μL heparin (5000 IU/mL), mice were sacrificed by cervical dislocation under anaesthesia (1% O₂ and 4% isoflurane), and the heart was excised. The isolated heart was retrogradely perfused for 10 min in a Langendorff system with oxygenated modified Tyrode’s solution (37°C) containing (in mM): NaCl 140, KCl 5.4, CaCl₂ 1.8, MgCl₂ 1.0, glucose 5.5, and HEPES 5.0; pH 7.4 (NaOH). Next, the heart was perfused for 10 min with Ca²⁺-free modified Tyrode’s solution containing 0.01 mM CaCl₂, supplemented with 10.7 mM creatine monohydrate (Sigma, C3630). Subsequently, the heart was digested using Liberase TM (0.18–0.21 U/mL; Roche) and elastase from porcine pancreas (2.5–2.9 U/mL; SERVA) in Ca²⁺-free modified Tyrode’s solution for around 12 min. The heart was then placed in Ca²⁺-free modified Tyrode’s solution containing 1% bovine serum albumin (BSA) (Sigma-Aldrich). Atria and right ventricular free wall were removed, and the left ventricle was mechanically dissociated to obtain single CMs. Isolated CMs were carefully washed with Ca²⁺-free modified Tyrode’s solution to remove BSA, and subsequently incubated at room temperature with 10 μM parthenolide (PTL; Sigma-Aldrich) or 20 μM EpoY (EpoY; Sigma-Aldrich) both dissolved in dimethylsulfoxide (DMSO), or DMSO as control. CMs were used for patch-clamp analysis or fixed for immunocytochemistry experiments between 3 and 5 h after the start of incubation.

2.3 Confocal microscopy

2.3.1 Immunocytochemistry

CMs were plated on laminin-coated coverslips and left to adhere for at least 30 min before fixation with 100% methanol. Cells were permeabilized with 0.1% Triton in phosphate-buffered saline (PBS) for 10 min, followed by a blocking step in PBS containing 2% glycine, 2% BSA, and 0.2% gelatin, of 30–45 min. Anti-α-tubulin antibody (Santa Cruz Biotechnology B-7 SCBT, 1:250 in blocking solution) and anti-detyrosinated α-tubulin antibody (Millipore Sigma AB3201, 1:200 in blocking solution) were incubated in fixed cells at room temperature for 1 h. Cells were then washed three times with PBS and subsequently incubated for 1 h at room temperature with the secondary anti-rabbit antibody conjugated with Alexa Fluor 488 (Invitrogen A-11008, 1:300 in blocking solution) and anti-mouse antibody conjugated with Alexa Fluor 568 (Invitrogen A-11004, 1:300 in blocking solution). Negative controls were performed under the same conditions but without primary antibodies.

2.3.2 Imaging

Images were acquired using a confocal microscope Leica SP8 with an HC PL APO CS2 63×/1.40 oil objective, spectral photo multiplier tube detector and with 488 and 568 nm lasers. To examine the amount of total α-tubulin and detyrosinated α-tubulin in cells, maximum intensity projections were generated using Leica Application Suite X (LAS X) software and the fluorescence intensity of an entire cell was measured. α-Tubulin and detyrosinated α-tubulin were visualized on external planes, towards the cell end.

2.3.3 Quantification

α-Tubulin and detyrosinated α-tubulin levels were quantified using LAS X software. For each CM, a region of interest (ROI) was manually drawn around the entire cell and mean fluorescence intensity within the ROI was calculated for each channel and background subtraction was applied.

2.4 Multi-colour super-resolution imaging by STORM

2.4.1 Immunocytochemistry

CMs were plated on laminin-coated coverslips and left to adhere for at least 30 min before fixation with 100% methanol. Cells were permeabilized with 0.1% Triton in PBS for 10 min, followed by a blocking step in PBS containing 2% glycine, 2% BSA and 0.2% gelatin, for 30–45 min. The primary rabbit polyclonal antibody against Na_v1.5 (Sigma-Aldrich S0819, 1:100 in blocking solution) and the mouse monoclonal antibody against α-actinin (Sigma-Aldrich A7811, 1:200 in blocking solution) were incubated for 1 h at room temperature. After washing three times with PBS, the secondary anti-mouse antibody conjugated with Alexa Fluor 647 (Invitrogen A-28181, 1:15 000) and anti-rabbit antibody conjugated with Alexa Fluor 568 (Invitrogen A-11011; 1:15 000) were incubated for 15 min at room temperature. Post staining, the coverslips were mounted on microscope glass slides with predrilled holes for imaging buffer exchange. Immediately before imaging, freshly prepared super-resolution imaging buffer consisting of 1 mg/mL glucose oxidase (Sigma, G2133), 0.02 mg/mL catalase (Sigma, C3155), 10% glucose (Sigma, G8270), and 100 mM mercaptoethylamine (Fisher Scientific, 100995) in PBS (pH 8.0), was injected.

2.4.2 Imaging

Samples were imaged using a custom-built platform based on an inverted microscopy setup (ASI-RAMM) similar to the scheme described before.³³ In brief, STORM imaging system was coupled to three excitation laser lines, 488 nm (Coherent, Sapphire 488 LPX), 561 nm (Coherent, Sapphire 561 LPX), and 639 nm (Ultralaser, MRL-FN-639–1.2).³⁴ Lasers were combined using dichroic mirrors and focused onto the back aperture of an oil immersion objective (Olympus, UApo N, 100×, NA = 1.49, TIRF) via multiband dichroic mirror (Semrock, 408/504/581/667/762-Di01). Samples were excited via Highly Inclined and Laminated Optical sheet illumination mode. Fluorescence emission collected back through the objective was directed to an sCMOS camera (Photometrics, Prime 95B). Acquisitions consist of a minimum of 2000 frames at 30 ms exposure per frame, with each channel captured sequentially using band pass filters for DiO (Semrock, FF01–531/40), AF568 (Semrock, FF01–607/36), and AF647 (Semrock, FF01-676/37) that were positions in an emission filter wheel (ASI) prior to sCMOS camera. A 405 nm laser line (CNI Laser, MDL-III-405-500) was used to enhance recovery of dark state AF647 fluorophores during acquisition. Images were acquired using Micro-Manager (v1.4) software. Images were mapped to correct for chromatic aberrations using a polynomial morph-type mapping algorithm via a custom Matlab script. Before each experiment, a calibration map was generated by imaging fluorescent beads in each channel. A second polynomial function was optimized to fit localizations in AF568 channels to their location in the AF647 channel.³⁵ This function is used to map molecule localizations of each experimental sample. Movies for each imaging were submitted to a home-built software in MATLAB (version R2018b, MathWorks) for precise single-molecule localization. Each frame of a raw image stack was box filtered with a box size of four times the FWHM of a 2D-Gaussian PSF. 7 × 7 pixel regions around local maxima from all frames were submitted for 2D-Gaussian multi-PSF fitting, performed by GPU, using maximum likelihood estimation algorithm (DAO-STORM). The fitting accuracy was estimated by Cramer-Rao lower bound (CRLB). Localizations which appear in consecutive frames within 2.5 × the localization precision were considered one blinking event and averaged into one localization weighted by the inverse of its own CRLB.^36,37 Note that the likelihood function was constructed by combining the Poisson shot-noise distribution and pre-calibrated pixel-specific Gaussian readout-noise distribution.³⁶ For display purposes, the representative images were generated by rendering the raw coordinates into 10 nm pixel canvas and convolved with a 2D-Gaussian (σ = 10 nm) kernel.

2.4.3 Cluster analysis

As previously described,³⁸ reconstructed super-resolved images were processed with a smoothing filter (‘Gaussian blur’ function in ImageJ version 1.52j), adjusted for brightness and contrast and filtered to a threshold to obtain a binary image. For the experimental datasets, ROIs were manually drawn for each image and saved as an ImageJ ROI file. LM clusters were defined as the clusters localized along the lateral edges of the CMs and ID clusters were defined as clusters localized at the cell ends. Cluster detection and parameters were obtained using the ImageJ function ‘Analyse particles’ as previously described.³⁹ Since the average background cluster size in the negative controls was <7000 nm², the minimum size/area of contiguous pixels defined as a ‘real’ cluster was 7000 nm². Signal-positive areas of smaller dimensions were considered as ‘background’ and not included. GraphPad Prism 7 was used for data visualization.

2.4.4 Distance analysis

Distances from Na_v1.5 clusters to α-actinin were measured by an automated script in Python, as previously described.³⁸ The images from the super-resolution fluorescence microscopy and their corresponding ROI files provided the input to the script. The script utilized the image processing packages scikit-image,⁴⁰ and ‘Mahotas’ (version 1.2), an open-source software for scriptable computer vision (http://dx.doi.org/10.5334/jors.ac). Two clusters were considered separate if one was at least 20 nm (1 pixel) apart from another in any direction.

2.5 Electrophysiology

2.5.1 Data acquisition and analysis

I_Na and action potentials (APs) were measured with the ruptured and perforated patch-clamp technique, respectively, using an Axopatch 200B amplifier (Molecular Devices). Voltage control, data acquisition and analysis were performed with pClamp10.6/Clampfit (Molecular Devices) and custom-made software for I_Na and APs, respectively. Borosilicate glass patch pipettes (Harvard Apparatus) with a tip resistance of 2–2.2 MΩ were used. Series resistance (Rs) and cell membrane capacitance (Cm) were compensated for ≥80%. Cm was determined by dividing the decay time constant of the capacitive transient in response to 5 mV hyperpolarizing steps from −40 mV, by the Rs. I_Na and APs were filtered at 5 kHz and digitized at 40 kHz.

2.5.2 Sodium current measurements

I_Na was measured at room temperature using a pipette solution containing (in mM): NaCl 3.0, CsCl 133, MgCl₂ 2.0, Na₂ATP 2.0, TEACl 2.0, EGTA 10.0, HEPES 5.0; pH 7.3 (CsOH). Measurements were performed in a bath solution containing (in mM): NaCl 7.0, CsCl 133, CaCl₂ 1.8, MgCl₂ 1.2, glucose 11.0, HEPES 5.0, nifedipine 0.005; pH 7.4 (CsOH). Current–voltage (I–V) relationships, voltage dependence of (in)activation and recovery from inactivation were determined using voltage-clamp protocols described in the relevant figures. In all protocols, a holding potential of −120 mV and a cycle length of 5 s were used. I_Na was defined as the difference between peak current and steady-state current. I_Na density was calculated dividing the current amplitude by the Cm. Na_v1.5 voltage dependence of activation and inactivation curves were fitted with a Boltzmann equation (y = [1 + exp{(V − V_1/2)/k}]⁻¹), where V_1/2 is the half-maximal voltage of (in)activation, and k the slope factor. Recovery from inactivation was analysed by fitting the data with a two-exponential function (y = y₀ + A_f{1−exp[−t/τ_f]} + A_s{1−exp[−t/τ_s]}), where A_f and A_s are the amplitudes of the fast and the slow components of recovery from inactivation, and τ_f and τ_s represent their respective recovery time constants.

2.5.3 Action potential measurements

APs were measured at 36°C using a modified Tyrode’s solution containing (in mM): NaCl 140, CaCl₂ 1.8, MgCl₂ 1.0, KCl 5.4, glucose 5.5, HEPES 5; pH 7.4 (NaOH) as bath solution. Pipettes were filled with (in mM): K-gluconate 125, KCl 20.0, NaCl 5.0, amphotericin-B 0.44, HEPES 10, pH 7.2 (KOH). APs were elicited at 2 Hz by 3-ms, ≈1.2× threshold current pulses through the patch pipette. We analysed resting membrane potential (RMP), AP amplitude (APA), maximal AP upstroke velocity (V_max), and AP duration (APD) at 20, 50, and 90% repolarization (APD₂₀, APD₅₀, and APD₉₀, respectively). Data from 10 consecutive APs were averaged, and potentials were corrected for the calculated liquid junction potential of 15 mV.⁴¹

2.6 Statistical analysis

Data were analysed using SigmaStat software version 3.5 (Systat Software Inc.) and GraphPad Prism version 8.4.3(686) for Windows (GraphPad Software). Values are shown as mean ± standard error of the mean. Normality was tested by Kolmogorov–Smirnov test. Unpaired Student’s t-test was used in case of normally distributed data. Mann–Whitney test was applied when normality and/or equal variance test failed. One-way ANOVA followed by Holm–Sidak test for pairwise multiple comparison or Kruskal–Wallis followed by Dunn’s test when data were not normally distributed, were used to compare multiple groups. Statistical significance for differences in current–voltage relations (I–V) curves was determined by performing a two-way repeated measures ANOVA, followed by Holm–Sidak test for post hoc analysis. The level of statistical significance was set to P < 0.05.

3. Results

3.1 Increased MT detyrosination in mdx CMs is attenuated by PTL

We first assessed the level of total α-tubulin and detyrosinated α-tubulin in ventricular mdx and WT CMs through immunocytochemistry. Isolated CMs were incubated for 3–5 h with either DMSO or PTL (10 μM) followed by immunolabelling with total α-tubulin and detyrosinated α-tubulin antibodies (Figure 1A). For analysis, we measured the mean fluorescence intensity of total α-tubulin, detyrosinated α-tubulin and then calculated the ratio of detyrosinated on total α-tubulin. These immunofluorescence experiments revealed a statistically significant increase of detyrosinated α-tubulin in mdx CMs compared to WT CMs, a clear trend towards an increase in total α-tubulin, in line with previous findings showing an increase in density of the MT network,^{6,7,10,11,14,15} and a significant increase in the ratio of detyrosinated α-tubulin on total α-tubulin (Figure 1B). While 3–5 h incubation with PTL had no effect on WT CMs, it significantly decreased detyrosinated α-tubulin, total α-tubulin, and their ratio in mdx CMs (Figure 1B). In addition to confirming the previously observed increased level of detyrosinated MTs in mdx CMs,⁶ these data demonstrate the ability of PTL to attenuate MT deyrosination and MT density in these cells.

3.2 PTL increases I_Na density and AP upstroke velocity in mdx but not WT CMs

We next assessed the effect of PTL on I_Na and AP properties in WT and mdx CMs. Firstly, we confirmed that I_Na and V_max were significantly decreased in mdx mice compared to WT (WT maximal peak I_Na: −61.8 ± 5.7 pA/pF; mdx maximal peak I_Na: −43.5 ± 2.2 pA/pF; 30% reduction, P = 0.02, Mann–Whitney test; WT V_max: 442.9 ± 17.3 V/s; mdx V_max: 379.0 ± 23.0 V/s; 15% reduction, P = 0.03, unpaired Student’s t-test; Figures 2A–D and 4A–D, Supplementary material online, Tables S1–S4), consistent with previous reports.^17,18,42,43 Secondly, we investigated the effect of 3–5 h incubation with PTL and found that while PTL had no effect on WT CMs (Figure 2A and C, Supplementary material online, Table S1), it significantly increased I_Na density in mdx CMs, restoring the current to WT level (Figure 2B and D, Supplementary material online, Table S2). PTL incubation did not affect I_Na voltage dependence of (in)activation, assessed as the half-maximal voltage of (in)activation (V_1/2) and the slope factor k, or recovery from inactivation in either WT or mdx CMs (Figure 3A–D, Supplementary material online, Tables S1 and S2), indicating that the increase in I_Na density is likely due to enhanced trafficking and is not an effect of PTL on e.g. channel conformation. PTL significantly increased APA and V_max in mdx CMs (Figure 4C and D, Supplementary material online, Table S4), restoring the latter to WT levels, while again not affecting WT CMs (Figure 4A and B, Supplementary material online, Table S3). Finally, RMP and APD at 20, 50, or 90% repolarization were not altered by PTL incubation either in WT (Figure 4A and B, Supplementary material online, Tables S3) or in mdx CMs (Figure 4C and D, Supplementary material online, Table S4).

3.3 Na_v1.5 cluster density is decreased at the LM and ID of mdx CMs and is restored by PTL

To investigate the impact of PTL on Na_v1.5 on a subcellular level, we performed single-molecule localization microscopy by STORM experiments on WT and mdx CMs, assessing Na_v1.5 cluster size and density at the LM and ID (Figure 5A and B). As previously shown,²⁹ Na_v1.5 cluster density and size were reduced at the LM in mdx CMs (Figure 5C and D), in line with the absence of dystrophin in this microdomain. Surprisingly, even though dystrophin is only located at the LM, we found that Na_v1.5 cluster density and size were also significantly reduced at the ID of dystrophin-deficient mdx CMs (Figure 5E and F), although this decrease was less prominent than at the LM. We then assessed the effect of PTL on Na_v1.5 cluster density and size. In line with confocal and patch-clamp results, we observed by STORM analysis that 3–5 h incubation with PTL did not affect Na_v1.5 cluster density or cluster size in WT CMs (Figure 5C–F), while in mdx CMs it significantly increased cluster density at both LM and ID (Figure 5C and E), without affecting cluster size (Figure 5D and F). These findings indicate that the reduced whole-cell I_Na density characteristic of mdx mice is not only the consequence of the lack of dystrophin-bound Na_v1.5 at the LM but is also due to reduced Na_v1.5 at the ID secondary to increased MT detyrosination.

3.4 PTL incubation increases Na_v1.5 clusters preferentially at the crests of LM in mdx CMs

The LM consists of crests interspersed by grooves, which are invaginations of the membrane. Most of Na_v1.5 at the LM of CMs are localized at the crests.^28,44 Under normal conditions, dystrophin co-localizes with Na_v1.5 at the groove;^{17,18,28–31} accordingly, in dystrophin-deficient mdx CMs Na_v1.5 clusters are decreased in the grooves but remain localized at the crests.^28,29 We therefore investigated whether following PTL treatment, Na_v1.5 could still be positioned at the grooves of LM or only at the crests. Since the grooves are anchored to the sarcomeric Z-line, of which α-actinin is the main component, we measured the distance from each α-actinin cluster to the closest Na_v1.5 cluster after incubation with either DMSO or PTL in WT and mdx CMs (Figure 6A and B). Na_v1.5 clusters within 50 nm from α-actinin were considered localized within grooves. As previously shown by Vermij et al.,²⁹ the proportion of Na_v1.5 clusters within 50 nm from α-actinin was found to be significantly reduced in mdx CMs as compared to WT CMs (WT mean: 28.55 ± 2.2% vs. mdx mean: 23.10 ± 1.6%; 19% reduction, P = 0.048, unpaired Student’s t-test). The frequency histogram (Figure 6C) shows that, in mdx CMs, PTL treatment further decreased the proportion of Na_v1.5 clusters within 50 nm from α-actinin in mdx CMs and, considering the proportion of Na_v1.5 clusters within 50 nm from α-actinin for each individual cell, there was a significant reduction (31%) compared to cells incubated with DMSO (Figure 6D), indicating that PTL increased Na_v1.5 clusters preferentially at the crests of LM of mdx CMs, while no effect was observed in WT CMs. Hence, reducing MT detyrosination promotes a dystrophin-independent redistribution of Na_v1.5 towards the LM crests.

3.5 EpoY increases I_Na density in mdx CMs confirming the modulatory effect of MT detyrosination

While we observed that PTL significantly decreased MT detyrosination in mdx CMs, it may also act on other pathways.^22,45 To rule out the possibility that the observed effects on I_Na density occur secondary to another mechanism, we also assessed the effects of EpoY, an inhibitor of vasohibin (VASH) and small vasohibin binding protein (SVBP), the enzyme complex responsible for α-tubulin detyrosination⁴⁶ and, therefore, a more specific pharmacological inhibitor of microtubule detyrosination. Incubation of mdx CMs for 3–5 h with 20 μM EpoY, a concentration previously shown to reduce the amount of detyrosinated tubulin,⁴⁶ specifically reduced the detyrosinated fraction of α-tubulin (Figure 7A and B). Similar to PTL, EpoY also significantly increased I_Na density without affecting Na_v1.5 voltage dependence of (in)activation (Figure 7C, Supplementary material online, Table S5), confirming a direct mechanistic link between MT detyrosination and I_Na.

4. Discussion

In the last decade, the subcellular distribution of Na_v1.5 and the differential interaction with distinct associating proteins within these domains has been increasingly investigated. At the LM of CMs, Na_v1.5 is known to be associated with syntrophin and dystrophin while at the ID it interacts with proteins such as plakophilin-2 and SAP97.^17,28–31 The functional relevance of these interactions within the distinct subcellular macromolecular complexes is underscored by the fact that loss of Na_v1.5-interacting proteins leads to decreased I_Na in the setting of pathophysiological disorders. While mutations in ID-related Na_v1.5-interacting proteins have been associated with arrhythmogenic cardiomyopathy, loss of the LM-based Na_v1.5-interacting protein dystrophin occurs in DMD.^47,48 This progressive muscular disorder affects not only skeletal muscle but also leads to cardiomyopathy as well as conduction disorders and cardiac arrhythmias caused at least in part by reduced I_Na.^{17,18,42,43,47} Hence, detailed insight into the mechanisms underlying DMD-related alterations in I_Na/Na_v1.5 is crucial. Previous studies demonstrated loss of Na_v1.5 specifically at the LM of mdx CMs,^29,32 in line with the specific localization of dystrophin at this subcellular domain. We here employed super-resolution (STORM) microscopy and found that Na_v1.5 was also decreased at the ID of mdx CMs, secondary to an increase in MT detyrosination. We furthermore show that reducing MT detyrosination in mdx CMs can also re-direct Na_v1.5 to the crest of the LM in a dystrophin-independent manner. Hence, we identify MT detyrosination as a potential target for modulating I_Na and subcellular Na_v1.5 distribution.

The magnitude of I_Na is larger at the ID as compared to the LM, and ID-based I_Na is considered essential for electrical propagation within the myocardium. Nevertheless, studies have shown that specific loss of I_Na/Na_v1.5 at the LM also compromises conduction.⁴⁹ The LM of CMs is divided into grooves, which harbour the T-tubules, interspersed by crests.^4,38,50 The grooves constitute invaginations of the LM anchored through costameres to the sarcomeric Z-line. Since α-actinin is the main component of the costameres, α-actinin filaments are considered a marker of the LM grooves. The Na_v1.5-dystrophin–syntrophin complex is localized at the grooves, yet only a small portion of the total I_Na originates from this subdomain,^4,28,51 with the majority of Na_v1.5 actually localizing at the crests.^28,44 As yet, it remains unknown which interacting proteins associate with Na_v1.5 at the crests, although an association with subsarcolemmal mitochondria has been described;³⁸ identifying these interacting protein(s) may uncover an interesting novel target for enhancing I_Na. In a previous study, Vermij et al.²⁹ demonstrated that in mdx CMs, Na_v1.5 is particularly reduced at the grooves in line with the loss of dystrophin in this subdomain, while the costamere structure is altered and the LM is flattened. Using STORM, we now show for the first time that not only Na_v1.5 cluster density but also cluster size is reduced at the LM of mdx CMs. The reduced Na_v1.5 cluster size is an important observation since smaller Na_v1.5 clusters have been linked to reduced functionality,^4,52 in line with the observed reduced I_Na density in mdx mice. Following incubation with PTL, which restored whole-cell I_Na, the percentage of Na_v1.5 clusters localized at the grooves of LM (defined as <50 nm distance from α-actinin) in mdx CMs was significantly decreased, indicating that PTL re-directed Na_v1.5 clusters to the crests in a dystrophin-independent manner. Hence, while the loss of dystrophin binding underlies the reduced Na_v1.5 at the grooves of mdx CMs, the decrease in Na_v1.5 at the crests may be explained, at least in part, by the increase in MT detyrosination. Dystrophin binds directly to MTs and the lack of dystrophin in mdx mice induces subsarcolemmal MT disorganization that impairs vesicular trafficking.¹³ This loss of dystrophin is associated with up-regulation of utrophin, the autosomal homologue of dystrophin, which also has been shown to play a role in the localization and function of Na_v1.5 at the LM.¹⁸ While it has been suggested that utrophin-based therapies may improve the mdx phenotype,^53–55 it has been demonstrated that utrophin is not able to bind MTs like dystrophin and that transgenic overexpression of utrophin is not able to rescue the MT disorganization in mdx mice.¹³

Interestingly, we observed that Na_v1.5 cluster density and size were reduced not only at the LM of mdx CMs but also at the ID, indicating that contrary to what was previously thought, the reduced I_Na density in these mice is not only due to the lack of dystrophin-Na_v1.5 interactions at the LM. Since PTL treatment restored I_Na as well as Na_v1.5 cluster density at the ID, we hypothesize that the increased MT detyrosination consequent to the loss of dystrophin leads to secondary remodelling of Na_v1.5 at other subcellular domains besides the LM, and that reduction of MT detyrosination promotes Na_v1.5 trafficking throughout the entire CM. The fact that PTL has no effect on either the level of detyrosinated MTs or on I_Na/Na_v1.5 clusters in WT CMs further underscores a link between this post-translational modification of α-tubulin and Na_v1.5 distribution and function. Hence, by targeting MT dynamics, Na_v1.5 trafficking can be (re)directed towards the sarcolemma in dystrophin-deficient CMs, thereby enhancing sodium channel availability, as demonstrated by our observation that PTL increased I_Na and AP upstroke velocity. These observations furthermore indicate that MT detyrosination may be an important modulatory link between alterations initially affecting one specific microdomain and subsequent remodelling of many processes throughout the entire CM. Indeed, increased MT detyrosination was previously associated with detrimental changes in contraction/relaxation and calcium homeostasis in mdx mice.^6,7,11 Hence, preventing MT detyrosination may prove to have important beneficial effects in maintaining electrical, contractile, and homeostatic balance, and as such prevent arrhythmias, in the setting of pathophysiological disorders, including HF and HCM.^4,5

The relation between modifications of the MT network, ion channel trafficking and plasma membrane expression has been demonstrated in studies using different agents affecting MT structure and/or dynamics.^56–59 Previously, we demonstrated that taxol, which has exactly the opposite effect on MTs as compared to PTL since it actually increases the amount of detyrosinated MTs, decreased I_Na in neonatal rat cardiomycoytes,¹⁹ strengthening our hypothesis of a mechanistic link between MT detyrosination and I_Na. Yet, PTL is known to impact on multiple pathways,^{5,7,22,24–27,45} necessitating assessment of this link using a more direct approach. Two enzymes are known to play a major role in the tyrosination/detyrosination cycle of MTs: tubulin-tyrosine carboxypeptidase, shown to be a complex of vasohibin and the small vasohibin binding protein (VASH-SVBP), responsible for the removal of tyrosine from MTs (i.e. detyrosination)^60,61 and tubulin-tyrosine ligase which adds back tyrosine to MTs.⁶² Hence, we also tested the effects of EpoY, an inhibitor of the VASH-SVBP and therefore a more specific pharmacological blocker of MT detyrosination.⁴⁶ Similar to PTL, EpoY significantly decreased detyrosinated α-tubulin levels and increased I_Na density in mdx CMs, confirming their mechanistic link. While PTL was originally thought to directly block the VASH-SVBP enzyme complex, this has been disproved recently,⁴⁶ and hence the exact mechanism by which PTL reduces MT detyrosination remains to be fully elucidated. PTL has been shown to bind directly to tubulin and covalently modify it, thereby destabilizing MTs;⁴⁶ indeed our results show that PTL also affects total α-tubulin levels. In contrast, EpoY inhibits specifically the VASH-SVBP complex and consequently only prevents detyrosination of the α-tubulin fraction but does not affect total α-tubulin levels.⁴⁶ While the mechanisms by which PTL and EpoY decrease MT detyrosination may differ, both interventions successfully increased I_Na density in mdx CMs. Crucially, no effects of PTL were observed in WT CMs (which display very low levels of MT detyrosination), indicating that this approach may specifically affect CMs with a pathological increase in MT detyrosination.

Multiple mechanisms may underlie the modulatory effects of increased MT detyrosination on trafficking and subcellular distribution of Na_v1.5. Kinesins and dyneins are motor proteins responsible for MT-dependent transport of e.g. ion channel proteins towards the membrane (anterograde transport) and backwards from the membrane (retrograde transport), respectively.¹⁴ While kinesin binds with higher affinity to detyrosinated MTs, the latter actually reduces the stepping and detachment rates of kinesin resulting in longer processive runs and slower motility.⁶³ The slower movement of kinesin on detyrosinated MTs compared to tyrosinated MTs, which has also been reported for dynein,^3,64 suggests that trafficking of Na_v1.5 is slower in mdx CMs, leading to reduced I_Na. Another possible mechanism relates to MT plus (+)-end tracking proteins, some of which, cytoplasmic linker CLIP170, MT end-binding protein 1 (EB1), and cytoskeleton-associated protein glycine-rich (CAP-Gly), are known to be important for the delivery of Na_v1.5 to the membrane.⁶⁵ CAP-Gly was found to be more present on tyrosinated MTs,⁶⁶ and the interaction between CLIP170 and EB1 has been shown to be facilitated on tyrosinated MTs,^66–69 indicating that tyrosinated MTs could facilitate the release of Na_v1.5 at its membrane destination. Elucidation of the exact mechanism(s) involved will require future in-depth biophysical studies.

In conclusion, we here show that I_Na reduction in mdx CMs is not only the consequence of the lack of dystrophin within the LM grooves but is also due to a decrease in Na_v1.5 at the LM crest and ID secondary to increased MT detyrosination. Overall, our findings point to a modulatory role for post-translational modifications of MTs impacting MT dynamics in the subcellular distribution of Na_v1.5, and identify MT detyrosination as an anti-arrhythmic target in the setting of pathophysiological conditions associated with decreased I_Na.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Acknowledgements

We thank Sagun Jonchhe, Dipika Gupta, and Shaun Christie from the Department of Biochemistry and Pharmacology of the NYU School of Medicine for expert technical support, Dr Michael Tanck from the Department of Epidemiology and Data Science (Amsterdam UMC) for statistical support and Dr Diederik Kuster (Department of Physiology, VUmc) for kindly providing EpoY.

Funding

This work was supported by Fondation Leducq (17CVD02), the Netherlands CardioVascular Research Initiative (an initiative with support of the Dutch Heart Foundation; CVON2015-12 eDETECT and CVON2018-30 PREDICT2), and the Netherlands Organisation for Health Research and Development (ZonMw; Innovational Research Incentives Scheme Vidi grant 91714371). The Rothenberg lab is funded in part by grants from the NIH (1R35GM134947-01, 1R01AI153040-01, and 1P01CA247773-01/549).

Data availability

The data underlying this article are available in the article and in its online supplementary material.

References

Author notes