All animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Maryland (IACUC) and adhere to NIH guidelines. All mice were obtained from Jackson Laboratories (Bar Harbor, MA) and include wild-type (C57BL.10/J, strain #000665) and mdx (C57BL.10/J mdx, strain #001801) mice at 4 and 16 weeks of age.

Homogenized cell lysates were processed via SDS-PAGE (Mini-PROTEAN TGX precast gels), transferred to a membrane (Millipore Immobilon-FL PVDF), stained (Revert 700 Total Protein Stain) for 3–5 min at room temperature, then washed in Wash solution (P/N 926-11012). After decanting solution, the membrane was rinsed in ultrapure water and imaged on the LICOR Odyssey CL-x system. Immediately after imaging the membrane was incubated with Revert destaining solution (P/N 926-11013) for 5 min then rinsed in ultrapure water before blocking (SuperBlock PBS 37515; Thermo Fisher Scientific) for 1 h at room temperature. The membrane was probed overnight for β-tubulin (T4026, Sigma-Aldrich), deTyr-tubulin (31-1335-00, RevMAb Biosciences), acetylated tubulin (T7451, clone 6-11B-1; Sigma-Aldrich), and gp91phox (Abcam; ab129068). Blots were washed 3 times for 5 min with 1x TBS + 0.1% Tween20. Blots were then incubated with appropriate corresponding secondary antibody (1:5000) for 1 h at room temperature, washed 3x for 5 min with 1x TBS + 0.1% tween 20 and imaged the LICOR Odyssey CL-x system.

Gastrocnemius muscles were collected from 5 mdx and 5 wild-type mice at both 4 and 16 weeks of age, snap-frozen in isopentane cooled on dry ice and stored at −80°C. Tissues were later powdered and homogenized in TRI-reagent (Zymo Research). Phase separation was performed using 0.2 mL of chloroform per 1 mL of TRI-reagent, with samples shaken vigorously for 2min then centrifuged at 12,000 × g for 10 min at 4°C. To precipitate RNA from the aqueous phase, 0.5 mL isopropyl alcohol per 1 mL TRI-reagent used for lysis was added and incubated at room temperature for 10 min before centrifuging for 10 min at the aforementioned settings. The resulting RNA pellet was washed with 75% ethanol, centrifuged for 5 min at 7,500 × g at 4°C, then dissolved in 30 µL DNase-RNase free water at which point RNA concentration was measured using a spectrophotometer. 2.5 µg samples of RNA were reverse transcribed to cDNA using the SuperScript IV First-Strand Synthesis System (Invitrogen), following the manufacturer protocol. cDNA was diluted 1:10 in RNase-free water before use for RT-qPCR. Primers used in this study are listed in Supplementary Table S1 and include primers for 3 housekeeping genes. All reactions were performed using Thermo Fisher Maxima SYBR Green/ROX qPCR Master Mix (2X) on the Applied Biosystems QuantStudio 3 machine with QuantStudio software for cycling and analysis. The DDCt Method was performed against WT values at each age to determine relative gene expression and fold-change.

Fixed FDB fibers were blocked for 2 h at room temperature in Superblock™ Blocking Buffer in PBS (Thermo Scientific) with 0.04% saponin. Fibers were incubated in an Eppendorf tube with primary antibodies to detect microtubule structure (beta-tubulin; T4026, Sigma-Aldrich) and the population of microtubule tubulin modified by detyrosination (deTyr-tubulin; 31-1335-00, clone RM444, RevMAb Biosciences United States, Inc.). To visualize myofibrillar structure, sarcomeric actin was decorated with phalloidin conjugated to Alexa Fluor 633 (A22284, Invitrogen). Primary Antibodies and phalloidin were used overnight at 4°C. The following day FDB fibers were incubated with the appropriate secondary antibodies (diluted in PBS containing 0.04% saponin and 0.1% sodium azide) for 2 h at room temperature, washed three times in PBS, then mounted onto slides with ProLong Gold + Dapi mountant (Invitrogen).

Fixed FDB fibers were imaged on an inverted Nikon C2+ confocal fluorescence system using an automated protocol developed in NIS Elements AR JOBS. Single fibers were identified by their actin labeling (i.e., phalloidin 633) from a full-slide tile scanned image (10x air obj.). Fibers without evidence of bends or hypercontraction under visual inspection (30–50 fibers per slide) were logged as regions-of-interest (ROI). Each identified fiber ROI was imaged using an automated routine that identified the muscle fiber surface and collected a full thickness z-stack (0.5 µm steps; 4 frame average) at 40x (1.4 N.A. Plan Apo air obj.) and 1.3 Airy units which yielded 0.31 mm/pixel resolution.

Statistical comparisons were with GraphPad Prism v9.3.1. Two group comparisons were with t-test and multiple groups comparisons were with ANOVA. The data is presented as Mean ± SEM.

Our lab’s previous work on dysregulated MT mechanotransduction and gross structural alterations was in mdx mice at 3–9 months of age when pathology is well-established and still progressing. In this study, we examined wild-type (C57BL.10/J) and mdx (C57BL.10/J mdx) mice at 4 and 16 weeks of age to elucidate the mechanisms that underlie the development of myofibrillar malformations.

Our initial experiments sought to establish the functional status of the muscle at these ages. Evaluating in vivo plantar flexor function, we confirmed deficits in maximal isometric force in the mdx at both 4 and 16 weeks (Figure 1A). Measuring the weight of the gastrocnemius muscle we identified no differences between genotypes at 4 weeks yet a significant increase in the mass of the mdx gastrocnemius at 16 weeks, a finding consistent with the pseudohypertrophy reported at this age (Figure 1B). Calculating the specific force (i.e., force normalized to mass) revealed no difference between genotypes at 4 weeks yet a significant drop in the specific force of the mdx was observed at 16 weeks (Figure 1C). Finally, evaluating isometric force loss following 20 in vivo eccentric contractions we again found no significant deficit between genotypes at 4 weeks; this however, progressed to a significant decrease in mdx at 16 weeks (Figure 1D). Taken together, we identified an acceleration in functional deficits after 4 weeks of age in the mdx.

Our group, and others (Iyer et al., 2017; Loehr et al., 2018; Nelson et al., 2018; Nelson et al., 2020), have implicated the proliferation MT arrays as negative disease modifiers in adult mdx mice with advanced pathology. Western blot profiling of gastrocnemius muscle from 4 week old mdx vs. WT mice identified a significant increase in tubulin expression (Figure 2B) and its modification by detyrosination (deTyr-tub; Figure 2C), yet we found tubulin acetylation unchanged (acetyl-tub; Figure 2E). Profiling 16 weeks WT muscle we found tubulin expression and levels of deTyr-tub and acetyl-tub unchanged vs. their 4 weeks counterparts. Profiling the 16 weeks mdx we found a similar elevation in tubulin expression (Figure 2B) and level of deTyr-tub as seen at 4 weeks with now a significant elevation in acetyl-tub. Normalizing the level of tubulins modification to its expression we find both deTyr-tub and acetyl-tub increase disproportionately to tubulin expression at 16 weeks (Figures 2D, F).

Alterations in myofibrillar structure are a consequence of deficient myofibrillar repair following acute muscle damage (McHugh and Tyler, 2019) or disease pathology (Head et al., 1997; Lovering et al., 2011; Buttgereit et al., 2013) that predisposes the occurrence of gross malformations in muscle structure (i.e., split and branched fibers) (Head et al., 1997). Our group previously reported a low percentage of grossly malformed (i.e., bifurcated, split, etc.) skeletal muscle fibers in 6–9 weeks mdx, with less than 10% abnormal fibers found in the flexor digitorum brevis (FDB) (Lovering et al., 2009; Goodall et al., 2012). These observations were made by visual inspection in brightfield where we found > 90% of the FDB muscle fibers had no detectable abnormalities. In the current study, we show that when labeled for myofibrillar structure (i.e., phalloidin labeled actin), and imaged with confocal microscopy, a more significant number of muscle fibers with abnormalities in myofibrillar structure becomes apparent (Figure 3).

Using an automated confocal strategy, we imaged single isolated FDB myofibers from WT and mdx at 4 and 16 weeks. Our qualitative visual inspection identified four distinct morphologies of myofibrils within the muscle fiber: 1) canonical aligned striations; 2) evidence of “braided” myofibril structure with misregistration; 3) misalignment characterized by myofibrils wrapping around the peripheral myofibrils; 4) fibers exhibiting gross malformations (i.e., branches, splits); as previously described (Goodall et al., 2012) (Figures 3A–C). In the 4 weeks mdx, myofibers with evidence of braided myofibrils make up 22% of the total FDB population. By 16 weeks, only 38% of myofibers in the mdx FDB exhibit canonical aligned striations; 39% of fibers have braided myofibrils, 14% have myofibrillar wrapping, and 9% of muscle fibers are branched (Figure 3C). Although muscle fibers with altered myofibrillar morphologies were identified in the wild-type, these comprised less than 11% of the total FDB population.

While fibers with canonical aligned striations were in the majority in both genotypes, a significant number of otherwise normal mdx fibers presented with separations between myofibrils, marked by bundles of microtubules (Figure 3C). In fact, these myofibrillar separations were evident in a majority of 16 weeks mdx muscle fibers and were observed in some wild-type FDBs albeit at a markedly reduced occurrence. These initial qualitative observations provided the basis for our adopting quantitative methods. Informed by our past observations of the MT densification often occurring in discrete areas in mdx muscle fibers (Khairallah et al., 2012 [Fig 2]; Kerr et al., 2015 [Fig 1]), and new qualitative evidence that this MT densification is coincident with altered myofibrillar structure, we posited a link between the disease altered MTs and the occurrence of myofibrillar malformations in DMD.

We next quantified the continuity of myofibrillar Z-line striations as a metric of myofibrillar structure. The quantitative assessment of Z-line striations in each image was performed using a custom MATLAB routine established by Morris. (2021) for profiling myofibrillar structure in developing cardiomyocytes and skeletal myotubes and adapted here for mature skeletal muscle fibers.

In brief, multi-channel Nikon confocal fluorescence images were converted into RGB TIFFs, and the Cy5 channel containing phalloidin-633 for actin was output to a new image stack. To decrease processing time, each image was cropped to the fiber of interest and rotated to align the myofiber long-axis to the x-axis of the frame. The maximal myofiber boundary was identified with an Otsu’s threshold (Otsu, 1979) of the maximum intensity projection (max-IP) of the z-stack (Figure 4A). In the resulting binary image, the myofiber area was calculated and the orientation axis was determined using the least mean square orientation estimation algorithm. Subsequently, each z-slice was binarized based on Otsu’s threshold and compared to the binarized maximum intensity projection to determine the ratio of “true” pixels in the z-slice of interest compared to the max-IP (Figures 4B–D). Z-slice ratios above a user-defined threshold (nthresh = 0.6) were selected for analysis (Figure 4E).

For each analyzed z-stack image, the minor axis length (i.e., width) of the myofiber was determined for every 20 pixels along the long axis of the muscle fiber (Figure 5A). Subsequently, the phalloidin labeled Z-line structure was detected via the ZlineDetection algorithm developed by Morris et al. (Morris et al., 2020). The continuity of each Z-line was determined by its length divided by the nearest minor axis length and plotted with color code (Figure 5B) with a continuous Z-line spanning the muscle fiber perpendicular axis yielding a measure of 1 (red), with interruptions in the continuity yielding lower values and cooler colors. The output for each fiber includes a mean striation continuity score for each z-slice as well as a boxplot for the entire z-stack (Figure 5C). The median striation length for the entire muscle fiber z-stack is also reported as the continuity score.

We show that striation continuity is a measure that effectively identifies the minor interruptions in the myofibrillar structure seen in WT muscle fibers as well as the more significant disruptions in the mdx (Figures 6A–F). We demonstrate continuity scores < 0.4 in fibers displaying areas of myofibrillar braiding (Figures 6B, D), while scores decrease to < 0.2 in fibers with significant myofibrillar wrapping (Figures 6C, F).

Using the automated imaging strategy and ZlineDetection to quantitate Z-line continuity, we screened FDB myofibers from WT and mdx mice at 4 and 16 weeks of age. At 4 weeks we find no significant difference in the mean striation continuity of the entire fiber between mdx and WT (Figure 6G), the 4-week mdx do achieve lower minimum continuity scores than the 4-week wild-type. Because severely misaligned myofibrils are a rare event at 4 weeks, the reduced continuity score is manifest from microtubule bundles between myofibrils causing separations as previously described (Figure 3A).

By 16 weeks, the differences between morphology in mdx and wild-type become more pronounced. While we found an increase in the average continuity score in wild-type, the continuity of the mdx decreased as altered morphologies emerge at a greater frequency (Figure 6G). This increase in separations between myofibrils in the mdx between 4 and 16 weeks suggests a progression of myofibrillar alterations with disease progression.

Examining FDB myofibers labeled for deTyr-tubulin and actin we find a significant increase in the density of deTyr-MTs in the mdx at 4 weeks that progresses at 16 weeks (Figures 7A, B). Revisiting our previous observation of bundled microtubules coinciding with myofibrillar separations, we quantified the density of deTyr-tubulin within the regions of myofibrillar separation versus in areas with otherwise normal myofibrillar connectivity (Figures 7C–E). At 4 weeks we find no difference in the density of deTyr -MTs between these areas in WT, but in the mdx, we find a significant increase in deTyr-MTs only in the regions of myofibrillar separation.

In 16 weeks WT muscle fibers we again find no difference in the density of deTyr -MTs within the regions of myofibrillar separation versus areas with otherwise normal myofibrillar connectivity. In the 16 weeks mdx we again find a significant elevation in deTyr-MTs in areas of myofibrillar separation but now find these changes in areas of otherwise normal myofibrillar connectivity as well. (Figure 7E). Taken together, these results suggest that deTyr-MTs become abundant first between myofibrils then progress more globally throughout the myofibrillar structure as disease progresses.

Detyrosination is the reversible enzymatic cleavage of the COOH-terminal tyrosine from α-tubulin by vasohibin 1 (VASH1) or vasohibin 2 (VASH2) and their partner the small vasohibin binding protein (SVBP) (Aillaud et al., 2017; Ramirez-Rios et al., 2022). Examining the transcripts of these proteins at 4 weeks (Figures 8A–D) finds no significant change in VASH1 yet a significant increase in VASH2 and SVBP in the mdx. At 16 weeks the VASH1 remained unchanged and VASH2 remained elevated in the mdx. In contrast was SVBP that was not different between genotypes at 16 weeks (Figures 8F–H).

Microtubules are essential for myofibrillar growth, maintenance, and repair (Pizon et al., 2005; Scholz et al., 2008; Denes et al., 2021; Dhanyasi et al., 2021). We identified bundles of deTyr-modified MTs associated with myofibrillar separations in 4 weeks mdx muscle fibers that then progress to more significant changes at 16 weeks. Given this association we sought to determine if an experimental increase in deTyr-MTs in 4 weeks WT fibers was sufficient to recapitulate the changes seen in the 16 weeks mdx. To this end we used electroporation to introduce our bicistronic cDNA construct of VASH2-GFP + SVBP (Aillaud et al., 2017), or sfGFP cDNA as a control, into the 4 weeks old mouse FDB and examined the muscle fiber properties 5–7 days later (Figures 9A–C).

In WT FDB fibers transduced with VASH2-GFP + SVBP, we find sfGFP localization in cytosol at both the m-line and with MT arrays (Figure 10D). We also find VASH2-GFP localized within the nucleus (Figure 9B; Figure 10D), a finding consistent with detyrosination as a regulator of mitotic spindle function (Barisic et al., 2015; Liao et al., 2019). Importantly, in WT FDB fibers transduced with VASH2-GFP + SVBP we find an increased abundance of deTyr-enriched MT arrays when compared to the controls expressing sfGFP controls (Figures 9A–C). In mdx muscle fibers we previously linked the elevated levels of deTyr-MTs to an increase in cytoskeletal mechanics (i.e., stiffness) (Kerr et al., 2015; Coleman et al., 2021). Using nanoindentation to measure the viscoelastic properties we show a significant increase in passive stiffness in VASH2-GFP + SVBP over expressing FDB muscle fibers compared to the sfGFP control (Figure 9D).

Our evidence suggested a link between the increased abundance of deTyr-enriched MT arrays and the altered myofibrillar structure in the 16 weeks mdx. We observed an increase incidence of sarcomere disruption in WT fibers overexpressing VASH2-GFP + SVBP compared to sfGFP expressing controls (Figures 10A, B). Consistent with this finding was a significant reduction in striation continuity in WT fibers overexpressing VASH2-GFP + SVBP determined by Z-line detection (Figure 10C). Concordant with our finding deTyr-enriched MTs co-registered with myofibrillar malformations in the mdx (Figure 7), our automated NIS-Elements analysis (Figure 10D) showed that VASH2-GFP + SVBP over-expression in WT myofibers yielded deTyr-tub enriched MT arrays only in areas with altered myofibrillar structure (Figure 10E). Further evidence supporting a link between deTyr-MTs and myofibrillar malformations came from visual inspection that revealed braided myofibrils in 21.3% of the VASH2-GFP + SVBP over-expressing myofibers (Figure 10F) with no evidence for these malformations in the GFP- controls. Given that VASH2-GFP + SVBP overexpression in WT muscle was sufficient to model the altered myofibrillar structure that arises in the 16 weeks mdx, we posit that the increased abundance of deTyr-enriched MT arrays is an early event in dystrophic pathology that predisposes the altered myofibrillar structure in dystrophinopathies.