Tropomyosin 4 Defines Novel Filaments inSkeletal Muscle Associated With MuscleRemodelling/Regeneration in Normal and

Diseased Muscle

Nicole Vlahovich,1,2 Galina Schevzov,3,4 Visalini Nair-Shaliker,1 Biljana Ilkovski,5

Stanley T. Artap,1 Josephine E. Joya,1 Anthony J. Kee,1,6 Kathryn N. North,4,5

Peter W. Gunning,3,4 and Edna C. Hardeman1,6*

1Muscle Development Unit, Children’s Medical Research Institute, Locked Bag23, Wentworthville, New South Wales 2145

2School of Natural Sciences, University of Western Sydney, Locked Bag 1797,Penrith South, New South Wales 1797

3Oncology Research Unit, The Children’s Hospital at Westmead,Locked Bag 4001, Westmead, New South Wales 2145

4Discipline of Paediatrics and Child Health, Faculty of Medicine, University ofSydney, New South Wales 2600

5The Institute for Neuromuscular Research, The Children’s Hospital at Westmead,Parramatta, New South Wales 2145

6Faculty of Medicine, University of Sydney, Sydney, New South Wales 2600,Australia

The organisation of structural proteins in muscle into highly ordered sarcomeresoccurs during development, regeneration and focal repair of skeletal muscle fibers.The involvement of cytoskeletal proteins in this process has been documented,with nonmuscle g-actin found to play a role in sarcomere assembly during muscledifferentiation and also shown to be up-regulated in dystrophic muscles whichundergo regeneration and repair [Lloyd et al., 2004; Hanft et al., 2006]. Here, weshow that a cytoskeletal tropomyosin (Tm), Tm4, defines actin filaments in twonovel compartments in muscle fibers: a Z-line associated cytoskeleton (Z-LAC),similar to a structure we have reported previously [Kee et al., 2004], and longitudi-nal filaments that are orientated parallel to the sarcomeric apparatus, present dur-ing myofiber growth and repair/regeneration. Tm4 is upregulated in paradigmsof muscle repair including induced regeneration and focal repair and in musclediseases with repair/regeneration features, muscular dystrophy and nemalinemyopathy. Longitudinal Tm4-defined filaments also are present in diseased mus-cle. Transition of the Tm4-defined filaments from a longitudinal to a Z-LAC orien-tation is observed during the course of muscle regeneration. This Tm4-definedcytoskeleton is a marker of growth and repair/regeneration in response to injury,disease state and stress in skeletal muscle. Cell Motil. Cytoskeleton 65: 73–85, 2008.' 2007Wiley-Liss, Inc.

Key words: tropomyosin; muscle; actin; muscular dystrophy; nemaline myopathy

Received 20 June 2007; Accepted 18 September 2007

Published online 29 October 2007 in Wiley InterScience (www.

interscience.wiley.com).

DOI: 10.1002/cm.20245

*Correspondence to: Edna C. Hardeman, Muscle Development Unit,

Children’s Medical Research Institute, Locked Bag 23, Wentworth-

ville, NSW 2145, Australia. E-mail: ehardeman@cmri.usyd.edu.au

Contract grant sponsors: Australian National Health and Medical

Research Council.

' 2007 Wiley-Liss, Inc.

Cell Motility and the Cytoskeleton 65: 73–85 (2008)

INTRODUCTION

The muscle and cytoskeletal filment systems ofskeletal muscle are essential for muscle integrity andfunction. The ordered arrays of contractile proteinsknown as the sarcomeres must be correctly aligned andassembled for skeletal muscle to function optimally. Amajor component of the sarcomere is the filamentousprotein actin, which forms the basis of the thin filament.In addition to the sarcomeric actin filaments, musclefibers contain cytoskeletal actin filaments, most notablythose present in the costameres. g-actin is a major com-ponent of the costameres, structures that lie beneath thesarcolemma providing a link between the plasma mem-brane and the sarcomeric apparatus. They connect themyofibrils to the sarcolemma playing a role in the main-tenance of sarcomeric structure and membrane stability[Rybakova et al., 2000; Ervasti, 2003]. g-actin filamentsalso are present in the interior of the fibers adjacent tothe Z-lines [Nakata et al., 2001]. In addition to the rolesthat g-actin filaments play in mature muscle fibers, theyalso participate in sarcomere assembly during fiber dif-ferentiation, existing as longitudinal filaments in newlyformed fibers that sequester the Z-line protein a-actinin[Lloyd et al., 2004].

It is known that the actin-associated protein tropo-myosin (Tm) defines the function of the actin cytoskele-ton [Gunning et al., 2005]. While there are only 6 actinisoforms, the Tm family of actin-associated proteins con-sists of more than 40 isoforms generated by alternateRNA splicing from four mammalian genes. There arethree striated muscle-specific Tms present in skeletalmuscle that associate with the sarcomeric actins in thethin filament of the sarcomere [reviewed in Clark et al.2002]. Therefore, the great majority of the Tm isoformsare classified as cytoskeletal Tms and associate withactin filaments comprised of the cyoskeletal actins, b-and g-actin. Recently, we reported the presence of cytos-keletal Tm isoforms in skeletal muscle fibers. There areat least two distinct cytoskeletal Tm/g-actin filament sys-tems as defined by the co-localization of a cytoskeletalTm from the g gene, Tm5NM1, and g-actin. One struc-ture is associated with the sarcolemma and the otherwith the Z-line called the Z-line associated cytoskeleton(Z-LAC) [Kee et al., 2004]. We found that disrupting theZ-LAC in transgenic mice elicited a mild dystrophicphenotype. This indicates the importance of cytoskeletalTm-defined structures in muscle fibers and the need tofurther define and characterise the Tms associated withthese structures.

We previously reported that another cytoskeletalTm, Tm4, is present at significant levels in adult mousemuscles [Gunning et al., 1990]. The current study dem-onstrates that in normal muscle, Tm4, an isoform from

the d-Tm gene, is present in the Z-LAC; however, inregenerating muscle and muscle from diseases character-ized by repair and regeneration, it is present in a novelcytoskeletal structure orientated perpendicular to the Z-line (longitudinal filaments). Examination of normalhuman muscle from different stages of development andmouse muscle undergoing regeneration and stretch-induced repair revealed that elevated Tm4 protein andthe presence of Tm4 in longitudinal filaments in musclefibers reflects active growth, regeneration and repairprocesses. Tm4 protein levels are elevated in and aswell, longitudinal filaments are present in human myo-pathic muscles and in muscles from a nemaline mousemodel. Notable is the sensitivity provided by Tm4 as ameans of detecting repair in nemaline muscles in whichfocal repair has only recently been identified [Sanoudouet al., 2006]. Identification of Tm4 as a marker of regen-erative processes in skeletal muscle may assist in the rec-ognition of muscle repair in muscle biopsy samples. Itspresence in longitudinal filaments in forming fibers andfibres undergoing repair indicate that it is the cytoskele-tal Tm isoform associated with g-actin filaments that areinvolved in sarcomere assembly.

MATERIALS AND METHODS

Antibodies

The following primary antibodies were used: Tm4specific antibody, WD4/9d (rabbit polyclonal antibody),g-actin (sheep polyclonal) both previously described bySchevzov et al. [2005], a-actinin (mouse monoclonal,clone EA-53, Sigma, St Louis), sarcomeric Tm (mousemonoclonal, clone 311, Sigma). The following second-ary antibodies were used: 488-conjugated goat anti-rab-bit, 594-conjugated goat anti-mouse and 594-conjugateddonkey anti-sheep (Alexa Fluor, Invitrogen, Carlsbad,CA), goat anti-rabbit and goat anti-mouse HRP-conju-gated antibodies (BioRad, Hercules, CA).

Muscle Samples

Human. Human muscle tissue was taken by biopsyor at autopsy and stored in liquid nitrogen. Samples werede-identified in accordance with The Children’s Hospitalat Westmead (CHW) ethical guidelines, and the use ofthe samples was approved by the CHW Ethics Commit-tee. Normal control biopsies: 19, 22, 25, and 36 weeksgestation, 8 days, 8 weeks, 5 years and 38 years of age(no muscle disease evident by pathology). Nemaline my-opathy samples: Patient NM1 [autopsy sample, 55 year-old with TPM3 (Met9Arg) mutation; Laing et al., 1995],Patient NM2 [biopsy sample, 22 year-old with ACTA1(V163M) mutation; Hutchinson et al., 2006]. Duchennemuscular dystrophy biopsies: Patients DMD1 (5 year-old), DMD2 (9 year-old). Becker muscular dystrophy

74 Vlahovich et al.

biopsies: Patients BMD1 (11 year-old), BMD2 (12 year-old).

Mice. FVB/N mice, TPM3(Met9Arg) nemalinemice [Corbett et al., 2001], and mdx mice [Tanabe et al.,1986] 2–4 months of age were used for all animal experi-ments. All studies using mice were performed in accord-ance with the ethical guidelines of the Animal WelfareCommittee, National Health and Medical ResearchCouncil, Australia.

Western Blot Analysis

Two methods were used to extract protein frommouse tissue. To enrich for Tms, tissue was homogenizedin 20 volumes of 50 mM Tris, pH 7.4, containing prote-ase inhibitors (Complete EDTA Free, Roche, Indianapo-lis), by sonication. This solution was boiled for 10 minand spun at 12,000 rpm for 5 min to precipitate andremove all proteins other than the heat stable and solubleTms. The second method was performed essentially asper Kee et al. [2004]. Briefly, muscle was homogenizedin 10 mM Tris, pH 7.6, 2%SDS, 20 mM 1,4-dithiothreitol(DTT) and western blotted as previously reported [Keeet al., 2004]. Densitometry was performed using a GS800calibrated densitometer (BioRad). Human biopsy sampleswere treated similarly with the exception that homogeni-zation was performed on frozen muscle cryo-sections (10sections, 8-l thick, each measuring approx. 10 mm2).

Immunohistochemistry

Human biopsy samples: Human muscle biopsysamples were sectioned in a cryostat (HM500OM; CarlZeiss MicroImaging, Oberkochen, Germany) at 7 lm,fixed in cold 2% parformaldehyde (PFA) for 2 min, andwashed in PBS.

Semi-thin sectioning: Immediately after dissection,mouse muscles were fixed in 4% PFA (hind- and fore-limb muscles were stretched and held during fixation)for 20–30 min and processed for cryoultramicrotomyaccording to a modification of Griffiths et al. [1984]. Af-ter fixation, muscles were cut into approximately 4–5-mm long 3 1-mm wide strips and transferred to 1.8 Msucrose/20% polyvinylpyrrolidone for overnight infusionat 48C. Muscle strips were trimmed further, mounted oncryopins, and snap frozen in liquid nitrogen. Semi-thin(0.5–0.8 lm) sections were cut at 2608C using an Ultra-cut UCT ultramicrotome (Leica, Wetzler, Germany)equipped with an EM FCS cryochamber (Leica). Sec-tions were suspended in 2.3 M sucrose, allowed to thaw,and placed on Starfrost poly-o-lysine coated slides(Knittle, Germany).

Both standard and semi-thin sections were blockedovernight (48C) in PBS containing 0.05% Triton X-100(PBST), 0.2% fish gelatine (Sigma), 2% BSA. Primaryantibodies were applied for 60 min at 378C or overnight

at 48C and slides were then washed with PBST. Second-ary antibody was applied for 60 min at RT, slides wereincubated with 40,6-Diamidino-2-phenylindole dihydro-chloride (DAPI) (Sigma) at a concentration of 1 lg/mLfor 5 min at RT to visualize nuclei and then washed withPBST, mounted with Vectashield (Vector Laboratories,Burlingame, CA) and viewed using either a confocallaser scanning microscope (oil immersion 633 objec-tive, model TCS SP2; Leica) and analyzed with confocalsoftware (Leica) or standard fluorescent microscopyusing an Olympus BX51 fluorescent microscope.

Notexin Induced Muscle Regeneration

Ketamine (Ketavet 100, Delvet, Asquith, NSW,Australia) and Xylazine (Ilium Zylazil-20, Troy Labora-tories, Smithfield, NSW, Australia) (100 mg and 10 mg/kg body weight, respectively) were used to anesthetizethe mice, hindlimbs shaved and incisions made to exposethe soleus or EDL muscles. 20 lL Notexin (Latoxan,France) at a concentration of 10 lg/mL in PBS wasinjected directly into each muscle [Plant et al., 2006].Skin was sutured and mice left to recover for a period of3, 5, 7, or 10 days with three mice used per timepoint.Mice were sacrificed and the muscles collected for west-ern or immunohistochemical analysis. Uninjected limbwas used as control tissue. Muscle was found to be com-pletely broken down 3 days post-Notexin injection, withno intact muscle fibers evident and an abundance ofmononucleated cells. After 5 days of recovery mono-nucleated cells were still evident; however, regenerationhad begun with many fibers found containing centralizednuclei. At 7 and 10 days post-injection mononucleatedcells were reduced and increased interfibrillar spaces evi-dent in comparison to uninjected muscle. Regenerationis almost complete at the 10 day time-point.

Mouse Hindlimb Immobilization

The hindlimbs of three adult FVB/N mice were im-mobilized for 5 days as described by Joya et al. [2004]such that the soleus muscle is lengthened and the EDL isshortened. Following 5 days of immobilization, soleusand EDL muscles were collected for western blottingand immunohistochemistry.

RESULTS

Cytoskeletal Tm4 Defines Two CytoskeletalStructures in Skeletal Muscle

We established that an actin cytoskeleton definedby the cytoskeletal Tm, Tm5NM1, exists in muscle fibersand that disruption of this structure results in a dystrophicphenotype [Kee et al., 2004]. To examine if other Tm iso-forms from the d gene are expressed in skeletal muscle, a

Tm4 in Muscle Disease and Repair 75

range of murine and human muscle samples were exam-ined by western blot using an antibody specific to exon9d. Tm4 is expressed in all muscles examined, but atvarying levels (Fig. 1A). In the mouse, Tm4 is highlyexpressed in the soleus (Sol), flexor digitorum brevis(FDB), diaphragm (Dia), intercostal (IC) and extraocular(EOM) muscles. The longissimus dorsi (LD) in the backexpresses a relatively lower level of this Tm along withhindlimb muscles such as the extensor digitorum longus(EDL) and tibialis anterior (TA). A higher molecularweight 38 kDa protein, recognized by the Tm4-specificantibody and abundant in mouse embryonic fibroblasts(MEFs), also is present in some muscles. A developmen-tal timecourse of human muscle was examined and thelevel of Tm4 is significantly higher in fetal and infantmuscle compared with adult muscle (Fig. 1B). This corre-lates with the parallel decrease in levels of Tm4 mRNAduring muscle development [Gunning et al., 1990].Taken together, these results reveal that Tm4 is differen-tially expressed between mouse muscles and during mus-cle maturation in humans.

Cytoskeletal Tms define distinct microfilamentpopulations in nonmuscle cells [Gunning et al., 1998]and associate with distinct membrane systems [Dalby-Payne et al., 2003; Percival et al., 2004]. In order todetermine if Tm localized to specific sites within musclefibres, we examined the distribution of Tm4 in sectionsof murine muscle (Fig. 2]. Two distinct structures con-taining Tm4 are evident in contrast to Tm5NM1 whichis evident in only one structure, the Z-LAC [Kee et al.,2004]. Tm4 is present predominantly in a Z-LAC orien-tation in the soleus (Figs. 2A–2C), ECU, EDL and LD

muscles (data not shown). In the diaphragm, Tm4 ispresent both in Z-LAC and longitudinal filaments thatare orientated perpendicular to the Z-line (Figs. 2D–2F).EOM (Figs. 2G–2I) and laryngeal muscles (data notshown) presented a mixture of the two patterns withsome fibers containing predominantly Z-LAC orientedstructures and others longitudinal structures (Figs. 2G–2I). A common feature of muscles that display Tm4-defined longitudinal structures is the presence of lowlevel chronic repair, such as in the EOM which has con-tinuous myofiber remodelling as evidenced by the up-regulation of activated satellite cell markers and devel-opmental myosin heavy chain isoforms [Lucas and Hoh,2003;McLoon et al., 2004].

Tms exist in association with actin filaments. Wehave shown that Tm5NM1 co-localizes with g-actin atthe Z-LAC region in muscle [Kee et al., 2004]. Wedetermined whether Tm4 also associates with a g-actinor alternatively with a b-actin-based filament system.Immuno-labelling of the soleus (Figs. 3A–3C) and dia-phragm (Figs. 3D–3F) muscles shows that Tm4 associ-ates with g-actin filaments. In the EOM, co-localizationof Tm4 with g-actin was found both in the Z-LAC regionand in longitudinal filaments (Figs. 3D–3F). We couldnot detect b-actin in this region (data not shown).

Longitudinal Structures Defined by Tm4 areEvident During Myofibrillar Assembly andRemodelling

To further explore whether Tm4-defined longitudi-nal filaments reflect repair in muscle, Tm4 localizationwas assessed in paradigms for muscle regeneration and

Fig. 1. Tm4 is expressed in mouse and human skeletal muscles.

Western blots of both mouse muscle (A) and human muscle samples

(B) show Tm4 (30 kDa) is present at varying levels. For mouse

muscles, 10 lg of total protein solubilized using a Tm enrichment pro-

tocol was loaded per well with the exception of the EOM where 5lgwas loaded. Human muscle samples from 19, 22, 25, and 36 weeks of

gestation and 8 days, 8 weeks, and 38 years post birth were analyzed

with approximately 10 lg of protein (10 3 10 mm2 sections) loaded

per well—8 week and adult blots were exposed for longer in order to

detect the adult signal. Coomassie staining of actin shows the relative

amounts of protein loaded per lane in mouse samples while sarco-

meric Tms were detected by the CH1 antibody in human samples

(Sarc Tm). Abbreviations: soleus (Sol), extensor digitorum longus

(EDL), tibialis anterior (TA), extensor carpi ulnaris (ECU), flexor dig-

itorum profundus (FDP), flexor digitorum brevis (FDB), diaphragm

(DIA), latissimus dorsi (LD), intercostal (IC), laryngeal (Lar), extra-

occular muscle (EOM), whole brain (Brain), mouse embryonic fibro-

blasts (MEFs).

76 Vlahovich et al.

stretch. Injection of the myotoxic agent notexin inducesmyofiber damage results in rapid degeneration followedby regeneration [Mendler et al., 1998]. Degeneration/regeneration was induced in the soleus and EDL musclesand Tm4 structures were monitored as the muscle fibersformed and matured. Tm4 levels increased at least three-fold during the time course of regeneration (Fig. 4A).Tm4 is present in longitudinal, but not Z-LAC structures5 days after muscle injury (Figs. 5A–5C). These longitu-dinal filaments persist at later stages of regeneration (7and 10 days post notexin injection; Figs. 5D and 5G,arrows); however, Z-LAC structures are also evident(arrowheads). This confirms that Tm4 is associated withsarcomere formation and alignment.

To examine the effects of altered stress on musclefibers resulting in sarcomere remodelling, we utilized astretch immobilization protocol. The hindlimbs of micewere immobilized such that the soleus and surroundingmuscles were held in a stretched position while the EDLremained relaxed. Tm4 levels are induced upon stretch(Fig. 4B). The localization of Tm4 in the stretched soleusmuscle (Figs. 6D–6F) differs from unstretched muscle(Figs. 6A–6C) in that the tight association of Tm4 withZ-LAC is lost and some Tm4 longitudinal structures areevident (Fig. 6D, arrow). Tm4-defined longitudinalstructures are not evident in control, unstretched muscle.The increase in Tm4 and altered localization in filamen-tous structures during regeneration and in stretch is con-

Fig. 2. Tm4 localization in Z-LAC and longitudinal structures dif-

fers in mouse muscles. Fluorescent antibody detection of Tm4

(A,D,G) and a-actinin (Z-lines) (B,E,H) in semi-thin (0.8 lm) sec-

tions reveal two distinct types of Tm4 structures (merged images

(C,F,I). Muscles of the hindlimb displayed predominantly Z-LAC

localization of Tm4 as shown in the soleus (A–C), which was fixed in

a stretched position. In the diaphragm (D–F), longitudinal structures

could be seen (F, arrows) and faint Z-LAC staining (F, arrowheads).

The EOM is a hybrid muscle with some fibers containing predomi-

nantly Tm4-defined longitudinal structures (I, star) and others showing

predominantly Z-LAC orientation (I, arrowhead). Scale bar5 10 lm.

Tm4 in Muscle Disease and Repair 77

sistent with a role in building, repairing and remodellingof myofibrils.

Longitudinal Tm Filaments is an Indicator ofSkeletal Muscle Disease

Regeneration and repair are hallmarks of musclediseases such as muscular dystrophy and to a lesserextent, nemaline myopathy. To determine if the localiza-tion of Tm4-defined filaments reflects these processes inmuscle disease as it does during muscle injury, we firstexamined Tm4 localization in the muscles of mouse mod-els for these diseases. Mouse models of muscular dystro-phy [mdx; Tanabe et al., 1986] and nemaline myopathy[TPM3(Met9Arg); Corbett et al., 2001] were analyzed forthe presence of longitudinal filaments in affected muscles.Longitudinal filaments are present in the muscle fibers ofthe mdx diaphragm, one of the most severely affectedmuscles (Figs. 7A and 7C, arrowhead). They are particu-larly prominent in fibres containing centrally locatednuclei and developing sarcomeres, indicated by small,closely spaced Z-lines. Affected nemaline fibers are char-acterized by the presence of a-actinin-rich structurescalled ‘‘rods’’ (Figs. 7A–7C, arrow). Tm4-defined longi-tudinal filaments are evident in the affected fibres(Figs. 7A and 7C, arrowhead), but are absent from the rods.

We confirmed these findings in human musclesamples from patients with myopathies in which repairand regeneration are characteristic features [Watchkoet al., 2002]. Biopsy and autopsy samples from patientswith Duchenne and Becker muscular dystrophy andTPM3 and ACTA1 nemaline myopathy were analyzed

Fig. 4. Tm4 protein levels are elevated in regenerating and stretched

muscle. Time courses of soleus muscles undergoing regeneration in

response to notexin injection (A) and stretch immobilization (B) wereanalyzed by western blot. Two muscles were pooled per timepoint in

(A) and three muscles for (B) and 20 lg loaded per well. Representa-

tive western blots are shown. An increase in the 30 kDa Tm4 isoform

is evident in regenerating muscles (R) at all timepoints in comparison

with the level in nonmanipulated (N) soleus muscles. An increase is

observed in soleus muscle that has undergone immobilized stretch for

5 days (I) compared with unimmobilized control muscle (U). Equal

amounts of proteins were loaded per lane as confirmed by Coomassie

staining of gels (data not shown).

Fig. 3. Gamma actin co-localizes with Tm4 in both Z-LAC and longitudinal structures. Tm4 (A,D) and

g-actin (B,E) were detected in sections of mouse soleus (A–C) and diaphragm (D–F). In the soleus the Z-

LAC orientated Tm4 structures contain g-actin as seen by the presence of yellow in the merged images

(C). Longitudinal structures are evident in the diaphragm, stained with both Tm4 (D, arrow) and g-actin(E, arrow) while faint Z-LAC staining is present (D and E, arrowheads). Scale bar 5 5 lm.

78 Vlahovich et al.

for the levels of Tm4 compared to normal control mus-cle. These samples have varying amounts of muscle ver-sus nonmuscle tissue infiltrate depending on the severityof the disease pathology. We determined the amount ofmuscle tissue present in each sample by measuring thesarcomeric Tm level as representative of contractile pro-tein present. The amount of Tm4 in the muscle in eachsample is expressed normalized to the amount of sarco-meric Tm present. Tm4 levels in dystrophic samples(from 5 to 12 year-old children) were elevated over the 5year-old control sample (Fig. 8). The nemaline musclesfrom adults displayed higher levels of Tm4 relative tocontractile protein (sarcomeric Tm) than adult control

muscles; although, less than that of the 5 year-old controlsample (Fig. 8).

Dystrophic and nemaline samples were examinedfor the presence of Tm4-containing structures by immu-nohistochemistry. Normal human muscle (Figs. 9A–9C)displays the Z-LAC localization. The novel Tm4-con-taining longitudinal structures are prominent in dystro-phic muscle (Figs. 9D–9F, arrowheads). The lack of Z-lines in Tm4 filament-rich regions of dystrophic fiberslinks these structures to sites of immature myofibrils.Longitudinal filaments are present, but less prominent innemaline-affected muscle (Figs. 9G–9H, arrowheads).The relative abundance of Tm4 longitudinal structures in

Fig. 5. Localization of Tm4 in structures changes during muscle

regeneration. A time course of soleus muscle regeneration in response

to notexin treatment was analyzed by immunohistochemistry using

antibodies to Tm4 (A,D,G) and a-actinin (B,E,H). DAPI was used to

visualize central nuclei shown in merged images (C,F,I). At 5 days

post injection, Tm4 is present in longitudinal structures (A,C arrow)

perpendicular to the a-actinin delineated Z-lines (B,C). At both 7 and

10 days post notexin treatment, Tm4 is present in both longitudinal

(D,G arrows) and Z-LAC structures (arrowheads). The merged images

show that longitudinal structures appear in areas of Z-line misalign-

ment (F,I arrows). Scale bar5 5 lm.

Tm4 in Muscle Disease and Repair 79

Figure 6.

Figure 7.

80 Vlahovich et al.

nemaline vs. dystrophic muscle most likely reflects thelevel of repair/regeneration occurring in these diseases.

DISCUSSION

Tm4 Defines Two Distinct Populations ofFilaments in Skeletal Muscle

Studies have shown that cytoskeletal Tms definefunctionally distinct actin filaments in a range of celltypes including neurons [Gunning et al., 1998; Bryceet al., 2003], fibroblasts [Percival et al., 2000], osteo-clasts [McMichael et al., 2006], epithelial cells [Dalby-Payne et al., 2003] and skeletal muscle [Kee et al.,2004]. The current study shows that skeletal muscle con-tains a cytoskeletal Tm isoform, Tm4, which definesactin filaments in association with the Z-line in maturemuscles and actin filaments that are present longitudi-nally along the muscle fiber in muscle undergoing regen-eration/repair/remodelling.

Tm4/Actin Cytoskeleton Plays a Role inthe Repair of Skeletal Muscle Fibers

The pre-myofibril model of myofibrillogenesis pro-poses that stress fibers in myobasts are transformed intopre- and then nascent myofibrils in the developing myo-tube which in turn form the mature myofibrils of theskeletal muscle [Sanger et al., 2005]. We propose thatTm4 is included in early stress fiber-like structures dur-ing muscle development and decorates actin filamentsduring the process of sarcomeric development and align-ment. Our data is not sufficiently sensitive; however, todetect Tm4 in nascent myofibrils. Nonmuscle g-actin hasbeen implicated in the development of the contractile ap-paratus during the process of muscle differentiation,since it is present in stress fiber-like filamentous struc-tures in C2C12 cultures [Lloyd et al., 2004]; although, itis not required for maintenance of the sarcomere [Sonne-mann et al., 2006]. In the present study, Tm4 was found

to be present in regenerating myofibers and diseasedmuscle in longitudinal structures similar to those definedby g-actin in C2C12 myotubes. We have also observedTm4 in stress fiber-like structures in newly formedC2C12 myotubes (Schevzov, unpublished). These longi-tudinal structures become less prominent as regeneratingmyofibers mature, disappearing in some muscles, con-comitant with an increase in prominence of the Tm4-defined Z-LAC structures. Tm4 therefore provides amarker for stress fiber-like structures in newly formedmyotubes and may aid in visualizing myofibril assembly.

It has been proposed that g-actin contributes toremodelling of the costameric cytoskeleton in the ab-sence of dystrophin [Hanft et al., 2006]. Western blottingof skeletal muscle from mdx mice showed a significantincrease in g-actin in the dystrophic muscle, observed byimmunohistochemistry to be located at both the periph-ery and within the cytoplasm of muscle fibers [Hanftet al., 2006].The increase of both Tm4 and g-actin indystrophic muscle and the co-localization of Tm4 and g-actin in muscles undergoing chronic repair, indicatesthat g-actin filaments decorated by this Tm isoform mayplay a role in this compensatory remodelling process.The Tm4-defined longitudinal structures may provide ascaffold for the assembly of the contractile proteins ashas been suggested for the g-actin based filaments[Lloyd et al., 2004]. It is also possible that these struc-tures serve a role in shuttling sarcomeric proteins to andfrom sites of repair/remodelling.

The Presence of Tm4-Defined LongitudinalFilaments is Diagnostic of Skeletal MuscleRegeneration and Repair

Formation of the contractile apparatus occurs inmyofibers during development, in regeneration follow-ing muscle injury or disease, and during focal repair andstretch. We have observed novel longitudinal structuresdefined by Tm4 in muscle fibers undergoing sarcomericregeneration/repair/remodelling. These structures are

Fig. 6. Stretching hindlimb muscles induces the formation of Tm4-defined longitudinal structures. Sol-

eus muscles immobilized for 5 days in a dorsi-flexion position were analyzed by immunohistochemistry

using antibodies to Tm4 (A,D) and a-actinin (B,E). Tm4 immunolabelling of unimmobilized control

muscles reveals the typical Z-LAC localization (A–C). Tm4 staining of stretched, immobilized soleus

muscle (D–F) revealed longitudinal structures (D, arrow) as well as a general reduction in the definition

of Z-LAC immunolabelling. Scale bar 5 10 lm.

Fig. 7. Tm4 defines longitudinal filaments in mouse models of muscular dystrophy and nemaline myo-

pathy.The diaphragm of mdx mice (A–C) and EDL muscle of TPM3(Met9Arg) mice (D–F) were ana-

lyzed for the localization of Tm4 (A,D) relative to a-actinin (B,E). Images merged to show co-localiza-

tion (C,F). Both dystrophic and nemaline muscle displayed Tm4-defined longitudinal filaments (arrow-

heads), which were more prominent in dystrophic muscle (D,F). a-Actinin-rich nemaline rods did not

contain Tm4 (D-F arrows).

Tm4 in Muscle Disease and Repair 81

particularly prominent in muscles undergoing extensiveregeneration such as in muscular dystrophies and arealso seen to a lesser degree in nemaline myopathy, a dis-ease with focal regeneration [Sanoudou et al., 2003].Longitudinal structures are also present in stretch-im-mobilized muscle, a condition where there is an increasein the synthesis of sarcomeric proteins leading to remod-elling of myofibrils [reviewed in Goldspink, 1999]. Weconclude that Tm4-defined longitudinal structures are afeature of sarcomeric remodelling as well as repair andregeneration due to the presence of these structures inthe various systems exhibiting these events.

Myopathies and injuries can cause focal repair aswell as degeneration and regeneration in muscle[Watchko et al., 2002; Jarvinen et al., 2005]. Recently,indicators of satellite cell activation and repair were alsofound in the human muscle disease nemaline myopathy.Microarray analysis of both human and mouse musclesamples from nemaline patients or TPM3 (Met9Arg)mice versus control muscle indicated an up-regulation ofsatellite cell markers suggesting repair is taking place in

these muscles [Sanoudou et al., 2003, 2006]. Here weshow that Tm4 is up-regulated at the protein level inboth muscular dystrophies and nemaline myopathy andthis is associated with the existence of longitudinal fila-ments in the diseased muscles. Muscle samples fromboth Becker and Duchene muscular dystrophy patientscontain a significant number of regenerating fibers anddisplay extensive longitudinal structures when immuno-labelled with the Tm4 antibody. Thus, Tm4 appears tobe a very good and sensitive indicator of regeneration/repair/remodelling and is potentially diagnostic of mus-cle disease where regeneration/repair/remodelling is afeature.

The Role of Tm4-Defined Filaments at the Z-LAC

Tm-defined filaments in the Z-LAC region may as-sociate with a number of structures located adjacent tothe Z-line, including the transverse tubules (T-tubules)or the sarcoplasmic reticulum (SR). The role of the T-

Fig. 8. The level of Tm4 protein is elevated in myopathic muscles.

Western blotting was performed with the Tm4 antibody using human

muscular dystrophy or nemaline myopathy patient samples and

muscles from TPM3 (Met9Arg) mice. Homogenates of 100 lm2 sec-

tions of muscle tissue from patient samples were loaded per lane. (A)Normal muscle: infant (8 weeks), child (5 years), adult (38 years).

Duchenne Muscular Dystrophy: DMD1 (5 years), DMD2 (9 years).

Becker Muscular Dystrophy: BMD1 (11 years), BMD2 (12 years).

Nemaline myopathy: NM1 [55 year-old with TPM3(Met9Arg) muta-

tion], NM2 [22 year-old with ACTA1(V163M)]. Sarcomeric Tm lev-

els detected with the CH1 antibody indicate the amount of contractile

proteins in each sample and act as a loading control showing the

amount of skeletal muscle protein present in the sample. (B) Quantita-tion of Tm4 protein levels relative to sarcomeric Tm.

82 Vlahovich et al.

tubule system is to propagate action potentials to triggercalcium release from the sarcoplasmic reticulum in orderto cause muscle contraction. This system is subjected toextreme mechanical forces during contraction and relax-ation of the muscle fibre and is supported by an underly-

ing cytoskeleton. This cytoskeletal structure has beenshown to contain molecules such as the focal adhesionproteins vinculin and talin as well as membrane mole-cules such as dystrophin and spectrin [Kostin et al.,1998]. The localization of both Tm4 and Tm5NM1 [Kee

Fig. 9. Tm4 is present in Z-LAC and longitudinal structures in myopathic human muscles. Myopathic

human samples were treated with Tm4 (A,D,G) and a-actinin (B,E,H) antibodies to examine the localiza-

tion of Tm4 structures in control (A–C), human dystrophic (D–F) and nemaline (G–I) muscle samples.

Tm4 is present only in Z-LAC structures in control muscle (A,C). Tm4 longitudinal structures are evident

in sections from human Becker muscular dystrophy muscle (D,F, arrowheads) and TPM3(Met9Arg) mus-

cle (G,I, arrowheads). Scale bar5 5 lm.

Tm4 in Muscle Disease and Repair 83

et al., 2004] in this region suggests that cytoskeletal Tmsmay play a role in supporting these structures.

CONCLUSIONS

We conclude that the cytoskeletal Tm isoform,Tm4, defines novel cytoskeletal compartments in bothhealthy muscle and muscle undergoing regeneration/repair/remodelling and that this Tm isoform is a sensitiveindicator of muscle disease that involves repair.

ACKNOWLEDGMENTS

We thank Ross Boadle, Electron Microscope Labo-ratory (Westmead Research Hub Facility), for his expertadvice with sectioning and microscopy, Gordon Lynchfor providing his notexin-induced muscle regenerationprotocol prior to publication and Tina Borovina foradvice and care of the mice. PWG is a NHMRC Princi-pal Research Fellow.

REFERENCES

Bryce NS, Schevzov G, Ferguson V, Percival JM, Lin JJ, Matsumura

F, Bamburg JR, Jeffrey PL, Hardeman EC, Gunning P, Wein-

berger RP. 2003. Specification of actin filament function and

molecular composition by tropomyosin isoforms. Mol Biol

Cell 14:1002–1016.

Clark KA, McElhinny AS, Beckerle MC, Gregorio CC. 2002. Striated

muscle cytoarchitecture: An intricate web of form and func-

tion. Annu Rev Cell Dev Biol 18:637–706.

Corbett MA, Robinson CS, Dunglison GF, Yang N, Joya JE, Stewart

AW, Schnell C, Gunning PW, North KN, Hardeman EC. 2001.

A mutation in alpha-tropomyosin (slow) affects muscle

strength, maturation and hypertrophy in a mouse model for

nemaline myopathy. Hum Mol Genet 10:317–328.

Dalby-Payne JR, O’Loughlin EV, Gunning P. 2003. Polarization of

specific tropomyosin isoforms in gastrointestinal epithelial

cells and their impact on CFTR at the apical surface. Mol Biol

Cell 14:4365–4375.

Ervasti JM. 2003. Costameres: The Achilles’ heel of Herculean mus-

cle. J Biol Chem 278:13591–13594.

Goldspink G. 1999. Changes in muscle mass and phenotype and the

expression of autocrine and systemic growth factors by muscle

in response to stretch and overload. J Anat 194 (Part 3):323–

334.

Griffiths G, McDowall A, Back R, Dubochet J. 1984. On the prepara-

tion of cryosections for immunocytochemistry. J Ultrastruct

Res 89:65–78.

Gunning P, Gordon M, Wade R, Gahlmann R, Lin CS, Hardeman E.

1990. Differential control of tropomyosin mRNA levels during

myogenesis suggests the existence of an isoform competition-

autoregulatory compensation control mechanism. Dev Biol

138:443–453.

Gunning P, Weinberger R, Jeffrey P, Hardeman E. 1998. Isoform sort-

ing and the creation of intracellular compartments. Annu Rev

Cell Dev Biol 14:339–372.

Gunning PW, Schevzov G, Kee AJ, Hardeman EC. 2005. Tropomyo-

sin isoforms: Divining rods for actin cytoskeleton function.

Trends Cell Biol 15:333–341.

Hanft LM, Rybakova IN, Patel JR, Rafael-Fortney JA, Ervasti JM.

2006. Cytoplasmic gamma-actin contributes to a compensatory

remodeling response in dystrophin-deficient muscle. Proc Natl

Acad Sci USA 103:5385–5390.

Hutchinson DO, Charlton A, Laing NG, Ilkovski B, North KN. 2006.

Autosomal dominant nemaline myopathy with intranuclear

rods due to mutation of the skeletal muscle ACTA1 gene: Clin-

ical and pathological variability within a kindred. Neuromuscul

Disord 16:113–121.

Jarvinen TA, Jarvinen TL, Kaariainen M, Kalimo H, Jarvinen M.

2005. Muscle injuries: Biology and treatment. Am J Sports

Med 33:745–764.

Joya JE, Kee AJ, Nair-Shalliker V, Ghoddusi M, Nguyen MA, Luther

P, Hardeman EC. 2004. Muscle weakness in a mouse model of

nemaline myopathy can be reversed with exercise and reveals a

novel myofiber repair mechanism. Hum Mol Genet 13:2633–

2645.

Kee AJ, Schevzov G, Nair-Shalliker V, Robinson CS, Vrhovski B,

Ghoddusi M, Qiu MR, Lin JJ, Weinberger R, Gunning PW,

Hardeman EC. 2004. Sorting of a nonmuscle tropomyosin to a

novel cytoskeletal compartment in skeletal muscle results in

muscular dystrophy. J Cell Biol 166:685–696.

Kostin S, Scholz D, Shimada T, Maeno Y, Mollnau H, Hein S,

Schaper J. 1998. The internal and external protein scaffold of

the T-tubular system in cardiomyocytes. Cell Tissue Res

294:449–460.

Laing NG, Wilton SD, Akkari PA, Dorosz S, Boundy K, Kneebone C,

Blumbergs P, White S, Watkins H, Love DR. 1995. A mutation

in the alpha tropomyosin gene TPM3 associated with autoso-

mal dominant nemaline myopathy. Nat Genet 9:75–79.

Lloyd CM, Berendse M, Lloyd DG, Schevzov G, Grounds MD. 2004.

A novel role for non-muscle gamma-actin in skeletal muscle

sarcomere assembly. Exp Cell Res 297:82–96.

Lucas CA, Hoh JF. 2003. Distribution of developmental myosin heavy

chains in adult rabbit extraocular muscle: Identification of a

novel embryonic isoform absent in fetal limb. Invest Ophthal-

mol Vis Sci 44:2450–2456.

McLoon LK, Rowe J, Wirtschafter J, McCormick KM. 2004. Continu-

ous myofiber remodeling in uninjured extraocular myofibers:

Myonuclear turnover and evidence for apoptosis. Muscle

Nerve 29:707–715.

McMichael BK, Kotadiya P, Singh T, Holliday LS, Lee BS. 2006.

Tropomyosin isoforms localize to distinct microfilament popu-

lations in osteoclasts. Bone 39:694–705.

Mendler L, Zador E, Dux L, Wuytack F. 1998. mRNA levels of myo-

genic regulatory factors in rat slow and fast muscles regenerating

from notexin-induced necrosis. Neuromuscul Disord 8:533–541.

Nakata T, Nishina Y, Yorifuji H. 2001. Cytoplasmic gamma actin as a

Z-disc protein. Biochem Biophys Res Commun 286:156–163.

Percival JM, Thomas G, Cock TA, Gardiner EM, Jeffrey PL, Lin JJ,

Weinberger RP, Gunning P. 2000. Sorting of tropomyosin iso-

forms in synchronised NIH 3T3 fibroblasts: Evidence for dis-

tinct microfilament populations. Cell Motil Cytoskeleton 47:

189–208.

Percival JM, Hughes JA, Brown DL, Schevzov G, Heimann K, Vrhov-

ski B, Bryce N, Stow JL, Gunning PW. 2004. Targeting of a

tropomyosin isoform to short microfilaments associated with

the Golgi complex. Mol Biol Cell 15:268–280.

Plant DR, Colarossi FE, Lynch GS. 2006. Notexin causes greater

myotoxic damage and slower functional repair in mouse skele-

tal muscles than bupivacaine. Muscle Nerve 34:577–585.

Rybakova IN, Patel JR, Ervasti JM. 2000. The dystrophin complex

forms a mechanically strong link between the sarcolemma and

costameric actin. J Cell Biol 150:1209–1214.

84 Vlahovich et al.

Sanger JW, Kang S, Siebrands CC, Freeman N, Du A, Wang J, Stout

AL, Sanger JM. 2005. How to build a myofibril. J Muscle Res

Cell Motil 26:343–354.

Sanoudou D, Haslett JN, Kho AT, Guo S, Gazda HT, Greenberg SA,

Lidov HG, Kohane IS, Kunkel LM, Beggs AH. 2003. Expres-

sion profiling reveals altered satellite cell numbers and glyco-

lytic enzyme transcription in nemaline myopathy muscle. Proc

Natl Acad Sci USA 100:4666–4671.

SanoudouD, CorbettMA,HanM,GhoddusiM,NguyenMA,VlahovichN,

Hardeman EC, Beggs AH. 2006. Skeletal muscle repair in a mouse

model of nemalinemyopathy. HumMolGenet 15:2603–2612.

Schevzov G, Vrhovski B, Bryce NS, Elmir S, Oiu MR, O’Neill GM,

Yang N, Verrills NM, Kavallaris M, Gunning PW. 2005. Tis-

sue-specific tropomyosin isoform composition. J Histochem

Cytochem 53:557–570.

Sonnemann KJ, Fitzsimons DP, Patel JR, Liu Y, Schneider MF,

Moss RL, Ervasti JM. 2006. Cytoplasmic gamma-actin

is not required for skeletal muscle development but its ab-

sence leads to a progressive myopathy. Dev Cell 11:387–

397.

Tanabe Y, Esaki K, Nomura T. 1986. Skeletal muscle pathology in X

chromosome-linked muscular dystrophy (mdx) mouse. Acta

Neuropathol (Berl) 69:91–95.

Watchko JF, O’Day TL, Hoffman EP. 2002. Functional characteristics

of dystrophic skeletal muscle: Insights from animal models.

J Appl Physiol 93:407–417.

Tm4 in Muscle Disease and Repair 85