Structural Organization of the Perimysium in Bovine Skeletal Muscle: Junctional Plates and Associated Intracellular Subdomains E

Structural organization of the perimysium in bovine skeletal muscle: Junctional plates and associated intracellular subdomains E. Passerieux, R. Rossignol, A. Chopard, A. Carnino, J.F. Marini, T. Letellier, J.P. Delage

To cite this version:

E. Passerieux, R. Rossignol, A. Chopard, A. Carnino, J.F. Marini, et al.. Structural organization of the perimysium in bovine skeletal muscle: Junctional plates and associated intracellular subdomains. Journal of Structural Biology, Elsevier, 2006, 154 (2), pp.206 - 216. 10.1016/j.jsb.2006.01.002. hal- 01758589

HAL Id: hal-01758589 https://hal.umontpellier.fr/hal-01758589 Submitted on 4 Apr 2018

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Journal of Structural Biology 154 (2006) 206–216 www.elsevier.com/locate/yjsbi

Structural organization of the perimysium in bovine skeletal muscle: Junctional plates and associated intracellular subdomains

E. Passerieux a, R. Rossignol a, A. Chopard b, A. Carnino b, J.F. Marini b, a a, T. Letellier , J.P. Delage ¤

a INSERM, U688 Physiopathologie Mitochondriale, Université Victor Segalen-Bordeaux 2, 146 rue Léo-Saignat, F-33076 Bordeaux Cedex, France b EA 3837, Laboratoire de Physiologie des Adaptations, Performances Motrices et Santé, Université de Nice Sophia Antipolis, Faculté des Sciences du Sport, 261 route de Grenoble, 06205 Nice Cedex, France

Received 19 October 2005; received in revised form 6 January 2006; accepted 7 January 2006 Available online 8 February 2006

Abstract

We analyzed the structural features of the perimysium collagen network in bovine Flexor carpi radialis muscle using various sample preparation methods and microscopy techniques. We Wrst observed by scanning electron microscopy that perimysium formed a regular network of collagen Wbers with three hierarchical levels including (i) a loose lattice of large interwoven Wbers ramiWed in (ii) numerous collagen plexi attaching together adjacent myoWbers at the level of (iii) speciWc structures that we call perimysial junctional plates. Second, we looked more closely at the intracellular organization underneath each plate using transmission electron microscopy, immunohistochemistry, and a three-dimensional reconstruction from serial sections. We observed the accumulation of myonuclei arranged in clusters surrounded by a high density of subsarcolemmal mitochondria and the proximity of capillary branches. Third, we analyzed the distribution of these perimysial junctional plates, subsarcolemmal mitochondria, and myonuclei clusters along the myoWbers using a statistical analysis of the distances between these structures. This revealed a global colocalization and the existence of adhesion domains between endomysium and perimysium. Taken together, our observations give a better description of the perimysium organization in skeletal muscle, and provide evidence that perimysial junctional plates with associated intracellular subdomains may participate in the lateral transmission of contractile forces as well as mechanosensing. © 2006 Elsevier Inc. All rights reserved.

Keywords: Perimysium; Myonuclear domains; Mitochondria; Muscle contraction; Skeletal muscle

1. Introduction tendinous junction (Borg and CaulWeld, 1980; Swasdison and Mayne, 1989; Trotter and Purslow, 1992). It is orga- The extracellular matrix (ECM)1 plays a fundamental nized as a regular network of thin Wbrillar collagen Wbers role, both in structural and functional aspects of skeletal composed mostly of type IV, as well as types III, VI (Listrat muscle. Its two components are the endomysium and the et al., 1999), and XII (Listrat et al., 2000) to a minor extent. perimysium. The endomysium is made of a continuous This network attaches myoWbers sarcolemma to speciWc sheath covering the full length of myoWbers until the myo- transmembrane proteins (Kovanen, 2002), thereby forming a regular mosaic pattern (Borg and CaulWeld, 1980). These * Corresponding author. Fax: +33 5 57 57 16 12. proteins, such as integrins or dystrophin–glycoprotein com- E-mail address: [email protected] (J.P. Delage). plexes (Rando, 2001), interact directly with the cytoskele- 1 Abbreviations used: COX, cytochrome c oxidase; DAPI, 4Ј,6-diamidi- ton (Berthier and Blaineau, 1997) and transmit the no-2-phenylindole, dihydrochloride; ECM, extracellular matrix; KS, contractile forces between adjacent myoWbers (Monti et al., Kolmogorov–Smirnov; PJP, perimysial junctional plate; SEM, scanning electron microscopy; TEM, transmission electron microscopy; 3D, three- 1999; Sheard et al., 2002) as far as the tendon (Purslow, dimensional. 2002). Taking into account this particular anatomy and

1047-8477/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2006.01.002 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 207 associated roles, the endomysium can be considered the easily visualize the perimysium between adjacent myoWbers, main ECM component involved in muscle Xexibility. or (ii) a fracture technique that does not alter the chemical In contrast, the second component of ECM, i.e., the peri- structure of connective tissue, and makes it possible to mysium, is formed by an areolar network of crimped colla- remove the endomysium and visualize perimysium connec- gen Wbers varying in diameter and composition, including tions at the surface of myoWbers. The digestion technique of essentially type I collagen in conjunction with types III, VI NaOH cell maceration was adapted from Ohtani et al. (Listrat et al., 1999), and XII (Listrat et al., 2000). However, (1991). For this, muscle samples were immersed in NaOH few studies have analyzed the perimysium’s Wne organization 6N and maintained at a temperature of 18°C for 5 days, so perimysium continues to be considered a simple packing before being rinsed for 3 days in water at 18 °C. Afterward, tissue, even though some early observations have suggested a the samples were Wxed in 2% tannin, freeze-dried, and frac- connection with endomysium (Borg and CaulWeld, 1980; tured. The sections were coated with gold and examined on Moore, 1983) and a possible role in the lateral transmission a Philips 515 scanning electron microscope. The fracture of contractile forces (Huijing et al., 1998). In particular, there technique consisted in the Wxation of muscle samples in 2% is a lack of detail concerning contact sites between perimy- osmic acid, followed by freeze-drying and fracture. For sium and endomysium, their organization, distribution along each analysis, a series of at least Wve diVerent samples were myoWbers, and underlying intracellular environment. We taken from three diVerent muscles. analyzed these features of the perimysium collagen network with a particular emphasis on endomysium contact sites, as 2.2. Transmission electron microscopy well as the associated intracellular subdomains in bovine Flexor carpi radialis muscle using scanning electron micros- The Flexor carpi radialis was stretched by a 300-g weight copy (SEM), transmission electron microscopy (TEM), to preserve muscle Wber straightness and immediately Wxed three-dimensional reconstruction from serial sections, and by the injection of solution A (0.5% glutaraldehyde, 2% immunohistochemistry techniques. paraformaldehyde, 7% saccharose, and 4% polyvinylpyroli- Our results show that the perimysium forms a network done in 0.1 M cacodylate buVer). This made it possible to of collagen Wbers with three hierarchical levels including (i) increase intramuscular pressure and separate muscle Wbers a regular lattice of interwoven Wbers with (ii) collagen plexi for better visualization of the connective tissue. The muscle at each angle that attach adjacent myoWbers at (iii) particu- was sectioned in 40 mm3 blocks, immersed for 1 h in Wxative lar domains that we call perimysial junctional plates (PJPs). solution A, and rinsed with water. These blocks were recut A three-dimensional reconstruction from serial sections into smaller samples of about 12 mm3 and immersed in a showed that, underneath each plate, there was an accumu- second Wxative solution B (2% osmic acid in 0.1 M cacodyl- lation of myonuclei arranged in clusters near capillary ate buVer) for 1 h. After dehydration, these blocks were branches. Surrounding these clusters, we observed a high embedded in epoxy (epon) resin and cut into longitudinal or density of subsarcolemmal mitochondria by immunohisto- transversal sections 0.1 !m thick. These sections were chemistry and TEM. Comparison of the data obtained by stained with a solution of uranyl acetate and lead citrate. these various techniques enabled us to propose a model of They were observed on a Philips CM10 microscope to ana- perimysium organization that may underline its role in lat- lyse the contact sites between perimysium and myoWbers as eral transmission of contractile forces and thus in various well as the intracellular subdomains. aspects of muscle Xexibility. 2.3. Three-dimensional reconstruction 2. Materials and methods Samples of embedded material (see above) were cut by All procedures were performed in accordance with insti- ultramicrotome parallel to myoWber direction to obtain a tutional guidelines for animal care. The Flexor carpi radialis series of 55 successive sections (2 !m thick). Sections were muscles of the foreleg were taken from ten diVerent cows, stained with toluidine blue and photographed on an light just after slaughter, and immediately Wxed for electron microscope (Zeiss). These photographs were then analyzed microscopy analyses or frozen for immunohistochemistry, with Photoshop 7.0 (Adobe) to isolate the surroundings of as detailed below. six adjacent myoWbers, and to determine the respective positions of myonuclei and capillaries. Images of the myo- 2.1. Scanning electron microscopy nucleus and capillary positions along the myoWbers from each section were submitted to a three-dimensional recon- The Flexor carpi radialis was immersion-Wxed in 10% struction (stereo pairs) using Image J software (http:// paraformaldehyde for 3 days before being sectioned in rsb.info.nih.gov). samples of 20 cm3. To study the perimysium collagen network’s Wne structure, we conducted an scanning electron 2.4. Immunohistochemistry microscopy (SEM) analysis on muscle samples submitted to either (i) a digestion technique (NaOH) that enabled to Samples of Flexor carpi radialis muscle were rapidly fro- free and detach partially the endomysium so we could more zen in liquid nitrogen-cooled isopentane and stored at 208 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216

80 °C. Longitudinal and transversal sections (10 !m thick) microscopy, and the corresponding nuclei were excluded were¡ obtained using a cryostat (Frigocut 2800, Reichert from the analysis. The distances between the centers of Jung) at 22 °C. The sections were Wxed with a solution of myonuclei clusters were measured and expressed in !m. For 4% paraformaldehyde¡ (Sigma–Aldrich) for 30 min at 25 °C. three-dimensional reconstruction, distances between the After two washes with PBS (5min each), the sections were centers of myonuclei clusters were measured on a series of 5 dehydrated by successive baths of 70, 90, and 100 ethanol sections (2 !m thick) to compare with the results of immu- (2 min each), followed by rehydration with successive baths nohistochemistry on 10 !m sections. Mitochondria were of 90 and 70% ethanol, followed by two washes in PBS stained green using an antibody directed against one sub- (2 min each). Sections were then incubated in a blocking unit of the cytochrome c oxidase (COX) complex of the solution [10% foetal calf serum (Hyclone) diluted in PBS] respiratory chain. The distances between the centers of sub- for 2 h at 25 °C. Next, they were incubated overnight with sarcolemmal mitochondria sections were measured and two primary antibodies (rabbit polyclonal anti-collagen III, expressed in !m. This entire analysis was performed on 10 Rockland and mouse monoclonal antibovine cytochrome c images of SEM, 20 immunohistochemistry photographs, oxidase subunit IV, Mitosciences). After two washes in PBS and 55 optic microscopy photographs (three-dimensional (10 min each), the sections were incubated in 10% normal reconstruction) using Image J software. We also measured goat serum (Sigma) in PBS for 1 h at 25 °C. Subsequently, the distances between the centers of the various structures sections were incubated with the secondary antibodies by pairs for an interobject distribution analysis. (Alexa Fluor 546 goat anti-rabbit IgG and Alexa Fluor 488 goat anti-mouse IgG, Molecular Probes) for 90 min at 2.6. Statistics 25 °C before being washed three times with PBS (15min each). Finally, sections were incubated with 300 nM 4Ј, An initial statistical analysis was performed to individu- 6-diamidino-2-phenylindole, dihydrochloride (DAPI, ally analyse the intra-distribution of various structures Molecular Probes) for 10 min at 25 °C, washed three times (PJPs, myonuclei clusters, and subsarcolemmal mitochon- in PBS (10 min each), and mounted with Mowiol 4–88 (Cal- dria). Another statistical analysis was also performed to biochem). The sections were observed on an epiXuorescence ascertain their inter-distribution and potential colocaliza- microscope (Nikon Eclipse E600), photographed by a digi- tion. We represented histograms of the distribution of distal camera (COHU High Performance CCD Camera), and tances between the centers of these structures as performed acquired using Visiolab software (Biocom). in Kassab et al. (1997) to study the intra-distribution of one type of structure along individual Wbers. We used two 2.5. Morphometric and colocalization analysis diVerent methods to study the inter-distribution and potential colocalization between diVerent types of structures. To determine the distribution of PJPs, images obtained First, we compared histograms obtained from the intra-dis- by SEM and immunohistochemistry (Xuorescence micros- tribution analysis using the Kolmogorov–Smirnov test copy) were analyzed using Image J software. The contours (Mehrotra et al., 2005) (KS test) and Statview software 5.0 of each PJP were traced on electron micrographs, and the (SAS Institute). This test determines whether or not two distances between their respective centers were measured datasets diVer signiWcantly, and has the advantage of mak- and expressed in !m. On images acquired by Xuorescence ing no assumptions about data distribution (non-paramet- microscopy, the PJPs were visualized using an antibody ric and distribution-free). The threshold of signiWcance for raised against type III collagen (red alexa Xuorescent stain- colocalization between the various structures was deWned ing). This type of collagen is contained in both the endo- at p 7 0.05. Second, we performed a statistical analysis of mysium and perimysium so that their focal colocalization the distances between the centers of the diVerent structures appears in the form of deep red domains. Successive dense by pairs. We did this by calculating the frequency of the regions of collagen were considered to belong to the same peak value as determined on the distribution histograms. PJP when there was a series of dots (72) interspaced at a distance of 650 !m. These criteria were obtained from 3. Results observing PJPs on an electron micrograph. For this analysis, we did not consider the regions corresponding to capil- 3.1. Perimysium collagen network and junctional plates lary walls that were identiWed by comparing with phase contrast images. The distances between the centers of Muscle digestion with NaOH 6 N at 18 °C resulted in successive PJPs were once again measured and expressed the loosening and partial removal of the endomysium in !m. The myonuclei were stained using 4Ј,6-diamidino- from embedded adjacent myoWbers (Figs. 1A–D). This 2-phenylindole (DAPI) to visualize DNA as deep blue dots. enabled us to directly observe the perimysium network Successive nuclei were considered to belong to the same and its connections with the myoWber surface. On lateral cluster when they were interspaced by a distance of (Fig. 1A) and front (Fig. 1B) views, the perimysium net- 650 !m as determined on the three-dimensional recon- work appeared to consist of a combination of several col- struction from serial sections. The presence of capillaries lagen elements including numerous large interwoven was identiWed on each section using phase contrast Wbers at an angle of about 60° with regard to myoWber E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 209

Fig. 1. The perimysium collagen network and junctional plates. Electron micrographs (SEM) of skeletal muscle Wbers prepared according to the NaOH digestion technique (6N at 18 °C) revealed (A) on lateral and (B) front views the overall lattice of collagen interwoven Wbers, (C) the distribution of plexi at the surface of myoWber (noted p), and (D) plexi adhering to the surface of adjacent myoWbers. Note the collagen Wbers crossing myoWbers at an angle of approximately 60°. Electron micrographs of skeletal muscle Wbers prepared according to the fracture technique show in greaster detail (E) the attachment of perimysium plexi to myoWbers at the level of a particular structure that we call “the perimysial junctional plate” (PJP). PJP are approximately 150– 200 !m in length. High magniWcation (F) shows the merging of the perimysium with the endomysium (noted e) at the extremities of two branches of the same plexus, which seems to connect with the outer surface of the sarcolemma at and between costameric structures. length. These Wbers formed a plexus of thin branches study the adhesion between perimysium and myoWber (noted p on Figs. 1C and D) at each angle bound to the endomysium in more detail, we performed a muscle frac- surface of adjacent myoWbers. This resulted in a patch- ture procedure (Figs. 1E and F). We again observed the work of plexi at the surface of myoWbers where large por- branching lattice of the perimysium collagen network, tions of membrane are devoid of any contact with the and more particularly the plexus that bound myoWber perimysium for distances of up to 300 !m (Fig. 1C). To endomysium for approximately 150–200 !m. This formed 210 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 an adhesive structure (Fig. 1E) that we called perimysial 3.2. Existence of intracellular subdomains associated with the junctional plate (PJP) consisting of about 10 principal perimysial junctional plates branches interspaced by a distance of 650 !m. Higher magniWcation revealed the attachment of branches at the To study PJPs and associated intracellular subcomposi- myoWber surface, forming adhesive regions both on the tion, we performed a TEM analysis on muscle samples top and between costameric structures (Fig. 1F). More- injected with a Wxative solution. On longitudinal sections over, we observed that, despite the fracture procedure, (Fig. 2A), we observed the association of myonuclei (noted Wbrils of the endomysium (e) remained present only in mn) and subsarcolemmal mitochondria (noted sm), in those regions where perimysium plexi were attached to proximity to bundles of collagen (noted p) and capillary myoWbers. At this particular level, the perimysium (which branches (noted Ca). Then, we looked more closely at the could not be distinguished from the endomysium), formed junction between connective tissue and these intracellular an important area for adhering to the plasma membrane domains using a higher magniWcation (Figs. 2B–D). The that may constitute focal points for attachment between cross-striation of collagen Wbers indicated a type I origin connective tissue and muscle. that is the major component of the perimysium, but absent

Fig. 2. Intracellular subdomains associated with PJPs. Electron micrographs (TEM) of (A–C–D) longitudinal and (B) transversal sections of skeletal muscle Wbers show (A and B) the particular association of perimysium collagen bundles (noted p) with the endomysium (noted e), a capillary (noted ca), and intracellular subdomains including subsarcolemmal mitochondria (noted sm) and myonuclei (noted mn). Note the uneven densiWcations (noted d) of sarcolemma at the junction with the perimysium and the presence of caveolae (noted cv). On the longitudinal sections, we highlight the association between the PJP with (C) mitochondria and (D) myonuclei. (C) Note the adhesion of collagen Wbers between Z-bands. E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 211 from the endomysium (Listrat et al., 1999). Thus, the regions of connective tissue observed between adjacent myoWbers (Fig. 2B) must belong to the perimysium in contrast with the endomysium (noted e) that solely envelopes myoWbers. Moreover, the presence of type I collagen Wbers inWltrated between adjacent myoWbers indicated that the perimysium network not only covered and connected longitudinal muscle surface (as seen in Fig. 1), but also strongly attached individual myoWbers within the interstitial space. In each instance, attachment occurred at the level of PJPs that formed strong adhesive structures on myoWbers as seen on Figs. 1F, 2B–D. At these junctions, the sarcolemma showed darker regions that corresponded to membrane densiWcations (noted d on Figs. 2B and C) with an accumulation of caveolae (noted cv) in unusual places that did not correspond to the costameric structures of cytoskeleton attachment on sarcolemma. We frequently observed myonuclei and subsarcolemmal mitochondria underneath these membrane densiWcations. To better identify and characterize the organization of intracellular subdomains with relation to extracellular environment, we realized a three-dimensional reconstruction of capillary bed position and the position of myonuclei in a muscle volume containing six adjacent myoWbers (Fig. 3A). Capillary bed reconstruction (Fig. 3B) from the adjoining faces of six adjacent myoWbers (noted A–B–C–D–E–F) revealed a longitudinal and linear network with the typical Fig. 3. Arrangement of a capillary bed and myonuclei in clusters. Stereo- branched organization described by Schraufnagel et al. scopic pairs of reconstructions from serial semi-thin sections of (A) mus- (1983) and Reina-De La Torre et al. (1998). The three- cle volume containing six adjacent myoWbers (noted A–B–C–D–E–F) dimensional view of the capillary bed showed short transver- showing the arrangement (B) of both capillaries and myonuclei at the W sal ramiWcations between adjacent myoWbers that often adjoining faces of six adjacent myo bers (A–B–C–D–E–F) seen from a lateral view, and (C) the position of myonuclei at the adjoining faces of formed branches every 150 !m as well as often colocalizing four adjacent myoWbers (A–B–C–D) of the same volume. Note the fre- with myonuclei. This indicates a possible topological associa- quent association between myonuclei and the capillary bed, and the tion between PJPs, myonuclei, and capillary branches. The arrangement of myonuclei in clusters overlapping adjacent myoWbers. position of myonuclei on the adjoining faces of four adjacent myoWbers (A–B–C–D) was examined (Fig. 3C) to identify the possible association between PJPs and myonuclei. We subsarcolemmal (high) mitochondria. We frequently observed that myonuclei belonging to diVerent individual observed myonuclei clusters surrounded by subsarcolem- Wbers formed clusters of approximately 15 U where myonu- mal mitochondria (Figs. 4C and D) under regions of deep clei were interspaced by a distance of 650!m. Each cluster red Xuorescence identiWed as PJPs. The presence of capillar- was approximately 150–200 !m in length, and separated ies in interstitial space was also observed both in transver- from subsequent clusters by a distance of approximately sal and longitudinal sections (noted Ca on Figs. 4A and B), 150 !m. This topological position of myonuclei is very much and conWrmed by phase contrast microscopy (Figs. 4E and in accordance with the SEM micrograph (Fig. 1D), with peri- F). However, the existence of capillary branches was only mysium plexi binding adjacent myoWbers. revealed by the 3D reconstruction described in Fig. 3. We repeated this structural analysis of PJPs, myonuclei, and subsarcolemmal mitochondria by Xuorescence micros- 3.3. Colocalization of perimysial junctional plates and copy on both longitudinal and transversal muscle sections. associated intracellular subdomains The PJPs corresponded to the deep red Xuorescent domains (please see criteria in the methods section) and appeared to We conducted a morphometric analysis on images be variable in length and number according to the myoWber obtained by SEM and immunohistochemistry to study the (Figs. 4A, B, and F). The myonuclei were visualized as blue intra-distribution of PJPs and associated intracellular com- dots (Figs. 4A–D and F) and showed the cluster organiza- ponents along myoWbers. The respective histograms of the tion previously determined by phase contrast microscopy mean distances between the centers of successive PJPs (see Fig. 3C). The mitochondria corresponded to green are given in Figs. 5A and B. These both show a wide Xuorescence (Figs. 4A–D and F). We observed diVerences distribution of PJPs with the same most frequent value of in COX expression between intermyoWbrillar (low) and 50–60 !m. Likewise, the distribution of distances between 212 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216

Fig. 4. Distribution of PJPs and associated intracellular components. Images of muscle sections from Xuorescence microscopy with speciWc labelling of type III collagen (red), nuclei (blue), and mitochondria (green). (A) Observations at low magniWcation on longitudinal sections show the existence of deep red domains corresponding to the PJPs associated with clusters of myonuclei. Images of (B) transversal or (C and D) longitudinal sections at higher mag- niWcation show the close association between PJPs, myonuclei, and subsarcolemmal mitochondria. Note the diVerence between subsarcolemmal mitochondria (deep green) and intermyoWbrillar mitochondria (paler green). (E) Phase contrast observation of a transversal section shows that the capillary (Ca) is also associated (F) with deep red domains, myonuclei clusters, and subsarcolemmal mitochondria. the centers of successive myonuclei clusters is indicated in from 0.63 to 0.8. This indicates that the distribution of PJPs Fig. 5C. This also shows a wide distribution, with a most in muscle closely corresponds to the distribution of myonu- frequent value of 50–60 !m. A similar result was also clei clusters and subsarcolemmal accumulations of mito- observed for subsarcolemmal mitochondria (Fig. 5D). chondria. Hence, we observed a strong similarity between the histograms obtained for PJPs, myonuclei clusters, and subsarco- 4. Discussion lemmal mitochondria. Moreover, the comparative statistical analysis of these histograms by pairs using the We analysed the Wne organization of the perimysium Kolmogorov–Smirnov test further indicated a signiWcant network in bovine Flexor carpi radialis muscle using various correspondence, with a p value of 0.88 between PJPs and sample preparation methods and microscopy techniques. myonuclei clusters, 0.99 between PJPs and subsarcolemmal We looked more into detail than previous studies at the mitochondria, and 0.61 between myonuclei clusters and contact sites between the perimysium and the endomysium, subsarcolemmal mitochondria. All these values are their distribution along myoWbers, and the underlying superior to 0.05, which implies a similar organization. To intracellular environment using SEM, TEM, three-dimen- ascertain potential colocalization, we studied the inter-dis- sional reconstruction from serial sections, and immunohis- tribution between PJPs, myonuclei clusters, and subsarco- tochemistry techniques. lemmal mitochondria. To do so, we once again conducted a Our observations of the perimysium collagen network morphometric analysis of the distances between the centers showed three hierarchical levels of organization that of the diVerent structures by pairs. The corresponding his- include a loose lattice of interwoven Wbers with collagen tograms are shown in Figs. 5E–G. They reveal a major plexi at each angle attaching adjacent myoWbers at the level colocalization of these three structures, as peak values were of speciWc domains that we call perimysial junctional plates obtained at a distance of <5 !m, with a frequency ranging (PJPs). Prior studies on the organization of the perimysium E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 213

Fig. 5. Colocalization of PJPs, myonuclei clusters, and subsarcolemmal mitochondria. Histograms of the intra-distribution of (A, by SEM and B, by Xuorescence microscopy) PJP, (C) myonuclei clusters, and (D) subsarcolemmal mitochondria. Histograms of the inter-distribution between (E) PJP and subsarcolemmal mitochondria, (F) PJP and myonuclei, and (G) PJP and subsarcolemmal mitochondria. Each bar corresponds to the mean distance between the centers of one type of structure (intra-distribution) or diVerent types of structures by pairs (inter-distribution). This is expressed in !m and was determined from measurements of 20 various longitudinal muscle sections observed in immunohistochemistry. in skeletal muscle essentially described this tissue as a loose regions between the endomysium and the perimysium. lattice surrounding myoWber bundles (Borg and CaulWeld, Micrographs taken by SEM clearly demonstrated the 1980; Moore, 1983; Rowe, 1981). All these works were per- nature of the strong attachment between the two compo- formed on various rat, bovine, and ovine skeletal muscles nents of the extracellular matrix, i.e., the endomysium and including the diaphragm, adductor major, gastrocnemius, perimysium which could indicate a focal region for delivery pectoralis, biceps, psoas major, soleus, semitendinosus, tension during muscle contraction. Moreover, we observed semimembranosus, and longissimus dorsi. They all gave the a sarcolemmal membrane densiWcation in these regions in same description of the perimysium that seemed to be simi- addition to those observed at the costameric junction of the lar between species, which may underly a possible function- endomysium (Young et al., 2000), and which also resem- ality of this tissue. We obtained more details on perimysial bled those of the myotendinous junction (Swasdison and structural features in bovine skeletal muscle, and speciWed Mayne, 1989). These two systems, i.e , tendon and endo- its organization while observing a predominant pattern of mysium, were previously described as playing a fundamen- interwoven Wbers at the surface of myoWbers. At high mag- tal role in the transmission of contractile forces (Trotter niWcation, we observed that each perimysium collagen Wber and Purslow, 1992; Tidball, 1991), which further suggests formed a plexus in close proximity to muscle surface and that PJPs could take part in this process. We also observed attached adjacent myoWbers at a level we call “perimysial the presence of numerous caveolae under each PJP, possi- junctional plates” (PJPs). These can be deWned as contact bly indicating a role for these domains in muscle plasticity 214 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 and physiological adaptation (Minetti et al., 2002; van notion of myonuclear domains (Hall and Ralston, 1989), Deurs et al., 2003; Woodman et al., 2004). and could suggest that perimysium organization is associ- We thus looked more closely at PJPs to show their ated with myoWber activity. This hypothesis is supported by potential role in the functioning of muscle physiology. In the existence of a functional collaboration between these particular, we investigated the composition, position, and myonuclei, subsarcolemmal mitochondria, and perimy- arrangement of the intracellular environment underneath sium. In addition, a functional connection between collagen each plate using TEM and immunohistochemistry tech- VI, a component of perimysium (Listrat et al., 1999), and niques. Interestingly, we observed that there was an accu- each of these structures was recently demonstrated for cal- mulation of myonuclei arranged in clusters, surrounded by cium transport and muscle contraction (Irwin et al., 2003). a higher density of subsarcolemmal mitochondria and Moreover, our study suggests that the association between nearby capillary branches, below each PJP. This cluster mitochondria and O2 supply (Mathieu-Costello et al., 2002; arrangement of myonuclei is in agreement with previous Suarez et al., 1991) may be completed by the presence of studies (Roy et al., 1999; Smith et al., 2000), but is also in adhesive plates (PJPs) and myonuclei clusters, possibly to slight contrast with the concept of “myonuclear domains,” form a functional domain involved in muscle contraction which considers individual myonuclei at the center of cyto- and mechanosensing. plasmic domains (Bruusgaard et al., 2003; Hall and Thus, our study reports the existence of particular 200 !m Ralston, 1989). Our statistical analysis of myonuclei cluster long domains where junctional structures and associated distribution showed that these domains exist, but instead of intracellular components are colocalized with each other that one they contain 10–15 nuclei. One might also expect the may participate in muscle integrity through the adhesion of Golgi apparatus and sarcoplasmic reticulum to be present adjacent myoWbers. Such a colocalization also supports pre- in these regions seeing as an association with myonuclear vious observations (Huijing et al., 1998) that demonstrate the domains was previously shown in rat gastrocnemius (Ral- perimysium’s role in the lateral transmission of contractile ston et al., 1999) or soleus muscle (Nori et al., 2003). In force. In this study, the contractile forces of rat Extensor digi- addition, the association of mitochondria and capillary torum longus were maintained even when the distal tendon branching was also reported in other studies (Mathieu- was cut, thereby proving a “non-myotendinous force trans- Costello et al., 2002; Suarez et al., 1991). Therefore, our mission.” Our structural observations suggest that such structural study showed that the perimysium network was forces could originate from contact sites between the perimy- regularly associated with the distribution of myonuclei and sium and myoWbers attached at the level of PJPs, and be subsarcolemmal mitochondria, but also with the capillary bed that was previously described as being included in a part of the perimysium network (Bosman and Stamenko- Perimysial vic, 2003). We conducted a statistical analysis on the distri- Collagen myofiber bution of PJPs, myonuclei clusters, and subsarcolemmal network mitochondria along Wbers to quantify our Wndings. To do so, we measured the distances between the centers of each 150 V structure and looked at their distribution in di erent indi- perimysium Plexus of Intracellular vidual myoWbers, as well as potential colocalization. Histo- subdomain branches myonuclei grams of these distances showed a disperse value ranging between 60 and 300 !m, with the most frequent distance of subsarcolemmal mitochondria 50–60 m for PJPs, myonuclei clusters, and subsarcolem- ! costamericmosaic mal mitochondria. In addition, the distribution of distances Perimysial membrane between the centers of each PJP was similar when densification Junctional performed on either electron micrographs or on immuno- sarcomere histochemistry images as veriWed using the Kolmogorov– plate Smirnov test (p >0.05). This also validated the use of type 5 µm III collagen dense regions as a means to identify PJPs in muscle sections. The histograms for these three diVerent Fig. 6. Model of perimysium organization in skeletal bovine muscle. The structures were similar, as veriWed using the KS test (with top panel represents a longitudinal view of a skeletal muscle covering p >0.05). This suggested the idea of a succession of domains 1 mm in length. We show the perimysium as a loose network of interwoven collagen Wbers as observed by SEM. At the angles of these interwoven where PJP, myonuclei clusters, and subsarcolemmal mito- Wbers, plexi branches originate and adhere to the surface of adjacent myo- chondria are located in close proximity to one another. This Wbers. This illustration in the second panel shows its branches as well as its view was supported by an inter-distribution analysis that association with a myonuclei cluster and the accumulation of subsarco- also revealed a frequent colocalization: 70% of cases with lemmal mitochondria in the intracellular subdomain. The bottom panel is t W a distance of <5 !m between PJPs, subsarcolemmal mito- a simpli ed drawing of the attachment region that we called the perimysial junctional plate (200 !m long) with several adhesion points on and chondria, and myonuclei clusters by pairs. These domains between the costameric structures, as well as associated membrane densiW- are 200 !m long and regularly repeated along Wbers. This cations as indicated in SEM and TEM views. A sarcomere has been particular repeated motif along muscle Wbers recalls the depicted on the bottom left to give an idea of the scale. E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216 215 transmitted within muscles through the well-ordered lattice Listrat, A., Lethias, C., Hocquette, J.F., Renand, G., Menissier, F., Geay, of perimysium collagen also observed in our study. Further- Y., Picard, B., 2000. Age-related changes and location of types I, III, XII and XIV collagen during development of skeletal muscles from more, we noticed that each PJP connected with the edges of V W genetically di erent animals. Histochem. J. 32, 349–356. costameric structures lying at the surface of myo bers. These Listrat, A., Picard, B., Geay, Y., 1999. Age-related changes and location of transmembranous proteins are likely to correspond to the type I, III, IV, V and VI collagens during development of four foetal integrins involved in signal transduction (Anastasi et al., skeletal muscles of double-muscled and normal bovine animals. Tissue 2003; Samson et al., 2004; van der Flier and Sonnenberg, Cell 31, 17–27. 2001). Interestingly, the presence of integrins at myotendi- Mathieu-Costello, O., Morales, S., Savolainen, J., Vornanen, M., 2002. Fiber capillarization relative to mitochondrial volume in diaphragm of nous junction with a possible implication in mechano-trans- shrew. J. Appl. Physiol. 93, 346–353. duction has been previously observed (Tidball, 1991). We Mehrotra, R., Singh, M., Gupta, R.K., Kapoor, A.K., 2005. Trends of prev- believe that the PJPs observed in our study may also partici- alence and pathological spectrum of head and neck cancers in North pate in a similar mechanosensing system in skeletal muscle. India. Indian J. Cancer 42, 89–93. Taken together, our observations give a better descrip- Minetti, C., Bado, M., Broda, P., Sotgia, F., Bruno, C., Galbiati, F., Volonte, D., Lucania, G., Pavan, A., Bonilla, E., Lisanti, M.P., Cordone, tion of the organization of the perimysium in skeletal mus- G., 2002. Impairment of caveolae formation and T-system disorganiza- cle. On the basis of numerous data provided by SEM and tion in human muscular dystrophy with caveolin-3 deWciency. Am. J. TEM mircrographs, as well as Xuorescence microscopy and Pathol. 160, 265–270. subsequent statistical analysis, we proposed a model repre- Monti, R.J., Roy, R.R., Hodgson, J.A., Edgerton, V.R., 1999. Transmission sented in Fig. 6. It highlights speciWc junctional structures of forces within mammalian skeletal muscles. J. Biomech. 32, 371–380. Moore, M.J., 1983. The dual connective tissue system of rat soleus muscle. (PJPs) with associated intracellular subdomains. Muscle Nerve 6, 416–422. Nori, A., Lin, P.J., Cassetti, A., Villa, A., Bayer, K.U., Volpe, P., 2003. Tar- Acknowledgments geting of alpha-kinase-anchoring protein (alpha KAP) to sarcoplasmic reticulum and nuclei of skeletal muscle. Biochem. J. 370, 873–880. The authors wish to thank INSERM, Université Victor Ohtani, O., Kikuta, A., Ohtsuka, A., Murakami, T., 1991. Organization of the reticular network of rabbit Peyer’s patches. Anat. Rec. 229, 251–258. Segalen Bordeaux 2, Association Française contre les Myo- Purslow, P.P., 2002. The structure and functional signiWcance of variations pathies (AFM), Association contre les Maladies Mitochon- in the connective tissue within muscle. Comp. Biochem. Physiol. A driales (Ammi) and Région Aquitaine for Wnancial support, Mol. Integr. Physiol. 133, 947–966. Service Commun de Microscopie (SERCOMI), Université Ralston, E., Lu, Z., Ploug, T., 1999. The organization of the Golgi complex Victor Segalen Bordeaux 2, for technical assistance, and and microtubules in skeletal muscle is Wber type-dependent. J. Neuro- sci. 19, 10694–10705. service veterinaire de Bordeaux for bovine muscle supply. Rando, T.A., 2001. The dystrophin-glycoprotein complex, cellular signal- ing, and the regulation of cell survival in the muscular dystrophies. References Muscle Nerve 24, 1575–1594. Reina-De La Torre, F., Rodriguez-Baeza, A., Sahuquillo-Barris, J., 1998. Anastasi, G., Amato, A., Tarone, G., Vita, G., Monici, M.C., Magaudda, Morphological characteristics and distribution pattern of the arterial L., Brancaccio, M., Sidoti, A., Trimarchi, F., Favaloro, A., Cutroneo, vessels in human cerebral cortex: a scanning electron microscope study. G., 2003. Distribution and localization of vinculin–talin–integrin sys- Anat. Rec. 251, 87–96. tem and dystrophin–glycoprotein complex in human skeletal muscle. Rowe, R.W., 1981. Morphology of perimysial and endomysial connective Immunohistochemical study using confocal laser scanning microscopy. tissue in skeletal muscle. Tissue Cell 13, 681–690. Cells Tissues Organs 175, 151–164. Roy, R.R., Monke, S.R., Allen, D.L., Edgerton, V.R., 1999. Modulation of Berthier, C., Blaineau, S., 1997. Supramolecular organization of the sub- myonuclear number in functionally overloaded and exercised rat plan- sarcolemmal cytoskeleton of adult skeletal muscle Wbers. A review. taris Wbers. J. Appl. Physiol. 87, 634–642. Biol. Cell 89, 413–434. Samson, T., Smyth, N., Janetzky, S., Wendler, O., Muller, J.M., Schule, Borg, T.K., CaulWeld, J.B., 1980. Morphology of connective tissue in skele- R., von der Mark, H., von der Mark, K., Wixler, V., 2004. The LIM- tal muscle. Tissue Cell 12, 197–207. only proteins FHL2 and FHL3 interact with alpha- and beta-sub- Bosman, F.T., Stamenkovic, I., 2003. Functional structure and composi- units of the muscle alpha7beta1 integrin receptor. J. Biol. Chem. tion of the extracellular matrix. J. Pathol. 200, 423–428. 279, 28641–28652. Bruusgaard, J.C., Liestol, K., Ekmark, M., Kollstad, K., Gundersen, Schraufnagel, D.E., Roussos, C., Macklem, P.T., Wang, N.S., 1983. The K., 2003. Number and spatial distribution of nuclei in the muscle Wbres geometry of the microvascular bed of the diaphragm: comparison to of normal mice studied in vivo. J. Physiol. 551, 467–478. intercostals and triceps. Microvasc. Res. 26, 291–306. Hall, Z.W., Ralston, E., 1989. Nuclear domains in muscle cells. Cell 59, Sheard, P., Paul, A., Duxson, M., 2002. Intramuscular force transmission. 771–772. Adv. Exp. Med. Biol. 508, 495–499. Huijing, P.A., Baan, G.C., Rebel, G.T., 1998. Non-myotendinous force Smith, H.K., Maxwell, L., Martyn, J.A., Bass, J.J., 2000. Nuclear DNA transmission in rat extensor digitorum longus muscle. J. Exp. Biol. 201 fragmentation and morphological alterations in adult rabbit skele- (Pt. 5), 683–691. tal muscle after short-term immobilization. Cell Tissue Res. 302, Irwin, W.A., Bergamin, N., Sabatelli, P., Reggiani, C., Megighian, A., Merlini, 235–241. L., Braghetta, P., Columbaro, M., Volpin, D., Bressan, G.M., Bernardi, Suarez, R.K., Lighton, J.R., Brown, G.S., Mathieu-Costello, O., 1991. P., Bonaldo, P., 2003. Mitochondrial dysfunction and apoptosis in myo- Mitochondrial respiration in hummingbird Xight muscles. Proc. Natl. pathic mice with collagen VI deWciency. Nat. Genet. 35, 367–371. Acad. Sci. USA 88, 4870–4873. Kassab, G.S., Pallencaoe, E., Schatz, A., Fung, Y.C., 1997. Longitudinal Swasdison, S., Mayne, R., 1989. Location of the integrin complex and position matrix of the pig coronary vasculature and its hemodynamic extracellular matrix molecules at the chicken myotendinous junction. implications. Am. J. Physiol. 273, H2832–H2842. Cell Tissue Res. 257, 537–543. Kovanen, V., 2002. Intramuscular extracellular matrix: complex environ- Tidball, J.G., 1991. Force transmission across muscle cell membranes. J. ment of muscle cells. Exerc. Sport Sci. Rev. 30, 20–25. Biomech. 24 (Suppl. 1), 43–52. 216 E. Passerieux et al. / Journal of Structural Biology 154 (2006) 206–216

Trotter, J.A., Purslow, P.P., 1992. Functional morphology of the endomys- Woodman, S.E., Sotgia, F., Galbiati, F., Minetti, C., Lisanti, M.P., ium in series Wbered muscles. J. Morphol. 212, 109–122. 2004. Caveolinopathies: mutations in caveolin-3 cause four dis- van der Flier, A., Sonnenberg, A., 2001. Function and interactions of inte- tinct autosomal dominant muscle diseases. Neurology 62, grins. Cell Tissue Res. 305, 285–298. 538–543. van Deurs, B., RoepstorV, K., Hommelgaard, A.M., Sandvig, K., 2003. Young, M., Paul, A., Rodda, J., Duxson, M., Sheard, P., 2000. Examination Caveolae: anchored, multifunctional platforms in the lipid ocean. of intrafascicular muscle Wber terminations: implications for tension Trends Cell Biol. 13, 92–100. delivery in series-Wbered muscles. J. Morphol. 245, 130–145.