Document Gz54jak3zqbZeVKgkym1obMyv

nature genetics LETTERS 20 03 N a tu re P u blish in g G ro u p http://www.nature.com/naturegenetics Mitochondrial dysfunction and apoptosis in myopathic mice with collagen VI deficiency William A Irwin1,2,6, Natascha Bergamin1,6, Patrizia Sahatelli3, Carlo Reggiani4, Aram. Megighian4, Luciano Merlini5, Paola Braghetta1, Marta Columbaro5, Dino Volpili1, Giorgio M Bressan1, Paolo Bernardi28cPaolo Bonaldo1 Collagen VI is an extracellular matrix protein that form s a microfilamentous network in skeletal muscles and other organs1"3. Inherited mutations in genes encoding collagen VI in humans cause two muscle diseases, Bethlem myopathy and Ullrich congenital muscular dystrophy4'5. We previously generated collagen V l-deficient (Coi6a1~'~) mice and showed that they have a muscle phenotype that strongly resembles Bethlem myopathy6. The pathophysiological defects and mechanisms leading to the myopathic disorder were not known. Here we show that Col6a1~!~ muscles have a loss of contractile strength associated with ultrastructural alterations of sarcoplasmic reticulum (SR) and mitochondria and spontaneous apoptosis. W e found a latent mitochondrial dysfunction in myofibers of Col6a 1~!~ mice on incubation w ith the selective F, F0 -ATPase inhibitor oligomycin, which caused mitochondrial depolarization, Ca2+ deregulation and increased apoptosis. These defects were reversible, as they could be normalized by plating Col6a 1~f~ myofibers on collagen VI or by addition of cyclosporin A (CsA), the inhibitor of mitochondrial permeability transition pore (PTP). Treatment of Co/fiaT"7" mice with CsA rescued the muscle ultrastructural defects and markedly decreased the number of apoptotic nuclei in vivo. These findings indicate that collagen VI myopathies have an unexpected mitochondrial pathogenesis that could be exploited for therapeutic intervention. Mechanical measurements showed that isolated Col6ai"!" muscles had less tension development (Fig. 1). Loss of contractile strength was most prominent in diaphragm, which is the most affected muscle0, but was also detected in flexor digitorum brevis (FDB) and other hindlimb muscles (data not shown). In diaphragm strips. Isometric tetanic ten sion was significantly lower (Fig. la ). Twitch tension was even more reduced and relaxation time was prolonged (Fig. Ib -d ), suggesting an abnormal SR function with less Ca-+ released on stimulation or less effective Ca2+ uptake during relaxation7. The decrease in tension development was not due to variations in fiber excitability (Supplementary Table 1 online) or to changes in fiber type composi tion (data not shown). The decreased contractile strength of Col6al muscles was confirmed by mechanical measurements on single skinned fibers during maximal activation (Fig. le). We detected struc tural alterations and irregular shape in about one-third o f the skinned Co/6fll"/_ fibers (Fig. If), These partially damaged fibers were never found in wild-type samples and developed significantly less tension than apparently healthy Col6al fibers (Fig. Ig). Taken together with the presence o f Evans blue-positive fibers6, this finding suggests that collagen VI- deficient muscles contain differently affected fibers that might represent various steps o f the dystrophic process. Electron microscopic analysis identified ultrastructural defects in various Col6al~!~ muscles. About 30% of fibers had structural alter ations o f mitochondria and SR (Fig. 2b). Mitochondria had abnormal cristae with tubular shape and altered matrix density associated with the presence o f dense bodies (Fig. 2d,e). We observed marked dilations in SR, especially' at the level of triadic system (Fig. 2g), but sarcomeres, sarcolemma and basal lamina appeared normal. We detected myonuciel with the typical hallmarks of apoptosis In Col6al"!" fibers with organelle alterations (Fig. 2i). None o f these defects were present in wild-type muscles (Fig. 2a,c,f,h). We investigated mitochondrial function in cultured FDB myofibers. Mitochondrial transmembrane potential ( 0 p ) was monitored by tetramethyirhodamine methyl ester (TMRM), a fluorescent probe that accumulates in polarized mitochondria and is released when (3|/n, decreases8. When measured for 1 h in standard conditions, the absolute TMRM fluorescence showed no significant differences between Co/baf and wild-type fibers, indicating that 0 )% levels were the same. But a mitochondrial dysfunction of Col6al"6" fibers could have been masked by the ATP synthase operating in a reverse mode, maintaining Qpn, during the experiment9. To test this hypothesis, we incubated the fibers with oligomycin, an inhibitor o f mitochondrial F]Fn-A.TP syn thase that Is widely used to investigate mitochondrial function in cul tured cells9"11. As expected, addition of oligomycin to wild-type fibers did not cause immediate 0.)im changes, and mitochondrial depolariza tion proceeded at slow rates even after extensive incubation (Fig. 3a). We observed fast mitochondrial depolarization in wild-type fibers only after addition of the protonophore carbonylcyanide-p-trifluoromethoxyphenyl hydrazone (FCCP), In contrast, addition of oligomycin to the Chiba / 1 fibers quickly caused marked mitochondrial depolarization (Fig. 3a,b). Depolarization was fast, proceeding at a rate Departments of 'Histology, Microbiology & Medical Biotechnologies and 2Biomedical Sciences, University of Padova, 35121 Padova, Italy. 3IT0i-CNR OR, 40136 Bologna, Italy. ''Departments of Human Anatomy & Physiology, University of Padova, 35121 Padova, Italy. bOrthopedic institute ' Rizzoli'. 40136 Bologna, Italy. These authors contributed equally to this work. Correspondence should be addressed to P. Bonaldo (bonaldo@bio.umpd.it) or P. Bernard! (bernardl@blo.unipd.lt). Published online 23 November 2003; doi:10.1038/ngl270 NATURE GENETICS VOLUME 35 j NUMBER 4 j DECEMBER 2003 3 67 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039346-00001  20 03 N a tu re P u blish in g G ro u p http://www.nature.eom/i LETTERS a m Wild-type M CoiSal-'- T b m Wild-type M CoiSal-'- E 100- 75- ,.,x,, t c 0.4- W wild-type M CoiSal-'- d 150- 1 0.3- A* yw?f 110000- Jl H Wilti-type ggj C oiS al-'- j j I i Figure 1 Loss of contractile strength in C o lS a l" muscles. Tension development in Half relaxation Time to time peak fused tetanus (a) or twitch (b), twitch/tetanus ratio (c) and time to half relaxation and time to peak tension (d) of intact diaphragm strips (DI A. n = 15 muscles for g each genotype), (e) Tension development during direct maximal Caz+ activation of skinned fibers Isolated from diaphragm (DIA, n ---- 32 fibers for each genotype) and filili? gastrocnemius (GCN, n = 27 fibers for each genotype), if) Microscope view of skinned fibers from C o l S a l gastrocnemius muscle. Top, a partially damaged but still contracting fiber with Irregular shape; bottom, a healthy-looking fiber. Scale bar, 20 com. (g) Tension development by single skinned fibers of C oiS ar'- ________ ,.KS ,s,. , w ' ' '' gastrocnemius muscles. Apparently healthy ColSal-'- fibers ( n --- 18), hatched bars; partially damaged Coi6al~*~fibers (n = 9), filled bars; wild-type fibers, open miiHWffifi bars. Cross-sectional area and sarcomere length was not significantly different between wild-type and CoiSal '''fibers. Data represent mean s.e.m. *P < 0.05, p fft **P < 0,01 for Col6al~y~compared with wild-type fibers; #P< 0.01 for partially damaged C oiS al-'-fibers compared with apparently healthy C oiS al-'-fibers. fjj| Wild-type CoiS al-1- comparable to that observed, in wild-type fibers treated with FCCP, indi cating that Col6al ~mitochondria were very permeable to protons. Under various pathological conditions, mitochondria undergo a large increase in inner membrane permeability due to opening of the FTP, This process depends on mitochondrial Ca2+ accumulation and leads to mitochondrial depolarization and other modifications that may trigger apoptosis12. Treatment with the FTP inhibitor CsA (Fig. 3c) or with the membrane-permeant Ca2+ chelator BAPTA-AM (Fig. 3d) greatly improved the 0 |/ m response to oligomycin in C ciaV /~ fibers, but cyclosporin H (CsH ), an analog that does not inhibit FTP13, was ineffective (Fig. 3c), The 2ij/mchanges in Col6al~^~ fibers also nor- malized when we plated the fibers on collagen VI, indicating that the abnormal mitochondrial response to oligomycin is the result of a reversible alteration (Fig. 3d). These experiments indicate that the mitochondrial dysfunction unmasked, by oligomycin in Col6al~^~ fibers involves FTP opening. Increased PTP opening may alter the dynamics o f intracellular Ca2+ fluxes14,15. Therefore, we monitored the resting [Ca2+] o f FDB fibers with Fura-2 (ref, 16; Fig. 4). Basal ICaz* ] c levels in wild-type and C oiSal-'- fibers were not significantly different when measured in standard conditions (data not shown) or in the presence o f 20 mV! extracellula r Ca2+ (Fig. 4a). Addition of oligomycin caused a marked Figure 2 Ultrastructural defects In collagen VI--deficient muscles. Electron micrographs of diaphragm and FDB muscles from wild-type (a,c,f,ii) and CoiSal~'~ (li,d,e,g,i) mice. (a,b) Low-power view of FDB longitudinal sections showing alterations in SR (arrows) and mitochondria (arrowheads) in CoiSal-1- fibers. Myofibrillar network is comparable in C o iS al1- and wild-type mice, (c--e) High-power view of diaphragm transverse sections showing abnormal mitochondria (IVlit) with altered crlstae and dense bodies (white arrow) in ColSal-'- fibers. Sarcolemma and basal lamina appear norma! (arrowheads and inset In e). (f,g) High-power view of the triadic system showing dilation of the terminal cisternae of SR and norma! T-tubules (arrowheads) in ColSal.-1- fibers. Swelling Is visible in some C o l S a l mitochondria (asterisk). (h,i) Peripheral chromatin condensation and irregular shape characteristic of apoptosis are detected In ColSal-1- myonuclei (Nu). Scale bars; a,b,h,i, 1 ocm; c-g, 0.5 am. 368 VOLUME 35 j NUMBER 4 | DECEMBER 2003 NATURE GENETICS Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039346-00002  20 03 N a tu re P u blish in g G ro u p http://www.nature.eom/i g -- Wild-type J 1.0 - 0.8 .S - -- Wild-type GsH -- Wild-type CsA -- Coia t '- CsH -- C ola'n- CsA FCCP LETTERS -- WiicMype -- Coi6aT'-- Col6a1-'- Coivi -- Co!6a1-- BA Time (min) Time (rnin) Figure 3 Altered mitochondria! response to ollgomycin in collagen VI--deficient myofibers. FD8 muscle fibers were loaded with TiVIRM {20 nM) for 10 min at 37 C. TMRM accumulates In mitochondria that maintain OX);,,,. Ollgomycin (Oim, 5 xM) and the protonophore FCCP {5 xM) were added at the indicated time points, (a) Measurement of in Col6al"1- and wild-type fibers incubated with oiigomycin. (b) Statistical analysis of the time required for a 20% decrease of the TMRM signal after addition of oiigomycin in C ol6al^ {five experiments) and wild-type (four experiments) fibers, (c) Effect of the PIP inhibitor CsA (2 xM). CsH {4 aM), an analog that does not Inhibit PTP, was used as a control. CsA or CsH was added at f = 0. (d) Effect of the intracellular Ca2+ chelator BAPTA-AM arid of collagen VI. Where indicated, BAPTA-AM {BA, 5 <xM) was added 15 min before the experiment or fibers were cultured on a substrate of collagen V! {ColVI) Instead of laminin. Neither treatment changed the response of wild-type fibers (data not shown). Data represent mean s.e.m., n= 2 -6 fibers for each time point and all results were confirmed in two or more independent experiments. *P < 0,01 for Col6al"/" compared with wild-type fibers. increase o f [Ca2+jc in C o l 6 a l but not wild-type fibers (Fig. 4b,c). As for mitochondrial depolarization, the [Ca2+]c response o f Col6al~!~ fibers was normalized by plating on collagen VI (Fig. 4b). To investigate the mechanisms o f oligomycin-dependent !C az+j c increase, we incubated C o l 6 a l myofibers with EG TA and various Ca2+ channel antagonists (Fig. 4d). The [Ca2+]c increase was still detected in nominally Ca2 :-free buffer or in the presence of sarcolemmal voltage-dependent Ca2^ channel antagonists, such as vera pamil, but it was completely prevented by dantrolene, a specific inhibitor o f the SR Ca24 release channels17 (Fig- 4d), CsA, but not CsH, delayed and reduced the oligomycin-dependent [Ca2+],. increase (data not shown). These data indicate that SR and PTP are involved in the abnormal [Ca24],. response of C oteal-1~ fibers to oiigomycin, in keeping with the mutual regulatory role o f mitochon dria and SR for Ca24 handling in excitation-contraction coupling18. We evaluated the occurrence of DNA strand breaks and detected TUNEL-positive (apoptotic) nuclei in both Col6al~^~ muscles and cul tured C o l 6 a l FDB fibers, but not in the corresponding wild-type samples (Fig. 5a,b). Col6al~l~ fibers had approximately seven times more apoptotic nuclei than wild-type samples (Fig. 5c). Addition of oi igomyci n increased the fraction o f apoptotic nuclei and caused a sig nificant decrease o f the total number o f myonuciei in Col6al~'~ fibers; both events were prevented by CsA (Fig. 5c,d and Supplem entary Table 2 online). As adjacent fiber segments are controlled by individ ual nuclei19, progressive depletion o f C o l S a l myonuciei may con tribute to muscle fiber loss and dystrophic processes. To test whether PTP has a crucial role in the pathogenesis o f the myopathy in vivo, we subjected wild-type and Col6al~'~ mice to intraperitoneal injection o f 5 m g CsA per kg body weight every 12 h. We chose this dose and frequency o f CsA treatment on the basis o f in vivo experiments in which PTP inhibition was estimated from the calcium retention capacity o f liver mitochondria20 (M. E. Soriano and P. Bernardi, unpublished data). After 4 d o f CsA treatment, we observed, substantially fewer apoptotic nuclei in both diaphragm sec- Figure 4 Altered [Ca2+]c response to oiigomycin in collagen VI--deficient myofi tiers. FDB muscle fibers were loaded with Fura-2/AM (5 xM) for 20 min at 37 C'C. The ratio of emission signals alter excitation at 340 rim and 380 nm (Fura-2 ratio) increases with Ca-- levels. Ollgomycin {01m, 5 xM) and ionomycin (Inm, 5 xM) were added where indicated, (a) Measurement of [Ca2+jc in Co/oak4 and wild-type fibers using oiigomycin-free Tyrode buffer supplemented with 20 mM CaCIp. (b) Measurement of !Caz+]c in C o l S a l and wild-type fibers incubated with ollgomycin and effect of collagen V! addition. Where indicated, fibers were cultured on a substrate of collagen VI (ColVI) instead of laminin. Collagen VI addition had no effect on wild-type fibers (data not shown), (c) Statistical analysis of the time required for a 50% increase of Fura-2 signal after addition of oiigomycin in CoiSal~l~ (nine experiments! and wild-type (six experiments) fibers, id) Effect of EGTA, verapamil arid dantrolene. Verapamil (Ver, 150 xM), dantrolene (Dan, 100 xM) or EGTA (5 mM) was added at t = 0 in standard Tyrode buffer (Ca) or in Tyrode buffer without CaCI2 (no Ca). CaCI2 (Ca2+, 10 mM) was added at the end of the experiment as a control. Data represent mean s.e.m., n = 2-6 fibers for each time point and all results were confirmed in two or more independent experiments. *P < 0,01 for Col6al~/~compared with wild-type fibers. NATURE GENETICS VOLUME 35 j NUMBER 4 j DECEMBER 2003 .3 69 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039346-00003 LETTERS 20 03 N a tu re P u blish in g G ro u p http://www.nature.com/naturegenetics Figure 5 Nuclear apoptosis in histological sections and cultured fibers from Co!6al"'" muscles, TUNEL staining of diaphragm (a) and of fiber cultures from FDB muscle (b). Hoechst-stained sections are shown below the corresponding TUNEL pictures. Several TUNEL-posItlve nuclei, labeled in red, are present In C o l B a l but not in wild-type fibers, (c) Frequency of TUNEL-positive nuclei in cultured fibers from ColBal"'" and wild-type FDB muscles. Fibers were cultured for the indicated times in Tyrode buffer with no additions {-), in the presence of 5 adV' oligomycin (Olm) or in the presence of 5 ocM oligomycin and 2 ocM CsA (Olm + CsA). Data represent mean s.e.rn. of three independent experiments, n = 40 -1 3 0 fibers, (d) Hoechst staining showing a prominent example of rnyonuclear loss in ColBal"'" fibers after oligomycin addition. Scale bars: a, 50 mm; b,d, 10 mm. *P < 0.01 for CoiBar'" compared with 'wild-type fibers; #P< 0,01 for ColBal"1"(Olm + CsA) compared with ColBal '" (Olm) fibers. WM-type (lm) ColSal (Olm) tions (Fig. 6a,b) and isolated FDB fibers o f ColBal"'" mice (Fig. 6c). Occurrence o f PTP inhibition in vivo was confirmed by measure ments o f the calcium retention capacity o f isolated mitochondria, which was always higher in mice treated with CsA (data not shown). We also treated mice with FK506, which inhibits calcineurin but not FTP1-. FK506 increased rather than decreased the incidence ofapoptotic nuclei in both wild-type and ColBal"''""' mice (Fig. 6a--c). When compared with diaphragm fibers from ColBal"1" mice treated with vehicle (Fig. 6d -f), those from ColBal 1 mice treated with CsA recovered considerably from the ultrastructural defects, as both mito chondrial and SR abnormalities disappeared (Fig, 6g--i). Rescue was virtually complete; we detected a small residual SR dilation in 1 o f 300 fibers examined for diaphragm and 9 o f 300 fibers for FDB muscle. Our results identify an unexpected mitochondrial pathogenesis in collagen VI deficiency. The key issue is the mechanism leading to mitochondrial defects and apoptosis. Ca2+ overload has been pro posed as an early event in certain muscular dystrophies21, and one possibility is that slight increases of sarcolemmal Ca2: influx in colla gen VI--deficient fibers might result in intracellular C a " overload, eventually causing PTP opening and apoptosis. Our results suggest that this may not be the case, because [Ca2+]r levels were indistin guishable in wild-type and ColBal"1" fibers, even at 20 mM extracellu lar Ca2+, and because addition of dantrolene or CsA, but not verapamil, ameliorated oligomycin-dependent Ca2+ deregulation. Another possibility is suggested by the interaction of collagen VI with integrins and other cell surface receptors2,2. Collagen VI prevents a Vehicle CsA Wild-type FK506 Co/eaD- Vehicle -- e ^ lilil "Wv. CsA p p h li T 370 Sierra Club v. EPA 18cv3472 NDCA SR ** III* Figure 8 In vivo administration of CsA prevents nuclear apoptosis and ultrastructural defects in coiiagen VI--deficient muscles, (a) TUNEL staining of diaphragm from wild-type and ColBal"1" mice treated with CsA, FK506 or vehicle alone. (b,ej Quantification of TUNEL-posItlve nuclei in diaphragm sections (b) and cultured FDB rnyofitiers (c) from wild-type and C o l B a l mice treated with CsA, FK506 or vehicle, (d-i) Electron micrographs of diaphragm transverse sections from ColBal"'" mice treated with CsA (lower panels) or vehicle (upper panels). Low-power view (d,gj shows a marked recovery of the ultrastructural defects in mice treated with CsA. Mitochondria (Mit) and SR of ColBal '" mice treated with CsA are morphologically normal and similar to wild-type (compare h,i with Fig- 2c,f). Extensive alterations in mitochondria and SR are detected in ColBal"1" mice treated with vehicle, similar to untreated ColBal '" mice (compare e,f with Fig.2d,g), Scale bars: a, 50 am; d-i, 0,5 <xm. *P < 0,01 for ColBal"'" mice treated with CsA compared with ColBal"'" mice treated with vehicle; NS, not significant (P> 0.05). VOLUME 35 j NUMBER 4 | DECEMBER 2003 NATURE GENETICS Attachments Prod 1 ED 002061 00039346-00004 LETTERS 20 03 N a tu re P u blish in g G ro u p http://www.nature.com/naturegenetics apoptosis of fibroblasts when integrin-matrix interactions are p er turbed or serum is withdrawn22'---'. Notably, addition o f collagen VI to fibroblasts causes downregulation of Bax2J, a proapoptotic protein that triggers FTP opening12. Lack of collagen VI might cause m ito chondrial dysfunction and increased PTP opening through an abnor mal engagement of integrins, in keeping with the recent finding that integrin-mediated signaling regulates mitochondrial function2'11, increased PTP opening would alter Ca2+ handling by SR35; cause Caz^ deregulation with further increase o f the PTP open time and thus set in motion a self-amplifying loop eventually leading to structural defects o f both mitochondria and SR. Our finding that the structural alterations and apoptotic defects of Coi6al"'" muscles in vivo can be rescued by CsA through a mitochondrial mechanism may open the way to a pharmacological treatment for Rethlem myopathy and Ullrich congenital muscular dystrophy. METHODS Mice, We backcrossed C ol6al+l'" mice in a mixed 129/Sv x C57BL/6J b ac k ground6 into C57BL/6J strain for eight generations. We obtained data by comparing sex-matched, 8-32-wk-old Col6ai~'~ mice with the correspond ing wild-type littermates. Mouse procedures were approved by the ethics committee of the University o f Padova. D rug treatm ent in vivo. We injected wild-type and Col6al~'~ mice intraperi(oneally every 12 h for 4 d with CsA (Novartis, 5 mg per kg body weight; n --8), FK506 (Fujisawa, 2.3 m g per kg body weight; n --4) or vehicle (olive oil; n --8). Muscle m echanics and electrophysiology. We dissected FDB, extensor digitorum longus and diaphragm strips from wild-type in -- 20) and Col6al~,'~ (n ----15) mice and recorded twitches and tetani as described26. We isolated single myofihers from diaphragm and gastrocnemius muscles, chemically skinned them as described^ and measured the tension during maximal isometric acti vation (pCa --4.5, T --20 C , sarcomere length --2.75 octn). We recorded resting and action membrane potentials intracellularly as described26. Transm ission electron microscopy. We longitudinally stretched diaphragm and FDB muscles from w ild-type in ----- 7) and Col6al 1 in ----- 7) mice on wax, fixed them with 2.5% glutaraldehyde in phosphate buffer 0.1 M (pH 7.4) and embed ded them in Epon E812 resin. We observed ultrathin sections, stained with uranyl acetate and lead citrate, in a Philips EM400 electron microscope at 100 kV. Isolation and culture o f skeletal myofibers. We isolated fibers from FDB m us cles o f wild-type (n --- 33) and Col6ai~'~ In -----28) mice as described8. We plated intact myofibers onto glass coverslips coated with mouse EHS laminin (3 ocg era-2) or collagen VI (15 ceg cm-2) and cultured them in DulbecccKs modified Eagle medium containing 10% fetal calf serum for 16-40 h before starting the experiment. Only apparently' healthy fibers (i.e.> fibers without structural alter ations and not stained by Evans blue) adhered to the coverslips. Q.prn and [Ca2+jc assay. We placed FDB my'ofiber cultures in i ml glucose- free Tyrode buffer and loaded them with appropriate probes. We used TMRM (Molecular Probes, 20 nM) to monitor 0.jJm as described8 and Fura-2/AM (Sigma, 5 ocM) to m onitor [Ca2l ]c according to established procedures10. We used oligomycin, CsA, FCCP, ionomycin, dantrolene, verapamil (all Sigma), CsH (a gift from Novartis) or BAPTA-AM (Molecular Probes) at the indicated times and concentrations, im aging was done with a Zeiss Axiovert epi-fluorescence microscope. We analyzed data with the MetaEluor Imaging software. TIJNEL. We prepared 7-cctn-thick sections o f muscles from wild-type (n --6) and Col6ai~'~ In -----6) nice after formalin fixation and paraffin embedding. We fixed FDB fiber cultures in 50% acetone/50% methanol. TUN EL was done using the ApopTag in situ apoptosis detection kit (Intergen). We stained sam ples with peroxidase-diaminobenzidine to detect TUNEL-positive nuclei and counterstained them with Hoechst 33258 (Sigma) to mark all nuclei. We deter mined the numbers o f total and TUNEL-positive nuclei in randomly selected fibers using a Zeiss Axioplan microscope. Statistical analysis. Data are expressed as mean s.e.m. We analyzed data with the unpaired Student's t-test. Values with P < 0.0:5 were considered significant. Note: Supplementary information is available on the Nature Genetics website. ACKNOWLEDG MENTS We thank T. Pozzan and R. Betto for advice about C a' measurements and fiber cultures, M. Ghidotti for animal husbandry arid A. Colombaiti for critical reading of manuscript. This work was supported by grants from Telethon, Italian MI UR, European Project Myocluster arid Italian Health Ministry. COMPETLNG INTERESTS STATEMENT The authors declare that they have no competing financial interests. Received 24 September; accepted 27 October 2003 Published online at http://www.nature.com/naturegenetics/ 1. Keene, D.R., Engvail, E. & Glarville, R.W. URestructure of type Vi co Hagen in human skin and cartilage suggests an anchoring function for this filamentous network. J. Cell Biol. 107, 1 99 5 -2 00 6 (1988). 2. Bonaldo, P., Russo, V., Bucciotti, F., Doiiana, R. & Colombatti, A. Structural and func tional features of the 3 chain Indicate a bridging role tor chicken collagen VI in con nective tissues. Biochemistry 29, 1 24 5 -1 25 4 (1990). 3. Tim pi, R. & Chu, M,L, Microfibrillar collagen type VI. n Extracellular Matrix Assembly and Structure (eds. Yurchenco, P.D., Birki D.E. & Mecham, R.P.) 208 -2 4 2 (Academic, Orlando, 1994). 4. Jobsis, G.J. e ta !. Type VI collagen mutations in Bethlem myopathy, an autosomal dominant myopathy with contractures. N a t Genet. 14, 1 1 3 -1 1 5 (1996). 5. Camacho Vanegas, 0. et a i Ullrich scieroatonic muscular dystrophy is caused by reces sive mutations in collagen type VI. Proc. Natl. Acad. Sci. USA 38, 7516-7521 (2001). 6. Bonaido, P. et a i Collagen VI deficiency induces early onset myopathy In the mouse: an animal model for Bethlem myopathy. Hum. Moi. Genet. 7, 2 1 3 5 -2 1 4 0 (1998). 7. Plant, D.R, & Lynch, G.S. Excitation-contraction coupling and sarcoplasmic reticu lum function in mechanically skinned fibres from fast skeletal muscles of aged mice. J. Physiol. 543, 1 59 -1 7 5 (2002). 8. Irwin, W. et a i Bupivacaine myotoxicity is mediated by mitochondria. J. Biol. Chem. 277,12221-12227(2002). 9. Nicholls, D.G. & Ward, M.W. Mitochondrial membrane potential and neuronal gluta mate excltotoxicity: mortality and millivolts. Trends Neurosci. 23, 155 -1 7 4 (2000). TO. D'hahan, N. et a i Pharmacological plasticity of cardiac ATP-sensitlve potassium channels toward diazoxide revealed by ADP. Proc. N a ti Acad. Sci. USA 96, 12152-12157 (1999). 11. Gugliucci, A. et a i Mitochondria are direct targets of the lipoxygenase inhibitor MK885. A strategy for cell killing by combined treatment with MK885 and cyclooxy genase inhibitors'. J. Biol. Chem'. 211, 3 1 7 8 9 -3 1 7 9 5 (2002). 12. Bernard!, P., Petronilh, V., Di Lisa, F. & Forte, M. A mitochondria! perspective on cell death. Trends Biochern. Sci. 26, 112 -1 1 7 (2001). 13. Griffiths, J . & Halestrap, A.P. Further evidence that cyclosporin A protects m ito chondria from calcium overload by inhibiting a matrix peptidyi-proiyl as-traris iso- merase. Biochem. J. 274, 6 1 1 -6 1 4 (1991). 14. Bowser, D.N., Minamikawa, T., Nagiev, P. & Williams, D.A. Role of mitochondria in calcium regulation of spontaneously contracting cardiac muscle cells. Biophys. J. 75, 2004-2014 (1998). 15. Bowser, D.N., Petrou, S., Panchal, R.G., Smart, M.L. & Williams, D.A. Release of mitochondrial Ca2*' via the permeability transition activates endoplasmic reticulum Ca2+ uptake. FASEBJ. 16, 1 10 5 -1 10 7 (2002). 15. Grynkiewicz, G., Poenie, M. & Tsien, R.Y. A new generation of Ca2+ indicators with greatly improved fluorescent properties. J. Biol. Chem. 260, 3 4 4 0 -3 4 5 0 (1985). 17. Zhao, F,. Li, P,. Chri, S.R., Louis, C.F, & Pruen, B.R. Dantrolene inhibition of ryan odine receptor Ca2'- release channels. J. Biol. Chem. 276, i 3 8 1 0 - 13816 (2001). 18. Robert, V. et al. Alteration in calcium handling at the subcellular level in max myotubes. J. Biol. Chem. 276, 4 54 7 -4 55 1 (2001). 19. Tews, D.S. Apoptosis and muscie fibre loss in neuromuscular disorders. Neuromuscui Disord. 12, 5 1 3 -5 2 2 (2002). 20. Kiohn, R.C. et a i Early resistance to cel! death and to onset of the mitochondria! per meability transition during hepatocarcinogenesis with 2-acetyiamlnofiuorene. Proc. N a ti Acad. Sci. USA 100, 1 0 0 1 4 -10 0 1 9 (2003). 21. Blake, D.J, Weir, A., Newey, S.E. & Davies, K.E. Function and genetics of dystrophin and dystrophin-related proteins in muscie. Physioi. Rev. 82, 2 9 1 -3 2 9 (2002), 22. Howell, SJ. & Doane, K.J. Type Vi collagen increases cel! survival and prevents anti- p i integrin-mediated apoptosis. Exo. Cell Res. 2 4 1 ,2 3 0 -2 4 1 (1998). 23. Rhl, M. et ai. Soluble collagen VI drives serum-starved fibroblasts through S phase and prevents apoptosis via down-regulation of Bax. J. Biol. Chem. 274, 34361-34368 (1999). 24. Werner, E. & Werb, Z. iritegrins engage mitochondrial function for signal transduction by a mechanism dependent on Rhc GTPases. J. Ceil Biol. 158, 3 5 7 -3 5 8 (2002). 25. Rossi, R,, Bottineili, R., Sorrentino, V. & Reggiani, C. Response to caffeine and ryan odine receptor isoforrns in mouse skeletal muscles. Am. J. Physioi. 281, C595-C602 ( 2 0 0 1 ). 25. Germinarlo, E. e! ai. Early changes of type 2B fibers after denervation of rat EDI skeletal muscle. J. A p p i Physioi'92, 2 04 5 -2 05 2 (2002). NATURE GENETICS VOLUME 35 j NUMBER 4 j DECEMBER 2003 3 71 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039346-00005