Ca2� dysregulation in Ryr1I4895T/wt mice causescongenital myopathy with progressive formationof minicores, cores, and nemaline rodsElena Zvaritcha, Natasha Kraevaa, Eric Bombardierb, Robert A. McCloyb, Frederic Depreuxc, Douglas Holmyardd,Alexander Kraeva, Christine E. Seidmanc, J. G. Seidmanc, A. Russell Tuplingb, and David H. MacLennana,1

aBanting and Best Department of Medical Research, Charles H. Best Institute, University of Toronto, Toronto, Ontario, Canada M5G 1L6; bDepartment ofKinesiology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1; cDepartment of Genetics and Howard Hughes Medical Institute, Harvard MedicalSchool, Boston, MA 02115; and dAdvanced Bioimaging Centre, Mount Sinai Hospital, Toronto, Ontario, Canada M5G 1X5

Contributed by David H. MacLennan, October 22, 2009 (sent for review October 15, 2009)

Ryr1I4895T/wt (IT/�) mice express a knockin mutation correspondingto the human I4898T EC-uncoupling mutation in the type 1 ryan-odine receptor/Ca2� release channel (RyR1), which causes a severeform of central core disease (CCD). IT/� mice exhibit a slowlyprogressive congenital myopathy, with neonatal respiratorystress, skeletal muscle weakness, impaired mobility, dorsal kypho-sis, and hind limb paralysis. Lesions observed in myofibers fromdiseased mice undergo age-dependent transformation from mini-cores to cores and nemaline rods. Early ultrastructural abnormal-ities include sarcomeric misalignment, Z-line streaming, focal lossof cross-striations, and myofibrillar splitting and intermingling thatmay arise from defective myofibrillogenesis. However, manifesta-tion of the disease phenotype is highly variable on a Sv129genomic background. Quantitative RT-PCR shows an equimolarratio of WT and mutant Ryr1 transcripts within IT/� myofibers andtotal RyR1 protein expression levels are normal. We propose aunifying theory in which the cause of core formation lies infunctional heterogeneity among RyR1 tetramers. Random combi-nations of normal and either leaky or EC-uncoupled RyR subunitswould lead to spatial differences in Ca2� transients; the resultingheterogeneity of contraction among myofibrils would lead tofocal, irreversible tearing and shearing, which would, over time,enlarge to form minicores, cores, and nemaline rods. The IT/�mouse line is proposed to be a valid model of RyR1-relatedcongenital myopathy, offering high potential for elucidation of thepathogenesis of skeletal muscle disorders arising from impaired ECcoupling.

calcium � central core disease � multiminicore disease �nemaline rod myopathy � ryanodine receptor

Congenital myopathies are a heterogeneous group of inher-ited skeletal muscle weakness disorders caused by mutations

in structural, contractile, or regulatory muscle proteins (1, 2).Early clinical symptoms include neonatal hypotonia, generalizedmuscle weakness, respiratory distress, and skeletal deformitiessuch as dislocated hips, pes cavus, and kyphoscoliosis. Thepenetrance of the disease is highly variable, but its course istypically slow or nonprogressive. Central core disease (CCD)and multiminicore disease (MmD), often referred to as coremyopathies, and nemaline myopathies (NM) are the most com-mon congenital myopathies. Their differential diagnosis is basedon histopathological analyses of skeletal muscle biopsies thatreveal cores, minicores, and nemaline rods (1–3). However,some cases of myopathy present with mixed histology so thatcores are seen in combination with minicores or with nemalinerods (4–8). The etiology and pathogenesis of lesion formation isunclear and their presence does not always correlate with theseverity of the clinical phenotype (5, 9, 10).

Many congenital myopathies have been linked to mutations inRYR1, encoding the type 1 ryanodine receptor/Ca2� releasechannel (RyR1), a key protein in excitation-contraction (EC)

coupling in skeletal muscle. RyR1 is a homotetramer of 565 kDasubunits, each containing 5,038 aa residues. In response tosarcolemmal depolarization, RyR1 channels are activated torelease Ca2� from the sarcoplasmic reticulum (SR), triggeringmuscle contraction. Mutations in RYR1 have been associatedwith malignant hyperthermia (MH) [Online Mendelian Inheri-tance in Man (OMIM) database no. 145600] (11), CCD (OMIMno. 117000) (11), MmD with external ophthalmoplegia (OMIMno. 255320) (12), congenital myopathy with cores and rods (6, 7),central nuclear myopathy (13), and neuromuscular disease withuniform type 1 fibers (14). For all of these disorders, Ca2�

dysregulation is a likely primary cause and a common etiology ispossible.

Insights into the etiology of RyR1-related myopathies havecome from in vitro functional studies of mutant RyR1 expressedtransiently in heterologous and homologous cell culture systems(15–16). CCD-associated RyR1 mutations impair Ca2� releasechannel function by rendering the channel either hypersensitiveto stimuli (leaky) or unresponsive to activating ligands anddepolarization of the sarcolemma (EC uncoupled). It is not yetclear how functionally different leaky and EC-uncoupled RyR1mutant channels can induce similar skeletal muscle lesions andskeletal muscle weakness.

Answers to question of the etiology and pathogenesis ofRyR1-related disorders are being sought through in vivo assess-ment of pathological changes produced in knockin (KI) mouselines carrying RyR1 mutations. These include two mouse lineswith leaky MH/CCD mutations R163C (17) and Y522S (18). Wegenerated a KI mouse line (19) carrying the EC-uncouplingRyR1 mutation, I4895T, which corresponds to one of the mostcommon CCD mutations in human RYR1, I4898T (11). Inhumans, occurrence of the heterozygous mutation is oftenassociated with a severe clinical phenotype, but penetrance isvariable (20). Ile-4898 is located in a highly conserved GGIG4899

motif, proposed to form the selectivity filter of the Ca2� releasechannel (21). In vitro functional studies have shown that allamino acid substitutions at position 4,898 reduce or disrupt Ca2�

release channel conductance (15, 16, 21). In particular, coex-pression of WT and I4898T mutant RyR1 cDNAs to mimic theheterozygous disease state reduced global Ca2� release by 60%(16). We reported (19) that homozygous Ryr1I4895T/I4895T (IT/IT)mice are born paralyzed in all skeletal muscles, are unable to

Author contributions: E.Z., A.K., C.E.S., J.G.S., A.R.T., and D.H.M. designed research; E.Z.,N.K., E.B., R.A.M., F.D., D.H., and A.R.T. performed research; E.Z. contributed new reagents/analytic tools; E.Z., N.K., A.K., C.E.S., J.G.S., A.R.T., and D.H.M. analyzed data; and E.Z., N.K.,A.K., C.E.S., J.G.S., A.R.T., and D.H.M. wrote the paper.

The authors declare no conflict of interest.

Freely available online through the PNAS open access option.

1To whom correspondence should be addressed. E-mail: david.maclennan@utoronto.ca.

This article contains supporting information online at www.pnas.org/cgi/content/full/0912126106/DCSupplemental.

www.pnas.org�cgi�doi�10.1073�pnas.0912126106 PNAS � December 22, 2009 � vol. 106 � no. 51 � 21813–21818

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021

breathe, and die perinatally. We also showed that the IT/ITmutation disrupts Ca2� release channel function without alter-ing the integrity of the RyR1 protein complex or of the su-pramolecular structure of the Ca2� release unit (CRU).

Here, we report our analyses of heterozygous Ryr1I4898T/wt

(IT/�) mice. We demonstrate that these mice exhibit a slowlyprogressive congenital myopathy with cores, minicores, androds. We conclude that the IT/� mouse line is a genetically andphenotypically valid model of an RyR1-related congenital my-opathy. A unifying theory is presented to explain how similarpathogenic phenotypes can arise from functionally differentRyR1 mutations.

ResultsIT/� Mice Exhibit a Variable Myopathic Phenotype. IT/� mice wereborn at the predicted Mendelian frequency, without apparentskeletal deformities, and survived with no neonatal lethality.Nevertheless, neonatal signs of myopathy were observed. Spe-cifically, IT/� mice (n � 24) were flaccid and cyanotic during thefirst minutes after delivery (Fig. S1 A), responded poorly tostimuli, and started breathing regularly 15–20 min after birth,compared to 5–7 min for WT littermates (n � 26). Progressionof the disease in IT/� mice was variable, ranging from mild tosevere. Impairment of mobility was first noted as weakness of thehindquarters at �6 months of age in some IT/�mice. By 10months, 24 (80%) out of 30 IT/� mice studied showed varyingdegrees of motor dysfunction (Fig. 1 A and B; Fig. S1 B–E). At�12 months, 2 males and 2 females (14%) suffered completehind limb paralysis, whereas 4 males and 2 females (20%)appeared asymptomatic and had a normal lifespan of �2 years.With age, IT/� mice developed dorsal kyphosis in the cervico-thoracic region, likely because of overuse of the forelimbs forlocomotion. At all ages, the size of IT/� mice was comparableto that of controls (Fig. S1 A and Fig. 1 A and B), but the averagebody weight of IT/� mice was �15% lower than WT. Forexample, at 10 months, IT/� mice weighed 27 � 3 g (n � 12) andWT mice weighed 32 � 2 g (n � 10). Gross pathologicalexamination did not reveal muscle wasting, but showed very lowfat deposits, which might account for the lower body weights ofIT/� mice.

Histopathology of IT/� Skeletal Muscles. Structural abnormalitiescharacteristic of congenital myopathy were sought for in hindlimb muscles of 6-weeks- to 24-month-old IT/� mice. Histolog-ical and EM analyses were carried out on animals that displayeda disease phenotype, which was recognizable by 8 months (seeabove). Increased fiber size variability, increased endomysialspacing, and mild fibrosis were observed in all muscle groups at�6–8 months, but were most prominent in soleus (Fig. S2 A andB). For this reason, we focused our attention on soleus muscle,which in 6-month-old IT/� mice displayed a high range of fiberdiameters, averaging 31 � 20 �m (n � 143), compared to 24 �6 �m (n � 128) in WT animals. Fiber type distribution, assessedby NADH-staining, did not differ between WT and IT/� micein all age groups (Fig. S2 C and D and Table S1). Type 1 fiberhypotrophy/atrophy, characteristic of core myopathies (3), couldbe detected in individual fibers as early as 6 weeks, whereas type2 fibers in young IT/� mice appeared enlarged. In 18-month-oldIT/� mice, both fiber types showed reduced diameters (TableS1), suggestive of generalized fiber atrophy. The extent of centralnucleation, indicative of skeletal muscle regeneration, did notdiffer between WT and IT/� mice of all age groups. Thisobservation is consistent with findings from core myopathypatients, who rarely exhibit central nucleation or active skeletalmuscle regeneration (22). A single case of RYR1-related centralnuclear myopathy (13) is apparently an exception.

Discrete foci of oxidative enzyme depletion, consistent withminicores, were observed in NADH-TR-stained longitudinal

sections of IT/� soleus fibers at �12 months of age (Fig. 1 C–E).The foci, 1–2 per 100 �m of myofiber length, were detected in�20% of both type 1 and type 2 fibers. They were predominantlyeccentric (Fig. 1C), small and rounded, with an average diameterof 10–20 �m. As in MmD (8, 23), the main axis of each minicore(Fig. 1D) was in transverse orientation to the main axis of thefiber. Larger areas of oxidative enzyme depletion, consistentwith cores, were observed in type 1 fibers of 20-month-old IT/�mice (Fig. 1 F and G and Fig. S3 A–C). These cores oftenoccupied a more central position and extended longitudinally formore than 50 �m. Outside of the minicore/core areas, theoxidative staining of the IT/� myofibers was uneven (Fig. 1 C–Gand Fig. S3 A–C). The cross-banding pattern, reflecting thenormal, highly-organized structure of the mitochondrial/sarcotubular network, was frequently distorted or absent and was

A

B

C D

E

F

G

Fig. 1. Morphological and histological abnormalities in IT/� mice. (A)Twelve-month-old WT and (B) IT/� female littermates. The IT/� mouse exhib-its dorsal kyphosis and hind limb paresis. Note the flattened posture andoutstretched hind limbs that fail to lift the hindquarters. (C–G) NADH-TR-stained longitudinal sections of IT/� soleus myofibers from 12-month- (C–E)and 20-month-old (F and G) mice showing minicores and cores. In C–E, arrowsshow minicores occupying eccentric (C), peripheral (E), and central (D) posi-tions. In D, the minicore (arrow) shows intense peripheral staining revealingthe distorted cross-striation of focally contracted myofibrils. In F and G, thearrows show cores extending longitudinally over multiple sarcomeres. (Scalebars, 10 mm.)

21814 � www.pnas.org�cgi�doi�10.1073�pnas.0912126106 Zvaritch et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021

substituted by prominent longitudinal streaks of oxidative en-zyme activity (Fig. 1 C, F, and G and Fig. S3 A–C). Whereasdepletion of oxidative enzyme activity represents loss of mito-chondria, increased intensity of oxidative staining peripheral tothe core reflects abnormal mitochondrial clustering. The pro-gressive formation of the patchy staining patterns outliningminicores and cores suggests an age-dependent expansion ofcore lesions in IT/� myofibers.

Examination of toluidine blue-stained muscle sections from6-month-old IT/� mice (Fig. 2A) revealed structural abnormal-ities consistent with minicore/core lesions. Minicore/core lesionswere identified as well-demarcated areas of spatial disruption ofsarcomeric and myofibrillar arrangement present within anindividual myofiber. There was no apparent disruption of cellmembrane integrity. They occurred with a frequency not greaterthan 1–2 per fiber, involved 5–40 consecutive sarcomeres, andwere observed in �14% of fibers. A linkage between theselesions and the minicores observed in NADH-TR staining (Fig.1 C–F) was established by similarities in frequency and location,which were predominantly peripheral and extended across ad-jacent sarcomeres in a transverse direction. The lesions often hada perinuclear location and were in close proximity to bloodvessels (Fig. 2 A and Inset), as described for minicores in coremyopathy patients (4). The lesions showed age-dependent ex-pansion (Table S2), so that by 18 months, they were present in�65% of fibers and extended over larger areas. Such lesions wererarely found in WT control myofibers (Table S2). Overall, theresemblance between the lesions in IT/� mice and minicore andcore lesions in core myopathy patients (4, 8, 22) was striking.

An unexpected finding in the muscles of aged 18-month-oldIT/� mice was the presence of rod-like inclusions, which wereclustered in areas apparently devoid of cross-striations andstreamed along the length of the myofibers (Fig. 2B). Theseinclusions were specific to IT/� muscle samples and were foundin �15–20% of soleus fibers in two out of five aged IT/� mice.

Electron Microscopy. Skeletal muscle fiber type was deduced fromultrastructural features specific for fast (type 2) and slow (type1) fibers: Z-line thickness, relative mitochondrial abundance,and M-line appearance (24). Ultrastructural abnormalities weredetected in both type 1 and type 2 myofibers of 6-week-old IT/�mice (Fig. 3 A and B). Fig. 3A shows gross structural defects in

myofibrillar organization in a type 2 soleus myofiber. Themyofibrils varied greatly in thickness and numerous sites ofsplitting and thinning were observed, suggesting possible defectsin myofibrillogenesis in IT/� mice. The focal abnormalitiesincluded sarcomeric shortening, insertion of an additional sar-comere, loss of sarcomeric register, and myofibrillar disorgani-zation with focal loss of cross-striation. Myofibrillar intermin-gling and the loss of intermyofibrillar space within compactedregions was a frequent feature. We propose that these com-pacted areas represent the initial stages of core formation.

A

B

Fig. 2. Toluidine blue-stained sections from 6-month- (A) and 18-month-old(B) mice. (A) A compact, well-demarcated minicore (arrow) in a single fiber,surrounded by fibers with normal cross-striation. The arrowhead designates ablood vessel. The Inset provides an enlarged image of the minicore. (B) Amyofiber featuring a central core area with a severe loss of cross striation.Arrows point to rod-like inclusions colocalizing within the core and runningthe length of the fiber. (Scale bars, 10 �m.)

A

B

C D

Fig. 3. Ultrastructural abnormalities in longitudinal sections of soleus myo-fibers from 6-week- (A and B) and 6-month-old (C and D) IT/� mice. (A) Afull-width section of a type 2 fiber showing myofibrillar splitting (arrows) andintermingling (bracket). The boxed area shows a focal loss of myofibrillarorganization over several sarcomeres. (B) A transitional area between anormally structured area and a compacted core area in a type 1 fiber. The corelesion shows Z-line streaming and focal loss of a Z-disk (arrowheads). Note thatintermyofibrillar spaces are reduced and mitochondria and SR are absent fromthis area (Upper Left diagonal), which is sharply delineated from a contiguous,normally structured region (Lower Right diagonal). Mitochondria in the un-affected area (asterisks) have a normal disposition and appearance. (C) A type2 fiber shows a well-demarcated lesion with gradual loss of sarcomeric orga-nization from a structured (Upper Right) to an unstructured area (Center).Double-headed arrows show the variability of sarcomeric lengths in the corearea. The arrowheads show how loss of sarcomeric register is transmitted toadjacent regions. (D) The central part of an structured core in a type 1 fiber.Z-lines (arrows) are wavy and disintegrating, intermyofibrillar space is re-duced, and mitochondria and SR are absent. [Scale bars, 5 �m (A), 2 �m (B andC), and 1 �m (D).]

Zvaritch et al. PNAS � December 22, 2009 � vol. 106 � no. 51 � 21815

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021

Fig. 3B features an area of transition between relativelynormal and compacted myofibrils in a type 1 soleus fiber. Incompacted areas, the myofibrillar bundles, with preserved sar-comeric banding, are closely apposed by lateral compaction sothat there is a loss of intermyofibrillar space. Focal myofibrillarcompactions are regarded as the ultrastructural correlate ofhistochemical cores in human CCD and MmD muscle biopsies(3, 9, 10). Fig. 3B shows extensive Z-line streaming and focal lossof Z-line material within the area of compacted myofibrils. Inthis area, there is loss of intermyofibrillar space, of SR, and ofmitochondria. The structure of the region contrasts with that ofan adjacent, normally organized area in the same myofiber, withregularly positioned mitochondria and well-defined intermyofi-brillar spacing.

The ultrastructural abnormalities that were first detected inmyofibers of 6-week-old, IT/� mice were more frequent andmore pronounced in 6-month-old mice (Fig. 3 C and D and TableS2). In both type 1 and type 2 fibers, we observed abrupttransitions from structured to unstructured areas within the corelesions. Fig. 3C illustrates a core lesion involving approximatelyfive sarcomeres in a few adjacent myofibrils in a type 2 soleusfiber. There is an insertion of two additional sarcomeres withinthe core area. The sarcomeres are out of register and theirshortening is highly irregular; some are overstretched, whileothers are shortened maximally by about twofold, with sarco-mere lengths of 0.76 �m vs. 1.7 �m in the unaffected area. Fig.3D shows the central part of a structured core area in a type 1soleus fiber. Although the sarcomeric pattern was relatively wellpreserved, the myofibrils were compacted and misaligned, in-termyofibrillar spaces were absent, and Z-lines were wavy anddisintegrated into small segments so that individual myofibrilscould no longer be discerned. The mitochondria were absentfrom most of the area.

EM imaging confirmed the presence of nemaline rods in 18-to 20-month-old IT/� mice. Fig. 4A shows extended areas of

myofibrillar misalignment involving the entire width of the type1 myofiber. In EM images, the inclusions observed in toluidineblue-stained sections (Fig. 2B) streamed throughout the lengthof the myofiber and aggregated in unstructured cores (Fig. 4 Aand B). The rods emanated from disintegrating Z-lines (Fig. 4B)and ranged in length from 0.2 to 2 �m; their lattice-like innerstructure had a periodicity of �13 nm by 17 nm (Fig. 4C). Theremnants of thin filaments extended linearly from both ends ofthe rods. Structural features of the rods in IT/� myofibers werestrikingly similar to those described in NM and CCD with rods(3–6).

Ryr1 mRNA Levels and Protein Expression. The relative abundanceof WT and mutant Ryr1 transcripts was determined in soleusmuscles of 2-month-old male WT (n � 3) and IT/� (n � 6) miceusing allele-specific primers (ASP) and real-time RT-PCR. TheWT and mutant Ryr1 transcripts differed by the two nucleotidebases that were introduced into the Ryr1 gene during mutagen-esis (19). To achieve the highest level of specificity, the ASPswere designed to anneal to the mutated region, to include bothnucleotide substitutions, and to terminate with a mismatch (SIMaterials and Methods). In all IT/� muscles sampled, the kineticsof PCR product accumulation from the mRNA pool were similarfor both ASPs (Fig. 5A), yielding an average expression ratio of1:1 between the allelic transcripts. Individual IT/� samplesshowed small stochastic f luctuations in the levels of both WT andmutant transcripts.

Western blotting (Fig. 5B) revealed no differences in theexpression of major contractile or Ca2� regulatory proteinsbetween 4-month-old IT/� mouse muscles and age-matchedcontrols. This result was expected because only 7% of all fiberswere affected in young mice (Table S2).

Whole Muscle Contractility. Measurements of the characteristics ofisometric contraction in isolated, intact, lumbrical (fast-twitch)and soleus (slow-twitch) muscles in 2-month-old IT/� males(n � 6) and age- and sex-matched WT mice (n � 5) at 28 °C (Fig.S4 A–D and Table S3) showed that force during a single twitchand submaximal tetanic contractions (i.e., �30 Hz) was 28–34%lower (P � 0.05) in IT/� mice, compared to WT. A significantdecrease (�37%) in peak twitch force and maximal twitch rateof contraction (�df/dt) in mutated muscles was observed. Nochange was observed in �df/dt. There was also a trend formaximal tetanic force (P � 0.09) and �df/dt (P � 0.12) to bereduced across all stimulation frequencies in IT/� mice. Thesedata suggest that contractile function is impaired in both fast-and slow-twitch IT/� muscle.

B

A

C

Fig. 4. Ultrastructural abnormalities in longitudinal sections of a type 1 fiberfrom an 18-month-old IT/� mouse. (A) Core-like lesions with rods streamingthe length of a full-width myofiber. (B) An unstructured core area with rods,which is contiguous with an area of relatively well-structured myofibrils,illustrating the origin of a ‘‘cornucopia’’ of rods in the Z-line. (C) A highermagnification of a rod from B, revealing its lattice-like structure. [Scale bars,5 �m (A), 2 �m (B), and 1 �m (D).]

Fig. 5. (A) Quantitative RT-PCR analysis of the relative abundance of allelicRyr1 transcripts in soleus muscle from 2-month-old IT/� mice. Representativeresults of the cDNA analysis of two IT/� mice (het 1 and het 2) and a normallittermate (WT) are shown. (B) Western blot analysis of the expression of Ca2�

regulatory proteins in microsomes from soleus muscle of WT and IT/� mice.The amount of total protein (�g) loaded per lane is shown Below the panels.

21816 � www.pnas.org�cgi�doi�10.1073�pnas.0912126106 Zvaritch et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021

DiscussionHere we show that the I4895T mutation causes a congenitalmyopathy with minicores, cores, and rods in KI Ryr1I4895T/wt

mice. The manifestation and progression of myopathy in themouse model recapitulates many features commonly observed inhuman congenital myopathies (1–3). Neonatal IT/� mice exhibithypotonia and respiratory distress, but recover; the course of thedisease is then slowly progressive. IT/� mice show impairedcontractility at 2 months in both fast fiber-rich (lumbrical) andslow fiber-rich (soleus) muscles (Fig. S4). After 8 months, manymice exhibit varying degrees of impaired locomotion and, in 14%of all cases (n � 30), a complete paresis of the hindquarters (Fig.1B and Fig. S1 B–E). Dorsal kyphosis, characteristic of congen-ital myopathies in humans (1–3), develops in 80% of aged IT/�mice. Variable penetrance is a common clinical feature of CCDand has been noted specifically for the RYR1I4898T mutation (20).

Histopathological analyses of IT/� skeletal muscle samplesreveal numerous abnormalities characteristic of congenital my-opathies. Specifically, increased fiber size variability and indi-vidual type 1 fiber atrophy observed in IT/� muscles (Fig. S2 Band D, and Table S1) are hallmarks of human congenitalmyopathies (3), whereas extended areas of oxidative enzymedepletion, defined as cores (Fig. 1 C–G and Fig. S3), arediagnostic of core myopathy (1, 2, 9). In contrast to humans, type1 fiber predominance is not a feature in the IT/� mouse model,nor is it in the �-tropomyosin related nemaline myopathy mousemodel (25). A possible explanation for this discrepancy might liein murine skeletal muscle physiology, which is strongly depen-dent on fast-twitch muscle activity (26).

An important finding is that structural abnormalities withinIT/� myofibers undergo an apparent age-dependent transitionfrom small, compacted areas resembling minicores in youngeranimals to cores and nemaline rods in adult and aged mice (Figs.2–4). Age-dependent changes in myofiber lesions have beensuggested in isolated CCD cases in which no abnormalities oronly minicores were found in muscle biopsies from youngerpatients, whereas cores were observed in subsequent musclebiopsies (5, 27) or in muscle biopsies of affected parents (28, 29).Nemaline rods have been detected in a few CCD cases (4–7).Accordingly, it has been proposed that there is a common originfor minicore, core, and rod lesions. Moreover, a strict classifi-cation of these myopathies has been questioned (4, 5). Never-theless, it has not been possible to determine whether mixedCCD cases represent a typical course of disease progression. Todate, minicores or nemaline rods have not been observed inhumans expressing the RYR1I4898T/� mutation. However, IT/�mice, carrying the analogue of this mutation, express a fullspectrum of minicores, cores, and rods in an age-dependentprogression. These findings provide independent new evidencesupporting earlier proposals that minicores, cores, and rodsrepresent consecutive stages in the formation of core lesions (4,5). They also support the hypothesis that structural abnormali-ties, conventionally attributed to distinct congenital myopathiesand classified as MmD, CCD, and myopathy with rods, do havea common etiology (4, 5) and pathogenesis.

Ultrastructural abnormalities in IT/� myofibers are observedin 6-week-old mice (Fig. 3 and Table S2). Irregularities inmyofibrillar organization, their focal thinning, misalignment,and intermingling, suggest sporadic, local abnormalities in myo-fibrillogenesis. Peak twitch contractions (Fig. S4) assessed inyoung IT/� mice show only a mild reduction in force generation,suggesting that the disease is slowly progressive.

Measurements of electrically evoked and 4-CMC-inducedRyR1 Ca2� transients in RyR1 I4898T mutant channels ex-pressed in the heterozygous state in dysgenic myotubes showedthat global Ca2� release was reduced by 60% (16), therebyaccounting for impaired EC coupling and muscle performance.

EC uncoupling and/or reduced RyR1-mediated Ca2� release arethought to play a key role in skeletal muscle disuse and wasting(30), in exercise-induced muscle damage (31), muscle aging (32),and denervation (33). Denervation and strenuous exercise areknown to produce core-like lesions (33–35), which are strikinglysimilar to those observed in core myopathies. It is reasonable tosuggest that the underlying cause of the ultrastructural abnor-malities observed in these conditions is reduced RyR1-mediatedCa2� release, with EC uncoupling as a possible common de-nominator.

As with human RyR1-related congenital myopathy, the dis-ease phenotype in IT/� mice is highly variable (Fig. 1 A and Band Fig. S1 B–E), even though genetic variability is minimizedon the stable Sv129 genomic background (19). Fluctuations intotal RyR1 protein content and in levels of expression of WT andmutant alleles are also minimal, because total RyR1 proteinlevels are normal in heterozygous IT/� muscles (Fig. 5B) andreal-time RT-PCR shows equimolar levels of mutant and WTRyr1 transcripts (Fig. 5A).

A recent study of the single channel properties of mutantRyR1 (36) has shown that functional variability is an intrinsicfeature of Ca2� release channels expressed in the heterozygousstate. It was then proposed that the random combination of WTand mutant subunits in an RyR1 tetramer might contribute tophenotypic variability in RyR1-related disorders. The RyR1population in heterozygotes would consist of six variants of RyR1tetramers: homotetrameric WT channels of normal activity,homotetrameric mutant channels of compromised or null activ-ity, and heterotetrameric channels of intermediate activity aris-ing from four possible tetrameric arrangements of WT:mutantsubunits (3:1; 1:3; 2:2, side by side; and 2:2, diagonally apposed).Our previous data showed that the homozygous I4895T muta-tion disrupted RyR1-mediated Ca2� release without altering thestructural integrity of the RyR1 protein or of the supramolecularCRU complexes (19). Assuming that the formation of RyR1tetrameric complexes in heterozygous IT/� mouse muscle oc-curs via random combination of WT and mutant subunitspresent in equimolar quantities, �1/16 of all RyR1 tetramerswould be fully active WT homotetramers; 1/16 would be func-tionally inactive IT/IT homotetramers; and 14/16 would exhibitan intermediate level of functional competence. At higher levelsof assembly, the CRU would be formed from random combi-nations of functionally diverse RyR1 tetramers with normaldihydropyridine receptor (DHPR) complexes. This would notonly increase the potential for inhomogeneity of Ca2� releasewithin a CRU, but might also lead to disparity between sarco-meric domains controlled by functionally different CRUs.

We propose that heterogeneity of Ca2� release channelfunction, created at the levels of RyR1 tetramer and CRUassembly, results in a long-term spatial heterogeneity of Ca2�

release that leads to a corresponding spatial heterogeneity ofcontractile force and, eventually, to impaired muscle activity.Stoichiometry of channel function is difficult to measure becauseof the intrinsic characteristics of ion gradients and channelgating. However, Feske et al. (37) have demonstrated thatmeasurable changes in Ca2� channel density can be correlateddirectly with measurable changes in Ca2� channel function ifconditions are created in which functional channels are limitingto ion flux. If we assume that RyR1 Ca2� release channelfunction in IT/� mice is limiting, then small f luctuations in theactivity of a CRU could have significant effects on local Ca2�

release and subsequent contractility of the sarcomeres governedby each CRU.

Our EM studies capture the apparent consequence of theheterogeneity of contraction that exists within IT/� myofibers(Fig. 3 A–C). The discordance of contraction between adjacentmyofibers is seen as disruption of sarcomeric register. Suchdiscordance would create physical stresses within myofibers that

Zvaritch et al. PNAS � December 22, 2009 � vol. 106 � no. 51 � 21817

MED

ICA

LSC

IEN

CES

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021

would lead to tearing and shearing between myofibrils and evento disruptions of the sarcotubular and mitochondrial systemsthat would, over time, cause physical and functional damage toCa2� regulatory, energetic, and contractile systems. We proposethat, as these processes progress, small foci of damage wouldcoalesce and manifest as minicores, cores, and rods that wouldcompromise muscle function and force. Most importantly, theheterogeneity of Ca2� release and contractile force that wouldoccur, regardless of whether the heterozygous RyR1 mutationenhanced Ca2� leak or caused EC uncoupling, could representa common mechanism for core formation arising from function-ally different RyR1 mutations. Thus we propose a unifyingtheory for the pathogenic mechanism leading to the formationof structural abnormalities in RyR1-related muscle disorders.

In conclusion, our results indicate that the IT/� mouse linerepresents a unique and phenotypically valid model of RyR1-related congenital myopathy with minicores, cores, and rods.Amenable to systematic sampling, the model offers high poten-tial for further unraveling of the abnormal molecular mecha-nisms that underlie the pathogenesis of core myopathies and ofskeletal muscle disorders arising from EC uncoupling. Ulti-mately, it will be valuable for testing the efficacy of therapeuticstrategies designed to combat progression of the diseasephenotype.

Materials and MethodsAnimal Handling. Experimental protocols for animal research were approvedby the Institutional Animal Care and Use Committees at all universities in-volved in this study. The generation and genotyping of Ryr1I4895T/wt mice was

described previously (19). For maintenance, IT/� mice, generated on an Sv129background, were crossed with WT 129S2/SvPasCrl mice (Charles River).

Histology. Mice were euthanized by cervical dislocation. Muscles were dis-sected within 10–15 min postmortem. Multiple transverse and longitudinalsections of vastus lateralis, gastrocnemius, EDL, tibialis anterior, and soleusmuscles from IT/� and WT littermates were examined for abnormalities.Specimens of buffered formalin- or glutaraldehyde-fixed tissue were embed-ded in paraffin, sectioned in transverse orientation (3–5 mm), and stained withhematoxylin and eosin (3). For NADH-TR reactivity, fresh muscle specimenswere frozen in isopentane cooled in liquid nitrogen, sectioned in an IECcryostat, and stained (3). To discriminate minicore/core lesions from possible‘‘edge’’ or ‘‘dissection contracture’’ artifacts only lesions located within cen-tral areas of sections and restricted to one myofiber were analyzed.

Electron Microscopy. Muscle samples were fixed immediately, without pin-ning, in 2% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.2,postfixed in 1% osmium tetroxide in the same buffer, dehydrated in a gradedethanol series, followed by propylene oxide, and embedded in a Quetol-Spurrresin mixture. Semithin longitudinal and transverse sections (0.6 �m) werestained with toluidine blue on a hot plate for 30 seconds. Thin sections (0.1�m) were cut on an RMC6000 ultramicrotome, stained with uranyl acetate andlead citrate, and viewed in an FEI CM100 TEM. Cross-sectional fiber diameterswere determined by standard methods (3).

ACKNOWLEDGMENTS. We thank Drs. Susan Brown, Imperial College, London,UK, and Nicole Monnier, University Hospital, Grenoble, France, for expertadvice and stimulating discussions. This work was supported by CanadianInstitutes of Health Research Grants MT 3399 and MOP 49493 to D.H.M.,National Heart, Lung, and Blood Institute, and National Institutes of Healthgrants to J.G.S. and C.E.S., and a Howard Hughes Medical Institute Grant toC.E.S.

1. Sharma MC, Jain D, Sarkar C, Goebel HH (2009) Congenital myopathies–a comprehen-sive update of recent advancements. Acta Neurol Scand 119:281–292.

2. Sewry CA, Jimenez-Mallebrera C, Muntoni F (2008) Congenital myopathies. Curr OpinNeurol 21:569–575.

3. Dubowitz V, Sewry CA (2007) Muscle Biopsy: A Practical Approach (Elsevier, Amster-dam) 3rd Edition.

4. Bethlem J, Arts WF, Dingemans KP (1978) Common origin of rods, cores, miniaturecores, and focal loss of cross-striations. Arch Neurol 35:555–566.

5. Vallat JM, et al. (1982) Coexistence of minicores, cores, and rods in the same musclebiopsy. A new example of mixed congenital myopathy. Acta Neuropathol 58:229–232.

6. Monnier N, et al. (2000) An autosomal dominant congenital myopathy with cores androds is associated with a neomutation in the RYR1 gene encoding the skeletal muscleryanodine receptor. Hum Mol Genet 9:2599–2608.

7. Scacheri PC, et al. (2000) A novel ryanodine receptor gene mutation causing both coresand rods in congenital myopathy. Neurology 55:1689–1696.

8. Ferreiro A, et al. (2002) A recessive form of central core disease, transiently presentingas multi-minicore disease, is associated with a homozygous mutation in the ryanodinereceptor type 1 gene. Ann Neurol 51:750–759.

9. Sewry CA, et al. (2002) The spectrum of pathology in central core disease. NeuromusculDisord 12:930–938.

10. Ferreiro A, et al. (2000) Multi-minicore disease–searching for boundaries: Phenotypeanalysis of 38 cases. Ann Neurol 48:745–757.

11. Robinson R, Carpenter D, Shaw MA, Halsall J, Hopkins P (2006) Mutations in RYR1 inmalignant hyperthermia and central core disease. Hum Mutat 27:977–989.

12. Monnier N, et al. (2003) A homozygous splicing mutation causing a depletion ofskeletal muscle RYR1 is associated with multi-minicore disease congenital myopathywith ophthalmoplegia. Hum Mol Genet 12:1171–1178.

13. Jungbluth H, et al. (2007) Centronuclear myopathy due to a de novo dominantmutation in the skeletal muscle ryanodine receptor (RYR1) gene. Neuromuscul Disord17:338–345.

14. Sato I, et al. (2008) Congenital neuromuscular disease with uniform type 1 fiber andRYR1 mutation. Neurology 70:114–122.

15. Tong J, McCarthy TV, MacLennan DH (1999) Measurement of resting cytosolic Ca2�

concentrations and Ca2� store size in HEK-293 cells transfected with malignant hyper-thermia or central core disease mutant Ca2� release channels. J Biol Chem 274:693–702.

16. Avila G, O’Brien JJ, Dirksen RT (2001) Excitation–contraction uncoupling by a humancentral core disease mutation in the ryanodine receptor. Proc Natl Acad Sci USA98:4215–4220.

17. Yang T, et al. (2006) Pharmacologic and functional characterization of malignanthyperthermia in the R163C RyR1 knock-in mouse. Anesthesiology 105:1164–1175.

18. Chelu MG, et al. (2006) Heat- and anesthesia-induced malignant hyperthermia in anRyR1 knock-in mouse. FASEB J 20:329–330.

19. Zvaritch E, et al. (2007) An Ryr1I4895T mutation abolishes Ca2� release channelfunction and delays development in homozygous offspring of a mutant mouse line.Proc Natl Acad Sci USA 104:18537–18542.

20. Lynch PJ, et al. (1999) A mutation in the transmembrane/luminal domain of theryanodine receptor is associated with abnormal Ca2� release channel function andsevere central core disease. Proc Natl Acad Sci USA 96:4164–4169.

21. Gao L, et al. (2000) Evidence for a role of the lumenal M3–M4 loop in skeletal muscleCa2� release channel (ryanodine receptor) activity and conductance. Biophys J 79:828–840.

22. De Cauwer H, Heytens L, Martin JJ (2002) Workshop report of the 89th ENMC Inter-national Workshop: Central Core Disease, 19th–20th January 2001, Hilversum, TheNetherlands. Neuromuscul Disord 12:588–595.

23. Jungbluth H, et al. (2000) Minicore myopathy in children: A clinical and histopatho-logical study of 19 cases. Neuromuscul Disord 10:264–273.

24. Shafiq SA, Gorycki MA, Milhorat AT (1969) An electron microscope study of fibre typesin normal and dystrophic muscles of the mouse. J Anat 104:281–293.

25. Corbett MA, et al. (2001) A mutation in alpha-tropomyosin(slow) affects musclestrength, maturation and hypertrophy in a mouse model for nemaline myopathy. HumMol Genet 10:317–328.

26. Close RI (1972) Dynamic properties of mammalian skeletal muscles. Physiol Rev 52:129–197.

27. Monnier N, et al. (2008) Null mutations causing depletion of the type 1 ryanodinereceptor (RYR1) are commonly associated with recessive structural congenital myop-athies with cores. Hum Mutat 29:670–678.

28. Morgan-Hughes JA, Brett EM, Lake BD, Tome FM (1973) Central core disease or not?Observations on a family with a non-progressive myopathy. Brain 96:527–536.

29. Dubowitz V, Roy S (1970) Central core disease of muscle: Clinical, histochemical andelectron microscopic studies of an affected mother and child. Brain 93:133–146.

30. Lynch GS, Schertzer JD, Ryall JG (2007) Therapeutic approaches for muscle wastingdisorders. Pharmacol Ther 113:461–487.

31. Proske U, Morgan DL (2001) Muscle damage from eccentric exercise: Mechanism,mechanical signs, adaptation and clinical applications. J Physiol 537:333–345.

32. Payne AM, Delbono O (2004) Neurogenesis of excitation-contraction uncoupling inaging skeletal muscle. Exerc Sport Sci Rev 32:36–40.

33. Schmitt HP, Volk B (1975) The relationship between target, targetoid, and targetoid/core fibers in severe neurogenic muscular atrophy. J Neurol 210:167–181.

34. Engel WK (1961) Muscle target fibres, a newly recognized sign of denervation. Nature191:389–390.

35. Newham DJ, McPhail G, Mills KR, Edwards RH (1983) Ultrastructural changes afterconcentric and eccentric contractions of human muscle. J Neurol Sci 61:109 –122.

36. Xu L, Wang Y, Yamaguchi N, Pasek DA, Meissner G (2008) Single channel properties ofheterotetrameric mutant RyR1 ion channels linked to core myopathies. J Biol Chem283:6321–6329.

37. Feske S, et al. (2006) A mutation in Orai1 causes immune deficiency by abrogating CRACchannel function. Nature 441:179–185.

21818 � www.pnas.org�cgi�doi�10.1073�pnas.0912126106 Zvaritch et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

8, 2

021