OUP user menu

Human disease caused by loss of fast IIa myosin heavy chain due to recessive MYH2 mutations

Homa Tajsharghi, David Hilton-Jones, Olayinka Raheem, Anna Maija Saukkonen, Anders Oldfors, Bjarne Udd
DOI: http://dx.doi.org/10.1093/brain/awq083 1451-1459 First published online: 24 April 2010


Striated muscle myosin heavy chain is a molecular motor protein that converts chemical energy into mechanical force. It is a major determinant of the physiological properties of each of the three muscle fibre types that make up the skeletal muscles. Heterozygous dominant missense mutations in myosin heavy chain genes cause various types of cardiomyopathy and skeletal myopathy, but the effects of myosin heavy chain null mutations in humans have not previously been reported. We have identified the first patients lacking fast type 2A muscle fibres, caused by total absence of fast myosin heavy chain IIa protein due to truncating mutations of the corresponding gene MYH2. Five adult patients, two males and three females, from three unrelated families in UK and Finland were clinically assessed and muscle biopsy was performed in one patient from each family. MYH2 was sequenced and the expression of the corresponding transcripts and protein was analysed in muscle tissue. The patients had early-onset symptoms characterized by mild generalized muscle weakness, extraocular muscle involvement and relatively favourable prognosis. Muscle biopsy revealed myopathic changes including variability of fibre size, internalized nuclei, and increased interstitial connective and adipose tissue. No muscle fibres expressing type IIa myosin heavy chain were identified and the MYH2 transcripts were markedly reduced. All patients were compound heterozygous for truncating mutations in MYH2. The parents were unaffected, consistent with recessive mutations. Our findings show that null mutations in the fast myosin heavy chain IIa gene cause early onset myopathy and demonstrate that this isoform is necessary for normal muscle development and function. The relatively mild phenotype is interesting in relation to the more severe phenotypes generally seen in relation to recessive null mutations in sarcomeric proteins.

  • muscle
  • myosin heavy chain
  • mutation
  • myopathy
  • recessive


Myosin is one of the most abundant proteins in the body and is indispensable for body movement and heart contractility. Three major myosin heavy chain (MyHC) isoforms are present in adult human limb skeletal muscle: MyHC I, also called slow/β-cardiac MyHC, is the gene product of MYH7 and is expressed in slow, type 1 muscle fibres as well as in the ventricles of the heart; MyHC IIa (MYH2) is expressed in fast, type 2A muscle fibres; and MyHC IIx (MYH1) is expressed in fast, type 2B muscle fibres. The three different muscle fibre types display distinct physiological properties and have unique roles in the function of skeletal muscle (Larsson and Moss, 1993). We describe the clinical and morphological characteristics of patients from three unrelated families lacking the production of MyHC IIa due to non-sense and truncating mutations in MYH2.

Materials and methods


Five patients were clinically assessed (Table 1). Three of the patients had mild to moderate generalized muscle weakness from early childhood, with minor progression. Two were subjectively asymptomatic. All had facial muscle weakness and marked external ophthalmoplegia and two had ptosis.

View this table:
Table 1

Clinical data

PatientII:1 (Family A) UKII:2 (Family A) UKII:3 (Family A) UKII:2 (Family B) FinlandII:1 (Family C) Finland
Age (years)4142445859
Age at onset of muscle weaknessEarly childhood. Little or no progressionAsymptomatic, except for ‘lazy eye’ noted in childhood. No subsequent ocular symptomsAsymptomaticGeneral muscle weakness from early childhoodPtosis, ophthalmoplegia and mild general weakness since early childhood
Distribution of muscle weaknessFacial muscle weaknessFacial muscle weaknessFacial muscle weaknessFacial muscle weakness Upper limbs MRC grade 4–5 Abdominal muscle weakness MRC grade 3 Mild proximal weakness in lower limbsFacial muscle weakness Upper limbs MRC grade 4–5 Abdominal muscle weakness MRC grade 3 Mild proximal weakness in lower limbs
Neck flexion weaknessNeck flexion weaknessNeck flexion weakness
Diffuse limb muscle slimness–mild weakness most marked proximallyElbow flexion and ankle dorsiflexionElbow flexion and ankle dorsiflexion
Other signs or symptomsSymptomatic joint hypermobilityAsymptomatic joint hypermobilityAsymptomatic joint hypermobilityCongenital pectus carinatum surgically corrected.
No improvement on strength training
EMGMyopathic—more marked in proximal musclesNot investigatedNot investigatedMild myopathicMyopathic
s-CKNormalNot investigatedNot investigatedNormalNormal
Muscle imagingNot investigatedNot investigatedNot investigatedModerate diffuse fatty degenerative change in thigh and in medial gastrocnemiusModerate diffuse fatty degenerative change in thigh and in medial gastrocnemius
  • MRC = Medical Research Council scale for grading of muscle strength (Aids to the Examination of the Peripheral Nervous System. Elsevier, 2000); s-CK = Creatine kinase in serum.

Muscle morphology

Muscle biopsy specimens were obtained from one patient of each family. In Patient II:1 (Family A) muscle biopsy specimens were obtained from the vastus lateralis of the quadriceps femoris muscle, at age 38, and from the deltoid muscle at age 40. In Patient II:2 (Family B) and Patient II:1 (Family C) muscle biopsies were obtained from the vastus lateralis of the quadriceps femoris muscle at age 55 and 58, respectively. Enzyme and immunohistochemical analyses, including MyHC isoforms, of freshly frozen muscle biopsy specimens were performed as previously described (Tajsharghi et al., 2002). In Patients II:2 (Family B) and II:1 (Family C) a new double immunostaining method for MyHC isoforms was performed that shows the expression of different MyHC isoforms in different muscle fibres in a single section (Raheem et al., 2010).

DNA analysis

Genomic DNA was extracted from frozen skeletal muscle or peripheral blood using DNA Extraction Kit (Qiagen, Hilden, Germany). Polymerase chain reaction (PCR) analysis was performed in a master mixture (ReddyMix PCR Master Mix; Abgene, Epsom, UK) after addition of 20 pmol of each primer and genomic DNA. PCR amplifications were performed as previously described (Tajsharghi et al., 2005). Nucleotide sequence determination was performed by cycle sequencing using a BigDye Terminator DNA sequencing kit (Applied Biosystems, Hercules, CA).

RNA analysis

The complementary DNA of MyHC isoforms, including the three adult skeletal isoforms, are highly homologous. In order to solely amplify fragments of MYH2 by PCR, we performed alignment of MYH2, MYH1, MYH7, MYH4, MYH3 and MYH8 complementary DNA (http://bio.lundberg.gu.se/edu/msf.html) to design MYH2 specific primers. Total RNA was extracted from muscle tissue of the patients using the Total RNA Isolation System (Promega, Madison, WI). Synthesis of first-strand complementary DNA was performed using Ready-To-Go You-Prime First-Strand Beads (Amersham Pharmacia Biotech, Uppsala, Sweden) according to the manufacturer’s instructions using 1 µg total RNA.

To analyse the splicing of exon 8 of MYH2 in Patient II:1 (family A), PCR was performed on complementary DNA with forward primer AGTGACGGTGAAGACTGAGGGA (corresponding to nucleotide 177–198 of human MyHC IIa complementary DNA sequence) combined with a backward primer ATCTGTGGCCATCAGTTCTTCCT (corresponding to nucleotide 986–1008 of human MyHC IIa complementary DNA). The resulting PCR products were analysed by sequencing after separation on 2% agarose gel and purification using QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany). In addition, PCR was performed on complementary DNA with forward primer AGGGAGCTGGTGGAGGGGCC (corresponding to nucleotide 1898–1917 of human MyHC IIa complementary DNA sequence) combined with a backward primer CTTGACATTCATGAAGGATCT (corresponding to nucleotide 2473–2493 of human MyHC IIa complementary DNA sequence) covering exon 15 through 20 to analyse the p.R783X mutation in Patient II:1 (Family A). This primer pair was also used to analyse the p.L802X mutation and the splicing of exon 16 in Patient II:2 (Family B) and Patient II:1 (Family C). The PCR amplifications consisted of an initial preheating step for 5 min at 94°C, followed by a touchdown PCR with denaturation at 94°C for 30 s, annealing at 65°C for 30 s and extension at 72°C for 1 min with a 1°C temperature decrement per cycle during the first 10 cycles. The subsequent cycles (40 cycles) each consisted of 94°C for 30 s, 55°C for 30 s and 72°C for 1 min.

To analyse the proportion of transcripts of the three major MyHC isoforms, PCR was performed on complementary DNA extracted from skeletal muscle and fragment analysis was performed as previously described (Tajsharghi et al., 2002).

Protein analysis

To analyse the expression of the MyHC isoforms, proteins extracted from muscle biopsy specimens were separated by 8% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) as previously described (Tajsharghi et al., 2002).

Haplotype analysis

Haplotype analysis was performed with micro-satellite markers.


Laboratory investigations

Morphological analysis of biopsy specimens from the quadriceps femoris and deltoid muscles of Patient II:1 (Family A) demonstrated type 1 fibre uniformity in the deltoid muscle and absence of type 2A fibres in both muscles (Fig. 1). A biopsy specimen from vastus lateralis of the quadriceps femoris muscle of Patient II:2 (Family B) demonstrated absence of MyHC IIa and myopathic features including increased variability of fibre size and internalized nuclei (Fig. 2). In Patient II:1 (Family C) a muscle biopsy of the vastus lateralis of the quadriceps muscle showed absence of muscle fibres expressing type IIa MyHC, as well as myopathic changes that included marked variability in fibre size, internalized muscle fibre nuclei, increased interstitial fat and connective tissue and type 1 fibre uniformity (Fig. 3A–C).

Figure 1

Muscle biopsy from quadriceps and deltoid muscles of Patient II:1 (Family A). (A–C) The quadriceps muscle include type 1 and type 2B fibres. (D–E) The deltoid muscle specimen shows type 1 fibre uniformity with expression of only slow/β cardiac myosin heavy chain. Bar corresponds to 100 µm.

Figure 2

Quadriceps muscle biopsy sections of Patient II:2 (Family B). (A–B) There is increased variability of muscle fibre size with atrophic and hypertrophic fibres and occasional fibres with internalized nuclei and lack of type 2A muscle fibres. (C) Immunohistochemical staining demonstrates muscle fibres with expression of either of myosin heavy chain I and IIx. No fibres expressing IIa MyHC are present. (D) Immunohistochemical of control muscle demonstrating muscle fibres expressing IIa myosin heavy chain (red fibres). Bars correspond to 50 µm.

MRI or CT of skeletal muscle in two of the patients showed diffuse fatty infiltration with an unusual pattern of predominant involvement of medial gastrocnemius in the lower legs, combined with predominant involvement of the semitendinosus, gracilis and vastus lateralis muscles in the thigh. The tibialis anterior muscle, which mainly consists of slow muscle fibres, showed normal appearance (Fig. 3D–G).

Figure 3

Muscle histopathology and MRI of Patient II:1 (Family C). (A–C) Sections of a muscle biopsy specimen from vastus lateralis of the quadriceps femoris muscle demonstrating fatty infiltration (arrow heads), hypertrophic and atrophic muscle fibres with internalized nuclei, type 1 fibre predominance, as well as slight disorganization of the intermyofibrillar network as revealed by NADH-tetrazolium reductase. (D–G) MRI of pelvis and legs at age 58 years demonstrating fatty infiltration in semitendinous, gracilis, vastus lateralis of the quadriceps femoris and medial gastrocnemius muscles.

Molecular genetics

The incentive to consider mutated MYH2 as a plausible cause of the disease was the ophthalmoplegia in the patients of Family A since in skeletal myopathy associated with a dominant missense mutation, p.E706K in MYH2, all patients had ophthalmoplegia and abnormal type 2A muscle fibres (Martinsson et al., 2000). In Families B and C it was the total absence of fast IIa fibres with the new double immunostaining technique (Raheem et al., 2010) in proximal muscle biopsy specimens that indicated a MYH2 defect.

Mutation analysis of MYH2 was performed in six individuals. In Patient II:1 (Family A), we identified two sequence variants. First, a heterozygous G to A change affecting a highly conserved nucleotide of the 5′ splice junction of intron 8 (c.904+1G>A). PCR analysis of complementary DNA in a region covering exons 2–10 of MYH2 revealed two different fragments: one fragment of normal size and a shorter fragment. Sequence analysis of the short fragment demonstrated skipping of exon 8, shifting of the reading frame and a premature stop codon (p. Tyr269-Glu302delfsX) (Fig. 4B). The second variant was a heterozygous nonsense mutation, c.2347C>T, changing Arginine at position 883 to a stop codon (p.Arg783X) in exon 19 (Fig. 4C). The same two mutations were also identified in siblings II:2 and II:3 (Family A). The unaffected father (I:1, Family A) had only the heterozygous 5′ splice site mutation of intron 8 indicating that the c.2347C>T mutation was inherited from the mother.

Figure 4

Pedigrees and DNA sequencing chromatograms of MYH2 in the patients. (A) Pedigree of Family A. (B) Complementary DNA (cDNA) sequence chromatogram of exons 7 and 9 demonstrating skipping of exon 8 in Patient II:1 (Family A). The normal sequence is illustrated in Supplementary Fig. 1. (C) Genomic DNA sequence chromatogram of exon 19 of Patient II:1 (Family A) carrying the heterozygous c.2347C>T mutation, changing the arginine at position 783 to a stop codon. (D) Pedigrees of Families B and C. (E) Complementary DNA sequence chromatogram of exons 15 and 17 showing skipping of exon 16 in Patient II:2 (Family B). The same results were obtained in Patient II:1 (Family C). (F) Genomic DNA sequence chromatogram of exon 19 of Patient II:2 (Family B) carrying the heterozygous c.2405T>A mutation, changing the leucine at position 802 to a stop codon. The same results were seen in Patient II:1 (Family C). Amino acid sequences in black indicate the normal sequences; sequences in red indicate the amino acid changes due to the mutations; and sequences in blue indicate the mutant allele. Filled symbols in the pedigrees show the individuals that are clinically and genetically affected.

In Families B and C, we identified in each of two patients (Patient II:2 of Family B and II:1 of Family C) two variants with truncating effects in MYH2. The two different variants were identical in both families. The first was a heterozygous A to G change affecting the highly conserved second nucleotide of the 3′ splice site of intron 15 (c.1975-2A>G) which resulted in skipping of exon 16 and shifting of the reading frame (p. Glu659-Gly687delfsX11) (Fig. 4E). The second variant was a heterozygous non-sense mutation, c2405T>A, changing leucine at position 802 to a stop codon (p.Leu802X) in exon 19 (Fig. 4F). Sequence analysis of complementary DNA demonstrated that the patients were compound heterozygous for the two truncating mutations. PCR amplification of complementary DNA of Patient II:2 (Family B) and Patient II:1 (Family C) in the region covering exon 15 through exon 20 of MYH2 generated two products: a large fragment derived from normal splicing and a small fragment with skipping of exon 16. Sequence analysis of the large fragment revealed normal splicing of exon 16 in combination with the c.2405T>A mutation in exon 19. Sequence analysis of the small PCR fragment revealed skipping of exon 16 and creation of a stop codon combined with wild-type c.2405T in exon 19.

Analysis of MYH2 transcripts

To determine the effect of the mutations on MYH2 gene expression, analysis of the relative level of expression of different isoforms of MyHC mRNA was performed by PCR on complementary DNA and fragment analysis. These results demonstrate that the three patients express very low levels of MYH2 transcripts (Fig. 5A).

Figure 5

Expression of myosin heavy chain isoforms. (A) Quantitative analysis of relative expression of MyHC I, MyHC IIa and MyHC IIx messenger RNA based on reverse transcription PCR analysis. The complementary DNA fragment with skipping of exon 16 differs from the wild-type cDNA fragment by 106 nucleotides. Small amounts of the transcripts of MyHC IIa with skipping of exon 16 in Patient II:2 (Family B) and Patient II:1 (Family C) are present. In the deltoid muscle of Patient II:1 (Family A) and the quadriceps muscle of Patient II:1 (Family C) MyHC IIx transcripts are undetectable whereas in the quadriceps muscle of Patient II:2 (Family B) MyHC IIx is expressed at levels comparable to those of the normal control. (B) Expression of MyHC isoforms by SDS–PAGE in muscle homogenate of Patient II:1 (Family A) (deltoid muscle) and Patient II:1 (Family C) (quadriceps femoris muscle) showing MyHC I predominance in both samples. A control muscle sample from a biceps muscle of an individual, without evidence of muscle disease, demonstrates the normal occurrence of three major MyHC isoforms.

Protein analysis

The expression of MyHC isoforms by SDS–PAGE analysis of the deltoid muscle of Patient II:1 (Family A) and the quadriceps muscle of Patient II:1 (Family C) confirmed the absence of MyHC IIa protein (Fig. 5B). There was a predominant expression of slow/β-cardiac MyHC (MyHC I) in these two muscle biopsy specimens.

Haplotype analysis

Haplotype analysis of Patient II:2 (Family B) and Patient II:1 (Family C) revealed that the two patients carried the identical haplotype over a distance ∼3.3 Mb on one chromosome with the c.1975-2A>G mutation, whereas sharing of a shorter segment (0.7–1.5 Mb) on the other chromosome indicates that the c.2405T>A mutation was more ancient.


We have identified the first patients with loss of a MyHC isoform, MyHC IIa and complete loss of one of the major muscle fibre types, type 2A. Our patients were compound heterozygous for truncating mutations in MYH2 resulting in loss of expression of MyHC IIa mRNA as well as any functional protein. Whether the reduced transcript expression was the result of non-sense mediated mRNA decay could not be established, but the functional consequences would be similar to complete inactivation of MYH2. The parents in all three families had no symptoms or signs of muscle dysfunction implying that all four mutations are recessive and that hemizygous loss of MyHC IIa expression does not lead to haploinsufficiency and disease.

In human limb muscle there are two fast MyHC isoforms: MyHC IIa (corresponding to MyHC IIa in the mouse) and MyHC IIx (corresponding to MyHC IId/x in the mouse). Mice also express a third fast MyHC isoform in limb skeletal muscle: MyHC IIb. Results from studies on MyHC IId/x and MyHC IIb null mice demonstrate that these genes are required for the normal muscle development and function of adult skeletal muscle in the mouse and that the different fast MyHC isoforms are functionally unique and cannot substitute for one another (Acakpo-Satchivi et al., 1997; Sartorius et al., 1998; Allen et al., 2001; Allen and Leinwand, 2001). MyHC IIa null mice have been reported but not characterized in detail (Geurts et al., 2006).

Our patients with loss of fast MyHC IIa expression exhibited muscle weakness and myopathic changes with predominant involvement of semitendinous, gracilis, vastus lateralis and medial gastrocnemius muscles in the lower limbs. The reason for preferential involvement of these muscles remains to be demonstrated but may reflect the relative proportion of MyHC IIa in these muscles, since the tibial anterior muscle, which is predominantly composed of slow fibres, showed normal appearance on imaging. In MyHC IIb knockout mice, two factors appeared to determine the extent to which a muscle was affected: the level of MyHC IIb and the amount of muscle activity (Allen et al., 2000). In the MyHC IId/x null mice there was no such correlation suggesting that other factors may also be of importance (Allen et al., 2000).

The expression of MyHC IIa can be detected from around 24 weeks gestational age to adulthood in humans and it is one of the major MyHC isoforms expressed in human skeletal muscle (Butler-Browne et al., 1990; Cho et al., 1994; Smerdu et al., 1994). This implies that our patients had disturbed development and maturation of skeletal muscle from around 24 weeks of gestational age, which is consistent with the early onset of symptoms. However, none of the patients were identified at birth as having a congenital myopathy. Analogous with the MyHC IIb and IId/x null mice, our patients showed slow progression of muscle wasting with increasing age (Acakpo-Satchivi et al., 1997; Allen et al., 2001; Allen and Leinwand, 2001). As in mice, the progression may be related to ongoing degeneration and regeneration as indicated by the pathological changes in muscle biopsy specimens. Why degenerative changes in the type 1 and type 2B fibres occur when one MyHC isoform and fibre type is lacking is not clear. A possible explanation could be that proper maintenance of muscle tissue requires all myosin isoforms and fibre types, and that muscle fibre degeneration is a consequence of less capability to sustain mechanical load during normal activity if one fibre type is lost.

The explanation of the apparent type 1 fibre uniformity and predominant expression of slow myosin in two of the investigated muscles (deltoid muscle of Patient II:1 in Family A and quadriceps muscle of Patient II:1 in Family C) is not clear. It is well known that in many congenital or early onset myopathies, such as nemaline myopathy and central core disease, there is predominance and sometimes uniformity of type 1 fibres. However, in our patients, type 1 fibre predominance was not a consistent finding in all muscles since Patient II:1 (Family A) had a normal amount of type 2B fibres in the quadriceps muscle and Patient II:2 (Family B) expressed MYH1 gene transcript at a nearly normal level in the quadriceps muscle, and up to 15% of fibres expressed MyHC IIx on immunohistochemistry. In the quadriceps muscle the normal proportion of type 1 fibres is between 44 and 57% (Lexell et al., 1983).

The clinical phenotype of our patients with compound heterozygous null mutations of MYH2 was rather mild. This was unexpected since several other myopathies caused by recessive null mutations of sarcomeric proteins that exist in different isoforms show a much more severe clinical phenotype. Absence of α-tropomyosin slow (TPM3) (Tan et al., 1999) or muscle troponin T slow (TNNT1) (Jin et al., 2003) is associated with severe forms of nemaline myopathy. Complete loss of β-tropomyosin (TPM2) is associated with Escobar syndrome with nemaline myopathy (Monnier et al., 2009) and absence of α-skeletal muscle actin (ACTA1) is associated with persistent expression of developmental actin and severe or intermediate nemaline myopathy (Nowak et al., 2006). In addition, the various isoforms of skeletal muscle MyHC genes and proteins show a high degree of conservation in genomic structure and amino acid sequences. The orthologous isoforms of MyHC in different species have a greater extent of conservation than different isoforms within a species (Weiss et al., 1999) suggesting an important functional diversity within the MyHC gene family.

The fact that MyHC IIa is expressed in extraocular muscle can explain the ophthalmoplegia observed in all of our patients (Pette and Staron, 1997; Pedrosa-Domellöf et al., 2000). In patients with autosomal dominant myopathy associated with the heterozygous MYH2 p.E706K missense mutation, there was a clear correlation between pathology and expression of MyHC IIa indicating a dominant negative effect of this missense mutation (Tajsharghi et al., 2002). In the patients with no fast IIa MyHC due to compound heterozygous truncating MYH2 mutations, the situation is different and illustrates the importance of expression of MyHC IIa, even if hemizygous loss is well tolerated. Total absence of MyHC IIa cannot be substituted for by an increased expression of another MyHC isoform but the consequence of the total loss is a surprisingly mild phenotype.


Swedish Research Council (7122 to A.O.); Association Francaise Contre des Myopathies (to A.O.); The Sahlgrenska Hospital Research Funds (to A.O.); Tampere University Hospital (to B.U.); Vasa Central Hospital Research Funds (to B.U.); and the Folkhälsan Genetic Research Foundation (to B.U.).

Supplementary material

Supplementary material is available at Brain online.


Hannu Haapasalo, MD, PhD, is acknowledged for providing morphological muscle biopsy images in the Finnish patients and MSc Helena Luque for performing genotyping of the MYH2 locus in the Finnish patients.


  • Abbreviations:
    myosin heavy chain
    polymerase chain reaction


View Abstract