Clinical, serological and electrophysiological findings
A 34 year old man of English and Irish descent was born from a normal pregnancy and had normal motor development. He had no history of early joint contractures. He was notably tall and thin in elementary school. In middle school, he started experiencing difficulty keeping up with peers in sprinting and jumping activities. Limitations to his range of eye movements were first noticed by peers at age 23. By age 26, he had difficulty climbing stairs, while a year prior to his clinical presentation, he developed mild shoulder muscle weakness in the absence of myalgia. He had no dysphagia or dyspnea. His parents and two sisters were unaffected by muscle weakness (father was deceased but had no history of weakness). Neurological examination was notable for mild bilateral ptosis and ophthalmoparesis (Fig. 1a, b),
mild weakness of facial and neck flexor muscles 4+/5, as graded by the Medical Research Council (MRC) scale, symmetric muscle atrophy and weakness of the shoulder girdle muscles (Fig. 1c, d), brachioradialis and finger extensors (4, 4+/5), severe weakness of hip flexors (3/5 bilaterally), less severe involvement of thigh adductors and abductors, and ankle dorsiflexors (4+/5). The quadriceps were symmetrically atrophic but with spared strength. Axial weakness was suggested by patient’s inability to sit up from a supine position without assistance. He demonstrated a waddling gait and was able to walk on toes but not on heels. Tendon reflexes and sensory examination were normal. He had a high arched palate, laxity of the metacarpal phalangeal joints and mild gynecomastia (Fig. 1d), but no contractures. No tremor was present.
Creatine kinase (CK) levels were elevated (457–626 U/L; normal < 320). The electrocardiogram and echocardiography were normal. He had a reduced maximal expiratory pressure (73% of predicted) and normal vital capacity on pulmonary function testing. Concentric needle electromyography demonstrated mixed small and large motor unit potentials in proximal limb and thoracic paraspinal muscles with early recruitment, fibrillation potentials in the gastrocnemius and complex repetitive discharges in proximal lower limb muscles. Nerve conduction studies were normal. 2-Hz repetitive stimulation of the facial, spinal accessory and peroneal nerves showed no decrement. (Limited information on this patient were reported prior to the pathological characterization of the myopathy [17]).
Muscle biopsy findings
Routine histological studies of the triceps biopsy demonstrated wide fiber size variability, muscle fiber splitting (Fig. 2a), internalized nuclei, rare necrotic or regenerating fibers and occasional rimmed vacuoles (Fig. 2b). The perimysial and endomysial connective tissue were increased suggesting chronicity. Several fibers, mainly atrophic fibers, contained prominent subsarcolemmal nuclei and/or large (up to 30 microns in diameter) central accumulation of eosinophilic material (Fig. 2c) that stained dark red in trichrome (Fig. 2d) and overreacted for acid phosphatase (Fig. 2e). At a higher magnification (Fig. 2f) these large central inclusions had the appearance of a filamentous tangle or a “ball of yarn” and often contained nemaline rods of various size. Immunohistochemical studies (performed as previously reported [22]) demonstrated that these tangles overreact for sarcomeric alpha-actinin and to a lesser degree for myotilin (Fig. 2g, h), but not for sequestome 1 (p62), desmin, alphaB-crystallin (Fig. 2i–k) or neural cell adhesion molecule 1 (NCAM 1) (not shown), all of which were diffusely increased in fewer scattered atrophic fibers. The tangles did not overreact for dystrophin. Under rhodamine optics, only extremely rare atrophic fibers showed focal large congophilia and a single non-atrophic fiber demonstrated small subsarcolemmal congophilic inclusions (not shown). Immunohistochemical studies using an antibody against human MyHC-IIA protein (A4.74, Developmental Studies Hybridoma Bank (DSHB) (University of Iowa, Iowa City, Iowa, USA) [35] demonstrated paucity and atrophy of type 2A fibers (Fig. 2l), and MyHC-IIA positivity of the fibers harboring the filamentous tangles, as observed by optical (Fig. 2m) and confocal microscopy (Fig. 2n). Additionally, sections reacted for reduced nicotinamide adenine dinucleotide (NADH) dehydrogenase showed a myriad of ring fibers, lobulated fibers and several atrophic fibers that were diffusely overreactive (Fig. 2o). NADH enzyme reactivity was preserved or increased in correspondence of the tangles.
High magnification (Fig. 3a) showed the filamentous appearance of the large inclusions and nemaline rods more clearly, while serial sections (Fig. 3b–f) confirmed that the filamentous tangles and associated nemaline rods occurred in type 2A fibers. Approximatively 1 to 9 type 2A fibers were present in each fascicle, with most fascicles harboring 4–5 type 2A fibers (as adjudged by combined findings in ATPase stains at pH 4.3, 4.6 and 9.4), but the estimated number of pure 2A fibers was complicated by the presence of fibers of poorly differentiated histochemical type. Immunocytochemical studies indeed showed presence of hybrid fibers co-expressing myosin IIX and IIA (Fig. 4a–d). Thus, hybrid fibers may also contain tangles. To better demonstrate the atrophy of type 2A fibers, we measured fiber size in sections immunoreacted for myosin IIA and IIX [16]. MyHC-IIA fibers had a mean fiber diameter of 28.0 ± 2.2 µm (Fig. 5) (mean ± SEM) with rare fibers as large as 61 µm. The mean diameter of hybrid MyHC-IIA/IIX fibers was 47.7 ± 4.5 µm with extremely rare fibers of up to 98 µm. MyHC-IIA and hybrid MyHC-IIA/IIX diameters were not significantly different (p = 0.254). MyHC-IIX fiber types had a mean diameter of 77.4 ± 3.2 µm (range:13–233 μm), while type 1 fibers had a mean diameter of 101.3 ± 4.0 µm (range: 30–243 µm). MyHC-IIA and MyHC-IIA/IIX fibers were less frequent and statistically smaller than pure MyHC-IIX and type I fibers.
Electron microscopy (EM) studies were performed on previously frozen muscle biopsy tissue. Tissue samples were taken directly from the − 80 °C freezer and separated into approximately 2 mm pieces. Individual pieces were quickly placed in ice-cold Trump’s fixative, making sure they remained frozen before placing into fixative. After approximately two hours, tissue was transferred to a 4 °C refrigerator and allowed to fix overnight. Following fixation, the tissue was washed with phosphate buffer, post-fixed in 1% osmium tetroxide, washed in H2O, dehydrated through a graded series of ethanol and acetone, and embedded in epon-araldite resin. Following a 24-h polymerization in a 60 °C oven, 0.1 µM ultrathin sections were prepared and then post stained with lead citrate and uranyl acetate. Electron micrographs were acquired using a JEOL 1400 Plus transmission electron microscope (JEOL, Inc., Peabody, MA) at 80 kV equipped with a Gatan Orius camera (Gatan, Inc., Warrendale, PA). The EM images (Fig. 6a–c) showed large areas of disorganized myofibrils oriented in multiple planes of direction, often trapping many nemaline rods, corresponding to the large tangles with rods seen on light microscopy. Occasionally nemaline rods were observed in nuclei (Fig. 6d).
Genomic sequencing
Whole exome next generation sequencing, performed in a commercial laboratory (GeneDx, Gaithersburg, Maryland, USA) identified two heterozygous variants in MYH2. The first MYH2 variant, c.3331C>T, p.Gln1111*, is predicted to result in truncation of the proximal tail region of the MyHC-IIA protein and is therefore predicted to be pathogenic. This variant has an extremely low allele frequency (3.98e−6, 1/251326) in large genetic databases (gnomAD) and shows possibly damaging (PolyPhen-2), damaging (SIFT, FATHMM, and fathmm-MKL), deleterious (LRT), disease causing (MutationTaster) effect through in silico analysis with a Combined Annotation-Dependent Depletion (CADD) score of 39. The second MYH2 variant, c.1546T>G, p.Phe516Val, affects a highly conserved amino acid within the highly conserved catalytic motor head relay loop. MYH2 c.1546T>G, p.Phe516Val is reported to have a very low allele frequency of 0.00000398 in large genetic databases (gnomAD, 1000 Genomes, and Exome Variant Server [EVS]), and shows probably damaging (PolyPhen-2), damaging (SIFT, FATHMM, and fathmm-MKL), unknown (LRT), and disease-causing (MutationTaster) effect through in silico analysis, with a Combined Annotation-Dependent Depletion (CADD) score of 25. Targeted Sanger sequencing showed that patient’s asymptomatic mother carries the c.1546T>G, p.Phe516Val variant, suggesting that the c.3331C>T, p.Gln1111* either arose from the asymptomatic deceased father or is de novo. The patient was also found to carry two heterozygous variants of unknown significance in MYO9A (c.5620A>C, p.Lys1874Gln and c.1148C>T, p.Ser383Phe), both detected also in the asymptomatic mother. No pathogenic or potentially pathogenic variants were detected in other genes so far known to cause neuromuscular diseases.
Western blot
Western blot analysis for MyHC-IIA protein concentrations were performed on total protein isolated from the proband patient’s muscle biopsy and muscle from an age-matched normal control subject. Protein lysates from 15 mg (dry weight) of muscle tissue were run at a concentration of 2.0 ug/ul per lane in quadriplicate using a 66–440 kDa separation module on a Wes Simple platform (ProteinSimple, San Jose, California, USA), according to the manufacturer’s protocol. MyHC-IIA protein bands were identified with mouse anti-human A4.74 antibody (1:10 dilution) as developed by Blau HM and colleagues [35] and obtained from the Developmental Studies Hybridoma Bank. MyHC-IIA band intensities were normalized to human vinculin protein (mouse anti-human antibody, 1:50 dilution) (R&D Systems, Minneapolis, Minnesota USA) in multiplexing analysis using the anti-mouse Luminol-S detection module and the Compass software (ProteinSimple). Mean MyHC-IIA protein illuminance in the affected patient sample, normalized to that of vinculin in each lane was reduced to 1.21 ± 0.15% (mean ± SEM) of that within the normal control sample (100.00 ± 9.15%) (Fig. 7a).
MyHC-IIA protein structure–function modeling
Protein structure homology modeling using the X-ray structure of cow cardiac myosin as template predicts that the p.Phe516Val substitution in MyHC-IIA destabilizes the motor head relay loop domain. Phe516 is located in a hydrophobic cavity, where it directly contacts Ile710 (Fig. 7b). With a valine in place of a phenylalanine at position 516, the hydrophobic cavity is disrupted and the loop domain becomes less compact. As the relay loop domain acts as a fulcrum that regulates conformational change between high and low affinity states, this structural change may in turn affect the angular and rotational movements of the motor head converter, actin-binding region and ATP binding sites.