Centronuclear myopathy related to dynamin 2 mutations: clinical, morphological, muscle imaging and genetic features of an Italian cohort

Abstract

Mutations in dynamin 2 (DNM2) gene cause autosomal dominant centronuclear myopathy and occur in around 50% of patients with centronuclear myopathy. We report clinical, morphological, muscle imaging and genetic data of 10 unrelated Italian patients with centronuclear myopathy related to DNM2 mutations. Our results confirm the clinical heterogeneity of this disease, underlining some peculiar clinical features, such as severe pulmonary impairment and jaw contracture that should be considered in the clinical follow-up of these patients. Muscle MRI showed a distinct pattern of involvement, with predominant involvement of soleus and tibialis anterior in the lower leg muscles, followed by hamstring muscles and adductor magnus at thigh level and gluteus maximus. The detection of three novel DNM2 mutations and the first case of somatic mosaicism further expand the genetic spectrum of the disease.

Fig. 1

Muscle MRI, T1-weighted images in 7 patients ranging from mild to severe muscular involvement. Age at MRI is reported for each patient. Earliest changes were observed in distal leg muscles with prominent involvement of soleus (#2, 6) and tibialis anterior muscles (#3, 6) followed by peroneal and gastrocnemius muscles (#2, 3). Tibialis posterior muscle was relatively spared except in more severely affected patients (#1, 7). At the level of the thigh posterior compartment was first involved: earliest changes were observed in biceps femoris, semitendinous and semimembranous muscles followed by adductor magnus (#6, 3, 5). Patients 2 and 7 showed a more severe involvement of the anterior compartment of the thigh with extensive fatty substitution of vastus lateralis, vastus intermedius and vastus medialis muscles. Rectus femoris, sartorius and gracilis were relatively preserved even in more severely affected patient (#1, 2, 5, 7). At the pelvic level, gluteus maximus was first involved followed by gluteus medius and minimus (#2, 3). A complete fatty infiltration of gluteal and paraspinal muscles was found in more severely affected patients and who had lost the ability to walk (#5, 1).

Fig. 2

Schematic diagram of the typical pattern of muscle involvement in DNM2-related CNM. In blue are represented the most affected muscles, in red the relatively spared muscles. A: In the leg there is a prominent involvement of soleus (So) and tibialis anterior (TA). B: In the thigh, biceps femoris (BF), semitendinosus (St) and adductor magnus (AM) are early affected, whereas gracilis (Gr), sartorius (S) and rectus femoris (RF) are relatively spared. C: In the pelvis gluteus maximum (GM) is more affected.

Fig. 3

Muscle biopsy from patient 9. Light microscopy of sections stained with H&E (A, B); NADHTR (C, D, E), SDH (F, G), COX (H, I, J). The fiber size variability with small type 1 fibers and the low percentage of centralized nuclei are showed in A and C. Atypical ‘necklace’ fibers: a slightly basophilic ring on H&E (A) and reactive with oxidative reactions (C, F, H) was evident underneath the sarcolemma and well represented at high magnification (B, D, E, G, I, J). In those fibers myonuclei appeared in the normal subsarcolemmal position.

Fig. 4

Electropherogram showing the DNM2 tandem duplication (c.1872dupAGCTGG) in patient 6 (A), the low level of mutation in his mosaic father (B) and wild-type mother (C). Evidence of the six base pairs sequence (AGCTGG) repetition along the exon 16 of DNM2 which could be responsible of nonallelic homologous recombination (NAHR) causing the DNM2 tandem duplication (c.1872dupAGCTGG) in patient 6 (D).

Fig. 5

SNaPshot obtained from blood of patient 6 and from blood, saliva, hair and urine of his asymptomatic father with evidence of somatic mosaicism. The values in the boxes represent the relative peak areas for the mutant sequence.