A nationwide survey on Marinesco-Sjögren syndrome in Japan

To elucidate the clinical characteristics of MSS, we conducted a nationwide, questionnaire-based survey in Japan. The first set of questionnaires, which focused on the experience of treating patients suspected of having MSS, was sent to a total of 5,452 Japanese specialists for neurology or pediatric neurology. The second set of questionnaires, which focused on the clinical information of patients with suspected MSS, was sent to their attending physicians. These patients’ medical records were carefully reviewed by 2 of our specialists (M.G., H.K.).

Biopsied muscle specimens were flash-frozen in isopentane cooled in liquid nitrogen. Transverse serial frozen sections of 10 μm in thickness were subjected to various types of histochemical staining, including hematoxylin and eosin (H&E), modified Gomori-trichrome (mGT), and ATPases. We obtained biopsied skeletal muscles from a total of 17 unrelated patients clinically suspected of having MSS.

Genomic DNA was extracted from either frozen muscle or peripheral blood lymphocytes using standard protocols. The PCR primers were designed to amplify all the exons of SIL1 together with their flanking intronic regions. The primer sequences are available upon request. Direct sequencing was performed using the BigDye Terminator v3.1 Cycle Sequencing system and an ABI3100 automated Genetic Analyzer (Applied Biosystems, Foster City, CA). The sequence data obtained were analyzed using the SeqScape (Applied Biosystems) program and compared with the genomic sequence of SIL1 in the database (NM_022464). Genetic analysis was performed on DNA from 27 unrelated patients and the parents of 3 of them. Two hundred control chromosomes from healthy individuals were examined for each novel mutation in SIL1 by direct sequencing.

For the haplotype analysis of the SIL1 genomic region, we performed direct sequencing of the following 11 common single nucleotide variants (SNVs) in the Japanese population: rs11748097, rs929775, rs10045761, rs1433008, rs11958050, rs7717375, rs7722413, rs3763016, rs6596456, rs700629, and rs12653845 (http://​www.​ncbi.​nlm.​nih.​gov/​SNP/​). Samples from 21 patients homozygous for the common c.936dupG mutation, the parents of 3 patients, and 92 control Japanese individuals were analyzed.

A total of 1,875 responses (34.4% response rate) were received to the first set of questionnaires. The second set of questionnaires was sent to a total of 37 attending physicians (2.0%) who had treated patients suspected of having MSS. The detailed clinical records of a total of 36 patients were carefully reviewed. The shortfall in the number of patients was due to an overlap of 1 case.

Frozen muscle or peripheral blood lymphocytes were obtained from 27 of the 36 patients suspected of having MSS, for genetic analysis. Mutations in SIL1 were identified in 24 out of the 27 patients. Twenty-one patients were homozygous for the previously reported c.936dupG (p.Leu313fs) mutation in exon 9. Patient 4 was homozygous for the previously reported c.603_607del5 (p.Glu201fs) mutation in exon 6 [11]. Patient 12 was homozygous for the previously reported c.331C > T (p.Arg111X) mutation in exon 4 [6, 7]. Patient 17 was a compound heterozygote for the novel c.617_618TC > AA (p.Leu206Glu) mutation in exon 6 and the c.936dupG (p.Leu313fs) mutation in exon 9 [12]. All mutations except for p.Leu206Glu are predicted to induce premature termination. Mutation of Leu206, which is highly conserved among species, was predicted to exert a deleterious impact on protein function by the Polymorphism Phenotyping v2 (PolyPhen-2; http://​genetics.​bwh.​harvard.​edu/​pph2/​) and the Sorting Intolerant From Tolerant software (SIFT: http://​sift.​bii.​a-star.​edu.​sg/​). None of these nucleic acid changes were found in the 200 chromosomes from healthy Japanese controls or in the Japanese Single Nucleotide Polymorphisms database (http://​snp.​ims.​u-tokyo.​ac.​jp/​index_​ja.​html).

The SIL1 c.936dupG mutation was identified in 43 of the 48 chromosomes (89.6%) in our cohort. The patients carrying this mutation were reported from different areas in Japan. To determine whether this was a result of a founder effect, we performed a haplotype analysis using 11 SNVs within or close to SIL1. The results revealed that 31 of the 32 chromosomes (96.9%) with the c.936dupG mutation had the same haplotype (P1-P24, Table 1). This haplotype was only found in 18 of the 184 chromosomes (9.8%) from the control group, suggesting a founder effect, although 1 chromosome from Patient 1 had a different haplotype.

Table 2 and Additional file 1: Table S1 show a clinical summary of the patients in our series.

The age at which the examination was performed in the 24 patients (10 men and 14 women) with SIL1 mutations varied from 2 to 52 (mean = 20.1 ± 18.1) years. Bilateral cataracts requiring prompt surgical intervention had appeared and rapidly progressed in all 24 patients at the mean age of 3.5 ± 1.2 years. Strabismus was also observed in 55.6% (10/18) of the patients. Cerebellar signs included hypotonia (21/24; 88%), ataxia (16/24; 67%), nystagmus (11/24; 46%), and dysarthria (8/24; 33%) were seen. Brain MRI demonstrated marked atrophy of the cerebellum, particularly the vermis, in all the patients examined (19/19). Mild to moderate intellectual disability, diagnosed by intelligence quotient/developmental quotient between 35 and 70, was seen in 91% (20/22) of the patients. Acquisition of meaningful words occurred at the age of 2.0 ± 0.8 years, and most of the patients had required special-needs education. Muscle weakness was observed in 95% (21/22) of the patients, with delays in motor milestones. Head control was first seen in all the patients at a certain time point between 4 and 18 (mean = 7.8 ± 3.7) months, and sitting at a certain time point between 10 and 36 (mean = 20.0 ± 8.7) months. Eighty percent (16/20) of the patients could stand with support at a certain point between the ages of 1 and 4 years, and walk with support at a certain point between the ages of 2 and 22 (mean = 5.8 ± 2.6) years; however, none of the patients acquired the ability to walk independently. Muscle weakness was slowly progressive and predominantly in the proximal muscles, with the patients becoming wheelchair-bound at a certain time point between the ages of 13 and 28 (mean = 17.4 ± 6.3) years. Serum creatine kinase levels were normal to moderately elevated (28–2000, mean = 389 ± 464; normal < 200 IU/L). Short stature (< -2 SD) was seen in 67% (12/18, mean = -3.6 SD) of the patients, and spinal deformity (8/22; 36%), flat foot (7/22; 32%), and short fingers (5/22; 23%) were also reported. Hypogonadotropic hypogonadism was seen in 3 of 8 (38%) patients (1 with microtestis, 2 with amenorrhea). No marked clinical differences were observed among patients with different SIL1 mutations. No patient had cardiac and respiratory problems.

The 3 patients with no SIL1 mutation (Patients 25, 26, and 27) and the 9 genetically unexamined patients showed clinical features indistinguishable from the patients with SIL1 mutations, including cerebellar signs with cerebellar atrophy on brain images, intellectual disability, congenital cataracts, and muscle weakness. Elevation of serum CK levels was also seen in 2 patients (Table 2).

Biopsied skeletal muscles were obtained from 16 patients with SIL1 mutations and one patient without (Patient 27). All muscle specimens showed myopathic changes of variation in fiber size and endomysial fibrosis. A few necrotic and regenerating fibers were seen in some patients with SIL1 mutations. No neurogenic changes, including fiber type grouping and grouped atrophy, were observed in any of the patients. Importantly, scattered rimmed vacuoles (RVs) were seen in all 16 patients with SIL1 mutations, but not in the patient without (Figure 1).