We report a family with five siblings affected with HIBM2, which we diagnosed primarily by linkage analysis and identification of mutations in GNE
. Recently, we used a similar approach to achieve a molecular diagnosis for a family with multiminicore disease, another rare myopathy that can be mistaken for LGMD [24
]. The exclusion in our linkage scan of almost all known myopathy genes [25
], despite fourteen regions showing linkage at a LOD score of approximately 1.4, confirms the value of linkage analysis as a diagnostic tool even in families not informative enough to generate suggestive or significant LOD scores.
Several features of the clinical and pathological presentation of the family made the diagnosis initially uncertain. First, the proband came to medical attention prior to the identification of mutations in GNE
as the cause of HIBM, so genetic testing for HIBM was not then available. Moreover, the symptoms were suggestive of LGMD, and LGMD is much more prevalent than HIBM [26
]. HIBM is predominantly found among Middle Eastern Jewish families [5
] and Japanese families (as DMRV/NM) [10
]; it appears to be much rarer among other populations [33
]. Only a few previous families of Asian Indian ancestry have been documented with mutations in GNE
]. Furthermore, an initial biopsy of the quadriceps of the proband appeared normal, which in retrospect is consistent with a quadriceps-sparing myopathy, but may have initially confounded diagnosis. A later biopsy revealed the pathologic hallmarks of HIBM; however, the characteristic rimmed vacuoles are not specific to HIBM [2
] and were negative for amyloid by Congo red staining, which is atypical for HIBM [34
]. Finally, because five of seven siblings were affected and their father appeared symptomatic, autosomal dominant inheritance was plausible, provided that the father's much milder symptoms could be explained by variable expressivity. Therefore, both dominant and recessive muscular dystrophies and myopathies were potential candidates.
Our linkage scans were performed under a recessive model, and our identification of compound heterozygous mutations in affected subjects confirmed this mode of inheritance. However, the father (Subject 3) was heterozygous for the c.1816+5G>A splice site mutation, and at age 79 he displayed slight weakness in his lower extremities similar in its pattern to his affected children. The predicted deletion of 61 amino acids (p.G545_D605del) in the kinase domain of GNE would be expected to result in an extremely hypomorphic or null allele, and thus the weakness in Subject 3 could possibly be a result of haploinsufficiency of GNE
. Mutations showing incomplete dominance have been documented for other recessive myopathies [35
]. However, without additional clinical examination and biopsy tissue we could not determine if the symptoms in Subject 3 represented a milder version of HIBM or normal age-related weakness. Alternatively, an unknown enhancer mutation in a modifier gene, or an effect of gender on the manifestation of the splice mutation, could potentially account for the myopathic symptoms in Subject 3 as well as the faster deterioration of his affected son than his four affected daughters.
The mechanism whereby mutations in GNE
result in muscle weakness is not completely understood [38
]. Approximately 80 different mutations in GNE
have been previously reported [18
], of which only two are in splice sites. Like the c.1816+5G>A mutation described herein, these mutations likely result in null alleles [10
]. The vast majority of mutations in GNE
are missense [33
], and no patients have been recorded as homozygous or compound heterozygous for two null mutations, suggesting a complete loss of function of GNE
might be lethal in humans [5
], as in mice [44
]. Accordingly, the p.V696M mutation reported both here and previously [5
], often in conjunction with very severe mutations, may be relatively benign, as otherwise the patients might not retain enough residual GNE
activity to permit life.