The limb-girdle muscular dystrophies are a genetically heterogeneous group of inherited progressive muscle disorders that affect mainly the proximal musculature, with evidence for at least three autosomal dominant and eight autosomal recessive loci. The latter mostly involve mutations in genes encoding components of the dystrophin-associated complex; another form is caused by mutations in the gene for the muscle-specific protease calpain 3. Using a positional cloning approach, we have identified the gene for a form of limb-girdle muscular dystrophy that we previously mapped to chromosome 2p13 (LGMD2B). This gene shows no homology to any known mammalian gene, but its predicted product is related to the C. elegans spermatogenesis factor fer-1. We have identified two homozygous frameshift mutations in this gene, resulting in muscular dystrophy of either proximal or distal onset in nine families. The proposed name 'dysferlin' combines the role of the gene in producing muscular dystrophy with its C. elegans homology.

Limb-girdle muscular dystrophies (LGMD) are a heterogeneous group of inherited neuromuscular disorders characterized by proximal muscular weakness of the pelvic and shoulder girdles and a variable progression with symptoms, ranging from very severe to mild. One autosomal dominant (LGMD1A, at chromosome 5q22.3-31.3) (ref. 3) and five autosomal recessive (AR) loci responsible for this phenotype have been identified: LGMD2A at 15q (ref. 4); LGMD2B at 2p (ref. 5), LGMD2C at 13q (ref. 6), LGMD2D at 17q (ref. 7) and LGMD2E at 4q (refs 8,9). In the muscle membrane, dystrophin associates with several proteins and glycoproteins organized in two main subcomplexes: the dystroglycan (DG) and sarcoglycan (SG) complexes. The genes for LGMD2C, LGMD2D and LGMD2E code for proteins of the SG complex. We recently mapped a sixth AR form of LGMD, LGMD2F, to chromosome 5q33-34 in two Brazilian families. In the same chromosomal interval we also mapped the delta SG gene, encoding a novel 35-kD component of the sarcoglycan (SG) complex. We now show that a homozygous mutation in the delta SG gene (a single nucleotide deletion that alters its reading frame) is the cause of LGMD2F.

The dystrophin-glycoprotein complex (DGC) serves as a link between cytoplasmic actin, the membrane and the extracellular matrix of striated muscle. Genetic defects in genes encoding a subset of DGC proteins result in muscular dystrophy and a secondary decrease in other DGC proteins. Caveolae are dynamic structures that have been implicated in a number of functions including endocytosis, potocytosis and signal transduction. Caveolin (VIP-21) is thought to play a structural role in the formation of non-clathrin-coated vesicles in a number of different cell types. Caveolin-3, or M-caveolin, was identified as a muscle-specific form of the caveolin family. We show that caveolin-3 co-purifies with dystrophin, and that a fraction of caveolin-3 is a dystrophin-associated protein. We isolated the gene for human caveolin-3 and mapped it to chromosome 3p25. We determined the genomic organization of human caveolin-3 and devised a screening strategy to look for mutations in caveolin-3 in patients with muscular dystrophy. Of 82 patients screened, two nucleotide changes were found that resulted in amino acid substitutions (G55S and C71W); these changes were not seen in a control population. The amino acid changes map to a functionally important domain in caveolin-3, suggesting that these are not benign polymorphisms and instead are disease-causing mutations.

Autosomal recessive limb-girdle muscular dystrophies (LGMDs) are genetically heterogeneous. A subgroup of these disorders is caused by mutations in the dystrophin-associated sarcoglycan complex. Truncating mutations in the 43 kDa beta-sarcoglycan gene (LGMD 2E) were originally identified in a sporadic case of Duchenne-like muscular dystrophy, and a common missense mutation (T151R) was identified independently in Indiana Amish pedigrees with a milder form of LGMD. To facilitate mutational analysis of larger numbers of patients directly from genomic DNA, as opposed to reverse transcribed RNA from muscle biopsies, we have determined the genomic structure of the beta-sarcoglycan gene. The open reading frame of the beta-sarcoglycan coding region extends over six exons. Primers were designed for PCR amplification of single exons from genomic DNA and subsequent single strand conformation polymorphism (SSCP) analysis. We screened 15 patients from the Brazilian LGMD patient population, 13 of whom followed a severe course. Most of the patients had been assessed previously for deficiency of alpha-sarcoglycan immunofluorescence on muscle biopsy sections as a marker for disease of the sarcoglycan complex. Novel mutations in two familial and two sporadic cases of severe childhood-onset LGMD were identified. Only one of these patients carried a truncating mutation (homozygous 2 bp deletion, FS164TER), while the other three carried missense mutations (homozygous R91P, homozygous M100K, heterozygous recessive L108R; only one allele could be identified in this family). All three missense mutations occurred in exon 3, coding for the immediate extracellular domain. Complete absence for all three of the known sarcoglycans was noted by immunohistochemistry on muscle biopsy sections of the patients.

Our studies show that LiCl can modulate Gsk3b transcription in vitro and in vivo. This observation suggest new regulatory effects of lithium over Gsk3b, contributing to the better understanding of its mechanisms of action, offering a new and complementary explanation for Gsk3b modulation and reinforcing its potential for the inhibition of key pathological pathways in Alzheimer's disease.