Expanded ATXN3 frameshifting events are toxic in Drosophila and mammalian neuron models.

Spinocerebellar ataxia type 3 is caused by the expansion of the coding CAG repeat in the ATXN3 gene. Interestingly, a -1 bp frameshift occurring within an (exp)CAG repeat would henceforth lead to translation from a GCA frame, generating polyalanine stretches instead of polyglutamine. Our results show that transgenic expression of (exp)CAG ATXN3 led to -1 frameshifting events, which have deleterious effects in Drosophila and mammalian neurons. Conversely, transgenic expression of polyglutamine-encoding (exp)CAA ATXN3 was not toxic. Furthermore, (exp)CAG ATXN3 mRNA does not contribute per se to the toxicity observed in our models. Our observations indicate that expanded polyglutamine tracts in Drosophila and mouse neurons are insufficient for the development of a phenotype. Hence, we propose that -1 ribosomal frameshifting contributes to the toxicity associated with (exp)CAG repeats.

HMSN/ACC truncation mutations disrupt brain-type creatine kinase-dependant activation of K+/Cl- co-transporter 3.

The potassium-chloride co-transporter 3 (KCC3) is mutated in hereditary motor and sensory neuropathy with agenesis of the corpus callosum (HMSN/ACC); however, the molecular mechanisms of HMSN/ACC pathogenesis and the exact role of KCC3 in the development of the nervous system remain poorly understood. The functional regulation of this transporter by protein partners is also largely unknown. Using a yeast two-hybrid approach, we discovered that the C-terminal domain (CTD) of KCC3, which is lost in most HMSN/ACC-causing mutations, directly interacts with brain-specific creatine kinase (CK-B), an ATP-generating enzyme that is also a partner of KCC2. The interaction of KCC3 with CK-B was further confirmed by in vitro glutathione S-transferase pull-down assay, followed by sequencing of the pulled-down complexes. In transfected cultured cells, immunofluorescence labeling showed that CK-B co-localizes with wild-type KCC3, whereas the kinase fails to interact with the inactive truncated KCC3. Finally, CK-B's inhibition by DNFB results in reduction of activity of KCC3 in functional assays using Xenopus laevis oocytes. This physical and functional association between the co-transporter and CK-B is, therefore, the first protein-protein interaction identified to be potentially involved in the pathophysiology of HMSN/ACC.

Molecular genetic studies of DMT1 on 12q in French-Canadian restless legs syndrome patients and families.

Converging evidence from clinical observations, brain imaging and pathological findings strongly indicate impaired brain iron regulation in restless legs syndrome (RLS). Animal models with mutation in (DMT1) divalent metal transporter 1 gene, an important brain iron transporter, demonstrate a similar iron deficiency profile as found in RLS brain. The human DMT1 gene, mapped to chromosome 12q near the RLS1 locus, qualifies as an excellent functional and possible positional candidate for RLS. DMT1 protein levels were assessed in lymphoblastoid cell lines from RLS patients and controls. Linkage analyses were carried out with markers flanking and within the DMT1 gene. Selected patient samples from RLS families with compatible linkage to the RLS1 locus on 12q were fully sequenced in both the coding regions and the long stretches of UTR sequences. Finally, selected sequence variants were further studied in case/control and family-based association tests. A clinical association of anemia and RLS was further confirmed in this study. There was no detectable difference in DMT1 protein levels between RLS patient lymphoblastoid cell lines and normal controls. Non-parametric linkage analyses failed to identify any significant linkage signals within the DMT1 gene region. Sequencing of selected patients did not detect any sequence variant(s) compatible with DMT1 harboring RLS causative mutation(s). Further studies did not find any association between ten SNPs, spanning the whole DMT1 gene region, and RLS affection status. Finally, two DMT1 intronic SNPs showed positive association with RLS in patients with a history of anemia, when compared to RLS patients without anemia.

NLGN3/NLGN4 gene mutations are not responsible for autism in the Quebec population.

Jamain [2003: Nat Genet 34:27-29] recently reported mutations in two neuroligin genes in sib-pairs affected with autism. In order to confirm these causative mutations in our autistic population and to determine their frequency we screened 96 individuals affected with autism. We found no mutations in these X-linked genes. These results indicate that mutations in NLGN3 and NLGN4 genes are responsible for at most a small fraction of autism cases and additional screenings in other autistic populations are needed to better determine the frequency with which mutations in NLGN3 and NLGN4 occur in autism.