Multiple Slits regulate the development of midline glial populations and the corpus callosum.

The Slit molecules are chemorepulsive ligands that regulate axon guidance at the midline of both vertebrates and invertebrates. In mammals, there are three Slit genes, but only Slit2 has been studied in any detail with regard to mammalian brain commissure formation. Here, we sought to understand the relative contributions that Slit proteins make to the formation of the largest brain commissure, the corpus callosum. Slit ligands bind Robo receptors, and previous studies have shown that Robo1(-/-) mice have defects in corpus callosum development. However, whether the Slit genes signal exclusively through Robo1 during callosal formation is unclear. To investigate this, we compared the development of the corpus callosum in both Slit2(-/-) and Robo1(-/-) mice using diffusion magnetic resonance imaging. This analysis demonstrated similarities in the phenotypes of these mice, but crucially also highlighted subtle differences, particularly with regard to the guidance of post-crossing axons. Analysis of single mutations in Slit family members revealed corpus callosum defects (but not complete agenesis) in 100% of Slit2(-/-) mice and 30% of Slit3(-/-) mice, whereas 100% of Slit1(-/-); Slit2(-/-) mice displayed complete agenesis of the corpus callosum. These results revealed a role for Slit1 in corpus callosum development, and demonstrated that Slit2 was necessary but not sufficient for midline crossing in vivo. However, co-culture experiments utilising Robo1(-/-) tissue versus Slit2 expressing cell blocks demonstrated that Slit2 was sufficient for the guidance activity mediated by Robo1 in pre-crossing neocortical axons. This suggested that Slit1 and Slit3 might also be involved in regulating other mechanisms that allow the corpus callosum to form, such as the establishment of midline glial populations. Investigation of this revealed defects in the development and dorso-ventral positioning of the indusium griseum glia in multiple Slit mutants. These findings indicate that Slits regulate callosal development via both classical chemorepulsive mechanisms, and via a novel role in mediating the correct positioning of midline glial populations. Finally, our data also indicate that some of the roles of Slit proteins at the midline may be independent of Robo signalling, suggestive of additional receptors regulating Slit signalling during development.

Molecular genetics of long QT syndrome.

Long QT syndrome (LQTS) is a cardiac disorder associated with sudden death especially in young, seemingly healthy individuals. It is characterised by abnormalities of the heart beat detected as lengthening of the QT interval during cardiac repolarisation. The incidence of LQTS is given as 1 in 2000 but this may be an underestimation as many cases go undiagnosed, due to the rarity of the condition and the wide spectrum of symptoms. Presently 12 genes associated with LQTS have been identified with differing signs and symptoms, depending on the locus involved. The majority of cases have mutations in the KCNQ1 (LQT1), KCNH2 (LQT2) and SCN5A (LQT3) genes. Genetic testing is increasingly used when a clearly affected proband has been identified, to determine the nature of the mutation in that family. Unfortunately tests on probands may be uninformative, especially if the defect does not lie in the set of genes which are routinely tested. Novel mutations in these known LQTS genes and additional candidate genes are still being discovered. The functional implications of these novel mutations need to be assessed before they can be accepted as being responsible for LQTS. Known epigenetic modification affecting KCNQ1 gene expression may also be involved in phenotypic variability of LQTS. Genetic diagnosis of LQTS is thus challenging. However, where a disease associated mutation is identified, molecular diagnosis can be important in guiding therapy, in family testing and in determining the cause of sudden cardiac death. New developments in technology and understanding offer increasing hope to families with this condition.

Mutations at KCNQ1 and an unknown locus cause long QT syndrome in a large Australian family: implications for genetic testing.

A large Australian family affected with long QT syndrome (LQTS) was studied. The medical characteristics of the 16 clinically affected members were consistent with LQT1. A previously identified mutation in KCNQ1 was found in 12 affected individuals and 1 unaffected infant but absent in 4 affected family members. A haplotype consisting of specific alleles for microsatellites flanking in KCNQ1 was associated with the mutation. This was absent from the four affected individuals without the mutation, who had three different haplotypes in this region, indicating that LQTS is unlikely to be segregating with KCNQ1 in these anomalous family members. A genome scan revealed 12 regions where all four of these individuals shared alleles. One region on chromosome 21 contained the KCNE1, KCNE2, KCNJ6, and KCNJ15 genes. A common variant of KCNE1 was segregating in the family but did not explain the anomalous cases. A candidate region on chromosome 7 contained the AKAP9 and KCND2 genes. A previously reported mutation in the N-terminal Yotiao region of AKAP9 was absent from the family. No evidence was found implicating any other known or suspected LQTS gene. This family shows that there remain unidentified genetic causes of LQTS which are clinically significant and highlights the difficulties associated with genetic testing in LQTS, since we cannot rule out risk in individuals who are negative for the known mutation in KCNQ1 without knowing the second disease locus.

Experimental and bioinformatic characterisation of the promoter region of the Marfan syndrome gene, FBN1.

Mutations in the FBN1 gene, encoding the extracellular matrix protein fibrillin-1, result in the dominant connective tissue disease Marfan syndrome. Marfan syndrome has a variable phenotype, even within families carrying the same FBN1 mutation. Differences in gene expression resulting from sequence differences in the promoter region of the FBN1 gene are likely to be involved in causing this phenotypic variability. In this report, we present an analysis of FBN1 transcription start site (TSS) use in mouse and human tissues. We found that transcription of FBN1 initiated primarily from a single CpG-rich promoter which was highly conserved in mammals. It contained potential binding sites for a number of factors implicated in mesenchyme differentiation and gene expression. The human osteosarcoma line MG63 had high levels of FBN1 mRNA and secreted fibrillin-1 protein to form extracellular matrix fibres. The human embryonic kidney line HEK293 and two breast cancer lines MCF7 and MDA-MB-231 had levels of FBN1 mRNA 1000 fold lower and produced negligible amounts of fibrillin-1 protein. Therefore MG63 appears to be the optimal cell line for examining tissue-specific, biologically relevant promoter activity for FBN1. In reporter assays, the conserved promoter region was more active in MG63 cells than in non-FBN1-expressing lines but additional elements outside the proximal promoter are probably required for optimal tissue-specific expression. Understanding the regulation of the FBN1 gene may lead to alternative therapeutic strategies for Marfan syndrome.