Converging evidence that sequence variations in the novel candidate gene MAP2K7 (MKK7) are functionally associated with schizophrenia.

Schizophrenia is a debilitating psychiatric disease with a strong genetic contribution, potentially linked to altered glutamatergic function in brain regions such as the prefrontal cortex (PFC). Here, we report converging evidence to support a functional candidate gene for schizophrenia. In post-mortem PFC from patients with schizophrenia, we detected decreased expression of MKK7/MAP2K7-a kinase activated by glutamatergic activity. While mice lacking one copy of the Map2k7 gene were overtly normal in a variety of behavioural tests, these mice showed a schizophrenia-like cognitive phenotype of impaired working memory. Additional support for MAP2K7 as a candidate gene came from a genetic association study. A substantial effect size (odds ratios: ~1.9) was observed for a common variant in a cohort of case and control samples collected in the Glasgow area and also in a replication cohort of samples of Northern European descent (most significant P-value: 3 × 10(-4)). While some caution is warranted until these association data are further replicated, these results are the first to implicate the candidate gene MAP2K7 in genetic risk for schizophrenia. Complete sequencing of all MAP2K7 exons did not reveal any non-synonymous mutations. However, the MAP2K7 haplotype appeared to have functional effects, in that it influenced the level of expression of MAP2K7 mRNA in human PFC. Taken together, the results imply that reduced function of the MAP2K7-c-Jun N-terminal kinase (JNK) signalling cascade may underlie some of the neurochemical changes and core symptoms in schizophrenia.

Progressive loss of PAX6, TBR2, NEUROD and TBR1 mRNA gradients correlates with translocation of EMX2 to the cortical plate during human cortical development.

The transcription factors Emx2 and Pax6 are expressed in the proliferating zones of the developing rodent neocortex, and gradients of expression interact in specifying caudal and rostral identities. Pax6 is also involved in corticoneurogenesis, being expressed by radial glial progenitors that give rise to cells that also sequentially express Tbr2, NeuroD and Tbr1, genes temporally downstream of Pax6. In this study, using in situ hybridization, we analysed the expression of EMX2, PAX6, TBR2, NEUROD and TBR1 mRNA in the developing human cortex between 8 and 12 postconceptional weeks (PCW). EMX2 mRNA was expressed in the ventricular (VZ) and subventricular zones (SVZ), but also in the cortical plate, unlike in the rodent. However, gradients of expression were similar to that of the rodent at all ages studied. PAX6 mRNA expression was limited to the VZ and SVZ. At 8 PCW, PAX6 was highly expressed rostrally but less so caudally, as has been seen in the rodent, however this gradient disappeared early in corticogenesis, by 9 PCW. There was less restricted compartment-specific expression of TBR2, NEUROD and TBR1 mRNA than in the rodent, where the gradients of expression were similar to that of PAX6 prior to 9 PCW. The gradient disappeared for TBR2 by 10 PCW, and for NEUROD and TBR1 by 12 PCW. These data support recent reports that EMX2 but not PAX6 is more directly involved in arealization, highlighting that analysis of human development allows better spatio-temporal resolution than studies in rodents.

A physical map of the genomic region on mouse chromosome 3 containing the hindshaker (hsh) mutation.

Hindshaker (hsh), a spontaneous, autosomal recessive mouse mutation, displays a developmentally dependent tremor of the hindquarters due to hypomyelination in the CNS. This myelin deficit is followed by progressive, but incomplete, recovery by postnatal day 42. Herein we describe the construction of a genomic contig spanning the interval between the markers D3Mit187 (42.4 cM) and D3Mit232 (45.2 cM) on mouse chromosome 3, which we have previously shown to contain the hsh mutation. A physical map, covering approximately 3.5 Mb, was constructed from a series of overlapping yeast and bacterial artificial chromosomes. A 1.2- to 1.4-Mb segment central to the contig was compared extensively with the syntenic regions in human (chromosome 1q21-q23) and rat (chromosome 2). We present new data on 10 genes erroneously assigned to this area and on another 6 genes previously assigned elsewhere. For absent genes, our work suggests that they are telomeric to the region encompassed in our map. Accordingly, our findings both map the area surrounding the hsh mutation and present important corrections to the current maps in an area rich in genes related to the nervous system.

Mapping of the dysmyelinating murine Hindshaker mutation to a 1.2-cM interval on chromosome 3.

Hindshaker (hsh) is a novel, spontaneous, autosomal recessive mouse mutation displaying a myelin deficit, predominantly in the spinal cord. It is characterized by developmentally dependent hypomyelination, first evident at postnatal day (P) 10, followed by progressive but incomplete recovery by P42. Hypomyelination is associated with a decreased number of mature oligodendrocytes, which fail to form complete myelin sheaths. Heterozygotes are phenotypically normal, and the hsh mutation shows considerable variation in penetrance and expression depending on genetic background, indicating the influence of modifying loci. Here, we followed an outcross/backcross breeding strategy in conjunction with genotyping for microsatellites and a novel marker for the gene S100a4. We describe the genomic mapping of the hsh mutation to within a 1.2-cM region near the centromere of mouse chromosome 3. We found that hsh is flanked between D3Mit187 proximally and S100a4 distally. The area containing hsh is gene-rich, with a high proportion of the genes specific to nervous tissue. Identification of the hsh mutation will aid our understanding of processes important in regional control of oligodendrocyte development and myelination.