10 May 2005
Rett syndrome (RTT) is a complex and serious neurological disorder that affects at least one in every 10,000 female births, making it the second most common cause of severe and profound learning disability in girls; clinical features appear 6-18 months after birth. A very small number of boys are also affected by the condition. In the majority of cases, Rett syndrome is caused by heterozygous mutations in the MECP2 gene on the X chromosome. Methyl-CpG binding protein 2 (Mecp2) is thought to selectively bind methyl-CpG islands in the mammalian genome and to function as a repressor of gene expression. It has previously been postulated that Mecp2 functions in imprinting, a process whereby certain genes are chemically modified (typically by methylation of CpG islands) according to whether they are maternal or paternal in origin. This is important because for normal development and function these genes are expressed primarily from the maternal or paternal alleles. Disruption of normal imprinting can cause disease due to altered and inappropriate gene expression; mutations in the MECP2 gene have been suggested to act in this way. Previous studies failed to find any loss of imprinting in RTT, looking at well known imprinted genes, but a new paper in Nature Genetics reports the loss of imprinting of the DLX5 gene in mouse models of RTT and human cell lines derived from individuals with the disease [Horike S et al. (2005) Nat. Gen. 37, 31-40].
The authors searched for Mecp2 target genes in the mouse brain by identifying and sequencing genomic Mecp2 binding sites (MBSs), working on the assumption that target genes would be close to such binding sites. 100 binding sites identified by chromatin immunoprecipitation, were sequenced; two MBSs were found to map to a cluster of imprinted genes on chromosome 6 (homologous to a region on human chromosome 7) which included the genes Dlx5 and Dlx6. To determine whether Mecp2 deficiency could cause dysregulation of these (and other) genes, the researchers used quantitative RT-PCR to compare their expression in the brains of wild-type and Mecp2 knockout mice. Expression of the Dlx5 and Dlx6 genes was found to be around two times higher in Mecp2-null mice than in normal mice.
To investigate whether this increase in Dlx5 expression was due to altered imprinting, Mecp2-null and control mouse lines were created with a single-nucleotide polymorphism (SNP) in Dlx5, which was used as a marker to monitor parental allele-specific transcription; there was no equivalent SNP available for analysis of the imprinted status of Dlx6. In the absence of Mecp2 there was complete loss of normal maternal imprinting of the Dlx5 gene in the mouse brain. The human gene, DLX5, is imprinted in normal human brain cells and lymphoblasts; the authors therefore examined lymphoblastoid cells from individuals with RTT who had mutations in MECP2 and found loss of imprinting in three out of four such samples.
Next, the researchers examined the methylation status of CpG islands in the DLX5 gene in both humans and mice, and found that they were unmethylated, suggesting that, despite the ability of Mecp2 to bind to these islands, regulation of DLX5 expression must be mediated by a different mechanism. A key factor in gene silencing and activation is chromatin structure; higher-order loops are associated with chromatin silencing. The authors investigated the role of Mecp2 in the formation of higher-order loops using a chromosome conformation capture technique combined with primers to test for MBSs brought together to form loops. They demonstrated that Mecp2 mediates the formation of an 11kb higher-order chromatin loop at the Dlx5-Dlx6 locus associated with silencing of the Dlx5 gene, via interaction with histone deacetylase 1 (Hdac1). In Mecp2-null mice, this region loses the ability to bind Hdac1 whereas in wild-type mice, this region is associated with a higher-order chromatin loop structure. This finding provides a potential mechanism for the loss of DLX5 imprinting in RTT.
Comment: Taken together, these results suggest a novel role for the loss of imprinting of the Mecp2 target gene DLX5 in the pathology of Rett syndrome, and provide avenues for further research into the underlying molecular mechanisms of the disease. DLX genes are expressed by neurons that respond to the neurotransmitter GABA in the brain; GABA neurotransmission has been linked to other neurodevelopmental disorders, including Angelman syndrome, which shares some clinical features with RTT and is also associated with a loss of imprinting. A degree of cross-regulation between different members of the DLX gene family has been previously reported; the authors postulate that small increases in levels of DLX5 expression due to a loss of maternal imprinting could significantly affect levels of other DLX genes leading to neurological pathology. A commentary accompanying the report queries whether DLX5 is in fact the key factor in pathogenesis of RTT, suggesting that it is probably only one of several gene targets regulated by Mecp2 [Pescucci C et al. (2005) Nat. Gen. 37, 10-11]. However, the targets and mechanisms of Mecp2 mediated gene expression are likely to be an important research area for unravelling the cause of RTT.