A number of genetic disorders seem to attract more than their fair share of attention from human and medical geneticists. Some diseases (such as cystic fibrosis) achieve this status because they are quite common and exact a substantial toll on both patients and the medical community. Others (such as fragile-X syndrome) do so because they present with unusual genetic or clinical features that defy conventional explanations. Discovery of the genetic and molecular basis of such disorders, therefore, promises to reveal new concepts or mechanisms of genetic disease, the significance and general interest of which extend far beyond the details of the particular disorder itself.

One such disorder is Rett syndrome (RTT), a childhood neuro-developmental disorder that affects females (almost exclusively). Its genetic basis has been difficult to establish, because most cases are sporadic. Moreover, affected girls are considered to have normal development for the first 6 to 18 months, followed by a period of regression (sometimes abrupt), marked in particular by loss of purposeful hand use and speech. Hand-wringing, ataxia and growth retardation often accompany a profound mental handicap.

X marks the Spot?

A number of models that account for the genetics of RTT have been proposed. The simplest explanation is that RTT is an X-linked dominant condition, lethal in hemizygous males. The absence, however, of a convincing deficit of males contrasts with other X-linked disorders of this type. A high proportion of de novo mutations (particularly in the paternal germ line) might account for the absence of obvious male lethality, but the incidence of RTT (an estimated 1 in 10,000) would require a very high mutation rate. The recognition of an X-linked dominant disorder with male sparing provides another model. Because of the sex-limited expression of RTT, most attempts to identify the causative gene have centered on the X chromosome. Tantalizing, but inconsistent, hints of skewed X-chromosome inactivation in females with RTT or their mothers also implicate the X. Whereas attempts to map the gene have been hindered by the frustratingly small number of familial RTT cases available, exclusion mapping based on comparison of X-chromosome haplotypes among affected sisters or half-sisters has focused attention on Xq28.

Ruthie Amir and colleagues now report the fruits of their labours over Xq28: the presence of several mutations in MECP2 in a proportion of RTT patients. The methyl CpG-binding protein 2 (MeCP2) can bind methylated DNA and has been implicated as a key player in assembling transcriptional silencing complexes. These data provide a link between the genetics of RTT and epigenetic silencing and establish RTT as the first human disease caused by defects in a protein involved in DNA methylation. Notably, they add RTT to the small, but growing, number of human genetic disorders that involve abnormal chromatin packaging and gene expression.

A relationship between chromatin structure, gene expression and DNA methylation has long been recognized, but the role of methylation in vertebrate development is poorly defined. Indeed, the only genes whose appropriate expression patterns are known to depend on methylation are those whose CpG islands need to become methylated for epigenetic silencing.

Silence the Noise

The MeCP2 protein silences methylated chromatin by recruiting a histone deacetylase complex. Unlike most other known transcriptional repressor proteins, however, the binding site of MeCP2 occurs frequently in vertebrate genomic DNA, as it requires only a single methylated CpG base pair to bind. What might be the role of such a ubiquitous transcriptional repressor? It has been proposed that MeCP2 acts as a global transcriptional repressor that prevents unscheduled transcription throughout the genome. Could the pathology of RTT patients be caused by excessive transcriptional ‘noise’ owing to a silencing defect? The fact that RTT patients do not suffer severe abnormalities during early development implies that no specific programmes of developmental gene expression (including X-chromosome inactivation) are disrupted in the absence of MeCP2. But if genomic noise is to blame, why is the brain the primary site of pathology? MeCP2 is more abundant in the brain than in any other tissue, so perhaps the brain is more sensitive to excess transcriptional noise than are other tissues. Alternatively, perhaps more MeCP2 is needed to keep noise to an acceptable level in the brain than in other tissues. But although a defect in gene silencing is a logical and exciting possibility in RTT, it remains unproven. Genes that are targets of MeCP2 (whether specific or more global) need to be identified and their possible over- or mis-expression in RTT, particularly in the nervous system, evaluated.

The finding that MECP2 is mutated in RTT fits well with what is known about MeCP2 deficiency in mice. Male mouse embryonic stem (ES) cells in which Mecp2 is disrupted cannot support development, consistent with the possible male lethality of RTT. In contrast, chimaeric mice, in which a small proportion of cells are derived from MeCP2-deficient ES cells, are viable. These animals might provide a model for RTT, as female RTT patients are also mosaic for MeCP2-expressing and MeCP2-deficient cells because of random X-chromosome inactivation. Conditional mouse mutants, in which Mecp2 is specifically disrupted in the brain, may provide further clues as to the reasons for the specific neurodevelopmental effects of MeCP2 deficiency.

MeCP2 is but one of three proteins known to both bind specifically to methylated DNA in vivo and to be capable of repressing transcription. Might other methyl-binding family members also factor in RTT, given that MECP2 mutations have been found in a proportion of patients? As the genes for other methyl-binding proteins are autosomal, mutation of them is unlikely to cause RTT, unless there are autosomal phenocopies. Nonetheless, the discovery that mutations in a gene that affects DNA methylation lead to human disease implicates the autosomal genes as candidates for other neurological disorders.

Documentation of MECP2 mutations by Amir et al. Identifies RTT as one of a small but growing number of human diseases involving abnormal chromatin assembly or remodelling, with consequent epigenetic effects on expression of one or more genes that are themselves not mutated. Other examples include patients with imprinting defects who demonstrate inappropriate gene expression due to alterations in epigenetic regulation. Similarly, patients with abnormal X chromosomes that are missing the X-inactivation centre fail to inactivate that X and therefore have functional disomy of X-linked genes. Recently, Allis and colleagues discovered a defect in phosphorylation of histone H3 in Coffin-Lowry syndrome, suggesting that its pathogenesis may be effected by global abnormalities in gene expression. So, these examples focus renewed attention on chromatin as a critical, but often overlooked, component in the cascade of regulatory mechanisms that not only underlie gene activation or silencing, but are also relevant to human disease.


Huntington F. Willard
Department of Genetics
Center for Human Genetics
Case Western Reserve University
University Hospital of Cleveland
Cleveland, Ohion 44106

Brian D. Hendrich
Institute of Cell and Molecular Biology
University of Edinburgh
Edinburgh EH9 3JR, UK

Reprinted by permission from the
Rett Syndrome Research Foundation
4600 Devitt Drive
Cincinnati, Ohio 45246
Phone: 513-874-3020
Fax: 513-874-2520

Back To The Diagnostic Categories Download this Article