Rett syndrome (RTT) is a severe neurodevelopmental disorder affecting almost exclusively girls. It is currently considered a monogenic X-linked dominant disorder due to mutations in MECP2 gene, encoding the methyl-CpG binding protein 2. A few RTT male cases, resulting from mosaicism for MECP2 mutations, have been reported. Male germline MECP2 mutations cause either severe encephalopathy with death at birth (usually in brothers of classical RTT females) or X-linked recessive mental retardation (XLMR). To date the wide phenotypic heterogeneity associated with MECP2 mutations in females (from classical RTT to healthy carriers) has been explained by differences in X chromosome inactivation. However, conflicting results have been obtained in different studies, with both random and highly skewed X-inactivation reported in healthy carrier females. Consequently it is possible that mechanisms other than X-inactivation play a role in the expressivity of MECP2 mutations. To explain the phenotypic heterogeneity associated with MECP2 mutations we propose a digenic model in which the presence of a "mutated" allele in a second gene, leading to a less functional protein, determines the clinical severity of the MECP2 mutation. The model is supported by the identification of the same mutation in XLMR and RTT cases. The carrier mothers of XLMR families are clinically asymptomatic and present balanced X chromosome inactivation. Therefore the same mutation arising in different genetic backgrounds can cause XLMR in males, remain silent in the carrier females and cause classic RTT in females. MECP2 mutations account for approximately 70-80% of classic RTT cases. MECP2 negative cases might result from mutations in noncoding regions of MECP2 gene. Alternatively, these cases might be due to mutations in other genes (locus heterogeneity). This hypothesis is supported by the identification of several chromosomal rearrangements in MECP2 negative patients with RTT and RTT-like phenotypes. MeCP2 is considered a general transcriptional repressor. However, conditional mouse mutants with selective loss of Mecp2 in the brain develop clinical manifestations similar to RTT, indicating that MECP2 is exclusively required for central nervous system function. The involvement of MeCP2 in methylation-specific transcriptional repression suggests that MECP2 related disorders result from dysregulated gene expression. Studies on gene expression have been performed in mouse and human brains. A relatively small number of gene expression changes were identified. It is possible that MeCP2 causes dysregulation of a very small subset of genes that are not detected with this method of analysis, or that very subtle changes in many genes cause the neuronal phenotype.