Skip Navigation


Human Molecular Genetics Advance Access originally published online on May 11, 2005
Human Molecular Genetics 2005 14(13):1851-1861; doi:10.1093/hmg/ddi191
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Material
Right arrow All Versions of this Article:
14/13/1851    most recent
ddi191v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Watson, C. M.
Right arrow Articles by Tam, P. P.L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Watson, C. M.
Right arrow Articles by Tam, P. P.L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org

Reduced proportion of Purkinje cells expressing paternally derived mutant Mecp2308 allele in female mouse cerebellum is not due to a skewed primary pattern of X-chromosome inactivation

Catherine M. Watson1, Gregory J. Pelka1,2,3, Tatiana Radziewic1, Mona D. Shahbazian4,{dagger}, John Christodoulou2,3, Sarah L. Williamson2,3 and Patrick P.L. Tam1,*

1Embryology Unit, Children's Medical Research Institute, NSW, Australia, 2Discipline of Pediatrics and Child Health, University of Sydney, NSW, Australia, 3Western Sydney Genetics Program, Children's Hospital at Westmead, NSW, Australia and 4Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX 77030, USA

* To whom correspondence should be addressed at: Embryology Unit, Children's Medical Research Institute, Locked Bag 23, Wentworthville, NSW 2145, Australia. Email: ptam{at}cmri.usyd.edu.au

Received March 24, 2005; Revised April 29, 2005; Accepted May 6, 2005


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Rett syndrome (RTT) is an X-linked disorder caused by mutations in the methyl CpG binding protein 2 (MECP2) gene. The pattern of X-chromosome inactivation (XCI) is thought to play a role in phenotypic severity. In the present study, patterns of XCI were assessed by lacZ staining of embryos and adult brains of mice heterozygous for a X-linked Hmgcr–nls–lacZ transgene on a mutant mouse model of RTT. We found that there was no difference between the lacZ staining patterns in the brain of wild-type and heterozygous mutant embryos at embryonic day 9.5 (E9.5) suggesting that Mecp2 has no effect on the primary pattern of XCI. At 20 weeks of age, there was no significant difference between XCI patterns in the Purkinje cells in the cerebellum of heterozygous mutant and wild-type mice when the mutant allele was inherited from the mother. However, when the mutant allele was paternally inherited, a significant difference was detected. Thus, parental origin of the mutation may have a bearing on phenotype through XCI patterns. An estimation of the Purkinje cell precursor number based on XCI mosaicism revealed that, when the mutation was paternally inherited, the precursor number was less than that in the wild-type mice. Therefore, it is likely that the number of precursor cells allocated to the Purkinje cell lineage is affected by a paternally inherited mutation in Mecp2. We also observed that the pattern of XCI in cultured fibroblasts was significantly correlated with patterns in the Purkinje cells in mutant animals but not in wild-type mice.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Rett syndrome (RTT) is an X-linked dominant neurodevelopmental disorder that mainly affects girls. Mutations in the methyl-CpG binding protein 2 (MECP2) gene (1Go) account for 80% of cases of classic RTT and a smaller percentage of the variant forms of RTT (2Go). RTT is characterized by a period of apparently normal development until the age of 6–18 months after which patients regress and speech and purposeful hand movements are lost. The clinical features that characterize the disorder are variable and include poor coordination, stereotypic hand movements, seizures, scoliosis, decreased head and body size and breathing irregularities (3Go).

The Mecp2/MECP2 gene is subject to X-chromosome inactivation (XCI) in female mice (4Go) and humans (5Go). The choice of which X-chromosome to inactivate is random in the somatic cells of early embryos of normal mice and humans (reviewed in 6Go). In most cases, the ratio of cells with the maternally to paternally derived X-active chromosome may vary between 80:20 and 20:80 (7Go). Individuals are classified as showing skewed patterns of XCI when >75% of the transcript is encoded by one X-linked allele (8Go,9Go). Skewing could be a result of stochastic processes during the establishment of XCI or cell selection at subsequent development. The XCI pattern can also be affected by various X-linked (6Go) and autosomal loci (10Go) in the mouse. In the human, the regions Xq28 (11Go) and Xp11.2–11.4 (12Go) have been shown to affect the pattern of XCI. In addition, the degree of skewing of XCI in peripheral blood cells appears to be affected by age (8Go), but this is not universally observed (13Go).

XCI mosaicism has been proposed as one mechanism to explain the phenotypic variability associated with MECP2 mutations. This appears to be true for cases such as carrier females with mild or absent RTT phenotypes (including the preserved speech variant), in which skewed XCI patterns have been observed in blood (14Go). However, most RTT cases display primarily random XCI patterns in the blood and the brain (15Go) and show no consistent correlation with the severity of symptoms (16Go). It is clear, then, that XCI patterns are not sufficient to explain the phenotypic variation resulting from mutation of MECP2. When several studies are combined, the frequency of skewing in 1095 normal subjects ranges from 17 to 38% if skewing is defined as more than 75:25 or from 4 to 17% when a ratio of more than 90:10 is taken as the criterion for skewing (13Go). Skewed XCI at more than 80:20 was found at a higher incidence (~50%) in several X-linked mental retardation disorders (XLMR). In these XLMR cases, the mutation-bearing X chromosome was preferentially inactivated, suggesting that the cells with an active mutation-bearing X chromosome may have been selected against (17Go).

Human males with mutations in the MECP2 gene are rare. The rarity of male cases has been assumed to be due to an X-linked dominant mode of inheritance with lethality in male hemizygotes (OMIM no. 312750 [OMIM] ). Several cases were reported in which affected males display variable phenotypes that may include some or all of the following features: severe neonatal encephalopathy, mental retardation, spasticity, tremor of the hands, distal atrophy of the legs and loss of speech (18Go). The greater incidence of females when compared with males with MECP2 mutations can be explained by a high number of sporadic mutations arising in the male germline resulting in only female offspring inheriting the affected X chromosome from their fathers (19Go). Two groups found that 96% (20Go) and 71% (21Go) of mutations in sporadic cases of RTT were of paternal origin. Affected males are usually found in familial cases where the mother has a skewed pattern of XCI and is only mildly affected (22Go) or has a random pattern of XCI in which case maternal germline mosaicism for the mutation appears to explain the absence of a somatic phenotype in the mother (23Go). The relationship between parental origin of the mutation and phenotypic severity has not been examined.

XCI patterns have been studied in two Mecp2-null mice models (24Go,25Go) of RTT. Braunschweig et al. (26Go) studied XCI patterns in heterozygous Mecp2 mutant mice brains using laser scanning cytometry to detect the proportion of cells expressing the MeCP2 protein. Their data showed a significant reduction in the proportion of Mecp2– cells in six regions of the brains of 10 mice indicating that the mutant allele is inactivated in most cells. In addition, it was found that the level of MeCP2 expression is reduced in the vicinity of the MeCP2– cells suggesting that Mecp2– cells influence protein expression levels in neighbouring Mecp2+ cells. These findings raise the possibility that XCI patterns influence RTT phenotype by cell autonomous as well as by non-cell autonomous (26Go).

XCI was also studied in a mouse model of RTT that carries a truncation mutation with a stop codon positioned after codon 308. The mutated allele is referred to as Mecp2308 (2Go). The phenotype of these mice is milder than that of Mecp2 null mice (24Go) and the males survive and are fertile (2Go). Similar to the Mecp2 heterozygous null mice, expression of the wild-type allele was favoured in 1-year-old Mecp2308 heterozygous females and >60% of these mice had >75% of their Purkinje cells expressing wild-type Mecp2 (27Go). They also observed that mice with a skewed pattern of XCI displayed less severe mutant phenotypes.

Although the impact of XCI on phenotype has been studied at the cellular and whole animal level, it is not clear at which developmental stages the XCI patterns are established and whether the parental origin of the mutation has any effect. We have investigated XCI patterns in female mice by exploiting the milder impact of the truncation mutation on male viability and fertility in Mecp2308 mice, which has allowed the generation of heterozygous mutant female mice carrying mutant allele inherited from either parent. This has not been previously feasible with the other two null-mutant mice models where male mice die too early to transmit the mutation efficiently to the next generation and limited the analysis of XCI to female heterozygous mice carrying only the maternally inherited mutation. In the present study, we have assessed the effect of parental origin of the mutant Mecp2 allele at two developmental time points: embryonic day 9.5 (E9.5) when the pattern of XCI is fully established; and postnatal week 20, when the Purkinje cell population has reached its adult size and is at a steady state of differentiation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
XCI pattern in embryonic brain is unaffected by Mecp2 mutation
To determine whether the initial pattern of XCI is affected by the presence of a mutated Mecp2308 allele during embryogenesis, we chose to study the pattern at E9.5 (Fig. 1A–G) as this is the earliest time at which it can be determined in morphologically distinct regions of the embryonic brain (28Go). In the E9.5 brain, Mecp2 is expressed widely (Supplementary Material, Fig. S1A–D) (29Go). Therefore, any impact of altered MeCP2 activity on XCI can be assessed in all types of the neural progenitors.



View larger version (105K):
[in this window]
[in a new window]
  		Figure 1. XCI patterns in the brains of E9.5 embryos and in the Purkinje cells of 20-week-old mice (A–E) Overall staining pattern for lacZ-expressing (Mecp2+ D4X derived allele, blue) and nuclear fast red-stained [Mecp2308 or Mecp2 129+ (129/SvEv) allele, red] cells in the E9.5 embryo. (A) Mecp2+/y male embryo expressing the X-linked lacZ transgene in all cells, (B) Mecp2+/129+ female embryo expressing maternally derived X-linked transgene, (C) Mecp2+/308 female embryo with paternal Mecp2308 allele and expressing maternally derived X-linked transgene, (D) Mecp2129+/+ female embryo expressing paternally derived X-linked lacZ transgene and (E) Mecp2308/+ female embryo with maternal Mecp2308 allele and expressing paternally derived X-linked lacZ transgene. (A, D and E) Offspring of crosses of Mecp2308/129+ female and D4X male mice; (B and C) D4X homozygous female and Mecp2308/y male mice. (F) High magnification view of the brain of (A) Mecp2+/y embryo and (G) of (E) Mecp2308/+ embryo showing the regions (H–M) where cells were scored. (I) The marking of cells in the selected regions for differential scoring of X-gal-stained (yellow dot) and nuclear fast red-stained (red dot) cells. (N) In the male Mecp2–/y cerebellum, the Purkinje cells and cells in the adjacent molecular layer of a cerebellar folium show uniform X-gal staining. (O) Mecp2+/129+ is a wild-type female with a paternally derived wild-type Mecp2 129/SvEv allele showing mosaic lacZ expression pattern in the Purkinje cells and other cerebellar cells in the molecular and granule layers. (P and Q) Female Mecp2+/308 heterozygote (the offspring of a D4X homozygous femalexMecp2308 male) with paternally derived mutant Mecp2308 allele: upper arrows mark red-stained Purkinje cells bearing a mutant Mecp2308 allele and lower arrows mark lacZ-stained wild-type Purkinje cells. Te, telencephalon; Di, diencephalon; Tc, dorsal midbrain, tectum; Tg, ventral midbrain, tegmentum; My, mylencephalon, hindbrain and Ma, mandibular arch.

 
Previously, we have shown that the expression of the Hmgcr–nls–lacZ transgene is a reliable indicator of XCI in the entire neuroepithelial cell population (28Go). Our assay system readily detects those cells in which the transgene is on the active X chromosome, because of strong expression of the lacZ transgene which is visible as blue staining in the nuclei (Fig. 1H–M). When the active X chromosome is the one that carries the lacZ transgene, it is therefore expected that the cis-wild-type (+) Mecp2 allele of D4X origin is also expressed. Conversely, nuclei in which the lacZ-bearing chromosome is inactivated will either express a wild-type (129+) Mecp2 allele or the mutant Mecp2308 (308) allele of 129/SvEv origin depending on the genotype of the animal. These cells are red-stained (Fig. 1H–M) and will be referred to as ß-gal– cells. The parental origin will be indicated by the maternal/paternal convention. For example, Mecp2129+/+ indicates that the maternally derived chromosome bears the wild-type Mecp2 allele of 129/SvEv origin (Fig. 1A–E). In our study, F1 animals were used and therefore, there is no chance for recombination between the lacZ transgene and the Mecp2 allele. Thus, although we do not test Mecp2 expression directly, our assay can clearly determine which X chromosome is active in a given cell.

A subset of cells was selected for the analysis from five brain regions: telencephalon, diencephalon, dorsal midbrain, ventral midbrain and basal part of the upper rhombencephalon (Fig. 1F and G). In the male lacZ-transgenic embryo, ~100% of all cells in the brain stained positively for X-gal (Fig. 1A, F and H), indicating that cells with an active X-chromosome carrying the transgene can be consistently detected by X-gal staining in this assay.

The number of X-gal and nuclear fast red-stained nuclei was scored for each region from several sections per region for each embryo. Twenty embryos of each genotype were scored. To ensure reproducibility of cell scores, a digital image of the selected section of the embryonic brain regions was taken and, using Photoshop 7 software, cells were marked for positive and negative staining on digital pictures of the sections (Fig. 1I) and then the relative number of blue- and red-stained cells was scored manually.

Standard statistical measures of the spread of data were determined and the means of the cell scores are presented in Table 1. The non-parametric Wilcoxon–Mann–Whitney rank sum test was performed to compare the scores of similar brain regions and the whole brains of wild-type and heterozygous Mecp2308 mutant specimens (Table 1). No significant differences were detected between the wild-type and the heterozygous Mecp2308 mutant embryos for any brain region and the whole brain irrespective of whether the mutant allele was inherited maternally or paternally. There is therefore no difference in the primary XCI pattern in the E9.5 embryonic brain of embryos of different genotypes of the same experimental cross.


View this table:
[in this window]
[in a new window]
  		Table 1. The pattern of XCI in the E9.5 embryonic brains of mice heterozygous for the Mecp2308 mutation
 
Curiously, the dorsal midbrain region seems to consistently have a higher proportion of ß-gal– cells relative to other regions. However, there is no significant difference between the heterozygous mutant Mecp2308 and the wild-type embryos in the dorsal midbrain region. The uniform positive X-gal staining of cells in the male transgenic embryo showed that the higher proportion of ß-gal– staining cells in females is not due to reduced lacZ expression or an artefact of staining in this brain region.

Preferential activity of the wild-type maternal allele in Purkinje cells of adult Mecp2308 mice
To ascertain whether the pattern of XCI had changed as the mice aged, we assessed the pattern of XCI in Purkinje cells of 20-week-old mice. Purkinje cells were selected for the analysis because they express the lacZ transgene very strongly relative to other cell types in the adjacent molecular layer in the folia of the cerebellum (Fig. 1N–Q) and they also express Mecp2 mRNA (Supplementary Material, Fig. S1E and F) (29Go) and MeCP2 protein (30Go), which will render them potentially susceptible to the effect of altered MeCP2 function on cell function and viability. In addition, the Purkinje cells are easily identified by their size, distinctive morphology and a readily defined localization at the junction of the molecular and internal granule layers (31Go). For each of 19–27 brain samples of the four types of mice (Table 2), at least 200 lacZ expressing (X-gal-stained) and red counterstained Purkinje cell nuclei (Fig. 1O–Q) were counted in three to five non-adjacent sections depending on the number of folia that could be analyzed in each section.


View this table:
[in this window]
[in a new window]
  		Table 2. The pattern of XCI in the Purkinje cells of the cerebella and cultured fibroblasts of 20-week-old mice heterozygous for the Mecp2308 mutation
 
In the brains of the two types of wild-type female mice (Table 2), the overall mean proportion of red counterstained Purkinje cell counts was <50% (38–41%), indicating that there may be a higher abundance of lacZ expressing (blue, ß-gal+) cells in this lineage. However, this percentage was still well within the range of values which is consistent with random XCI (7Go). This bias for more ß-gal+ Purkinje cells was found for both Mecp2+/129+ and Mecp2129+/+ female and heterozygous Mecp2+/308 and Mecp2308/+ specimens. A possible factor influencing the XCI ratio may be the strength of the activity of the XCI controlling element (Xce). The D4X transgenic mice were generated by crossing H253 [(C57BL6xCBA/2) F1] mice with the D4 [(129 R1xICR) F1] mice. These mice have been inter-crossed for more than 10 generations and will therefore be of a mixed background of the four founding strains. The Xce genotype of the D4X mice is likely to be a/b, a/a or b/b, in contrast to the Xce a/a genotype of the 129/SvEv mice that harbour the Mecp2 mutant allele (32Go,33Go). Consequently, mice of the experimental crosses will therefore be either Xce a/b or Xce a/a. Given that the two Xce alleles are associated with low to medium level of expression of cis-locating genes on the X chromosome, and that the data were collected from at least 19 different adult brains, the Xce genotype may have insignificant impact on the XCI outcome between female mutant and wild-type mice of the same experimental cross. This hypothesis is supported by the finding that the percentages of ß-gal– cells are not statistically different between maternal 129-wild-type and paternal 129-wild-type brains (Wilcoxon–Mann–Whitney rank sum test, P>0.05). However, it is not known if the XCI ratio may be influenced by other autosomal or X-linked loci (10Go,34Go). All Purkinje cells in male controls stained blue (Fig. 1N) suggesting that every ß-gal– cell can be effectively detected by our histochemical procedure, and cells that were not stained by X-gal (ß-gal–) represent the population not expressing the X-linked lacZ transgene and are not due to a staining artefact.

The Wilcoxon–Mann–Whitney rank sum test was performed to compare the scores of the percentage of ß-gal– Purkinje cells (i.e. cells that have an active X-chromosome with the 129/SvEv derived wild-type or Mecp2308 allele) in the wild-type and heterozygous mutant Mecp2308 cerebellum (Table 2). No significant difference in the abundance of ß-gal– Purkinje cells was detected when the Mecp2308 allele was maternally inherited relative to that of the wild-type 129/SvEv allele. In contrast, there was a significant reduction in the proportion of Mecp2308 expressing Purkinje cells in the brains of mice that inherited a paternally derived mutant Mecp2308 allele versus the wild-type 129/SvEv allele (Table 2). A comparison of the score of percentage of ß-gal– Purkinje cells also revealed a significant difference between the maternal and the paternal heterozygous mutant mice. This finding, when considered together with the lack of the effect of parental origin of the X-chromosome of different genetic background in the wild-type D4X/129 mice, strongly suggests that the reduced proportion of Purkinje cells expressing the paternally inherited Mecp2 mutant allele is principally due to the parent-specific effect of the Mecp2 mutation and not to other genetic loci that may influence XCI (10Go,32Go–34Go).

The paternally derived Mecp2308 Purkinje cell progenitor pool may be reduced
The significant difference found between wild-type and mutant Mecp2+/308 heterozygotes prompted us to ask whether this difference arose at the time of allocation of progenitors of the Purkinje cell lineage between E11 and E13 or whether it happened over time after the Purkinje cells had reached their final location.

To estimate the number of progenitors of the Purkinje cells, we followed the protocol outlined by McMahon et al. (35Go), which is based on the model that the random selection of cells from a common pool of progenitors of different active X-chromosome gives rise to the variability in the ratio observed in the relative abundance of cell types with different X-linked transgene activity in different animals (36Go). The computation of the number of progenitors that is present at the time of lineage allocation is based on the degree of variance of score of the cell types between samples in each animal and that of the mean score between animals. Analysis of these two variances for the week 20 Purkinje cell population revealed that 39 brain samples (n=17) and 36 Purkinje cell progenitors (n=15) per half cerebellum may be present in mice with maternally inherited wild-type and Mecp2308 alleles, respectively. The value calculated for the Purkinje cell progenitors of mice with paternally derived wild-type Mecp2 129/SvEv allele was 76 cells (n=15) which is higher than mice with either maternally inherited wild-type or mutant Mecp2 gene. Of significance is that a value of 66 cells (n=19) was obtained for the paternally derived Mecp2308 allele, which was 13% less than that of the wild-type control. Our findings therefore suggest that the expression of a paternally inherited mutant Mecp2 allele may lead to a reduction in the number of Purkinje cell progenitors present at lineage allocation.

CFB XCI pattern as a predictor of the Purkinje cell XCI pattern
To determine whether the difference between paternally derived heterozygous Mecp2308 mutant and wild-type XCI patterns was confined to Purkinje cells, we examined XCI in cultured fibroblasts (CFB) from tail tip tissues. CFB have been previously shown to express Mecp2 (37Go). Cells were not passaged in order to reduce any bias on cell populations due to stochastic influences or varied cell viability that could be manifested with prolonged culture. At least 100 cells per culture were scored for positive or negative expression of lacZ (Supplementary Material, Fig. S2). The Wilcoxon–Mann–Whitney rank sum test was performed and no significant differences were found between the wild-type and the heterozygous Mecp2308 mutant cells with either a paternally (P=0.07) or a maternally (P=0.54) inherited mutant Mecp2308 allele (Table 2).

Linear regression analysis was performed to determine whether a correlation exists between the XCI patterns in CFB and the Purkinje cells in the cerebellum. R-values derived from the linear curve fitted to the scatter plots are not statistically significant for either the maternally or the paternally derived wild-type 129/SvEv Mecp2 allele (Table 2), suggesting that there is not a consistent correlation between the XCI pattern in CFB and the Purkinje cells in the same animal that is wild-type for the Mecp2 gene. However, there is a statistically significant correlation between the percentage of ß-gal– cells in mutant Mecp2308 CFB and Purkinje cells irrespective of the parental origin of the mutation. If we combine the wild-type data of both parental origins (n=25), a R-value of 0.07 is obtained, which is not significant as the P-value for the F statistic is 0.74 (Fig. 2A). However, when the mutant Mecp2308 heterozygote data are combined (n=24), a highly significant correlation (R=0.75, P=0.00002 for F statistic) is found (Fig. 2B). The mean percentage of ß-gal– cells is always 12–15% higher in CFB when compared with Purkinje cells. This may be due to differences in detecting blue cells versus red cells in the two cell types, the CFB are also proliferative cells and this affects the intensity of blue staining in CFB, whereas the Purkinje cells at 20 weeks are not proliferating. In male controls, ß-gal– cells ranged from 2% to as high as 37% for one animal with an average of 16% (n=8), which is the background rate for mis-scoring ß-gal+ cells as ß-gal– cells. Despite this caveat, a CFB pattern skewed in favour of the wild-type allele may allow a prediction about the direction of skewing of XCI in the brain. Therefore, the CFB XCI pattern may give a reasonable estimate of the Purkinje cell pattern in mutant Mecp2308 heterozygotes.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 2. Linear regression analysis of percentage of ß-gal– cells in CFB versus Purkinje cell samples for (A) wild-type mice and (B) Mecp2308 heterozygotes with 129SvEv-derived wild-type and mutant allele of either parental origin. Confidence limits for combined wild-type values: lower 95% Y-intercept 20.1 and upper 95% Y-intercept 51.6. Confidence limits for combined heterozygous mutant Mecp2308 values: lower 95% Y-intercept-4.1 and upper 95% Y-intercept 16.7.

 
The number of animals displaying a skewed XCI ratio of ≥80:20 was determined (Table 2). Only one animal was skewed in the wild-type CFB and Purkinje cells, whereas in the group that inherited the mutant Mecp2308 allele from their father seven were skewed. In general, the skew was in favour of the wild-type allele being on the active X chromosome, however, three heterozygous mutant Mecp2308 animals were skewed in favour of an active mutant allele in CFB.

A potential bias for expressing the wild-type allele in RTT patients with paternally transmitted mutation
To explore whether there may be XCI leading to preferential inactivation of the paternal X-chromosome when the mutation arises from the paternal origin, we studied 20 RTT patients known to have a de novo c.473C>T (T158M) missense mutation. We found two of them to be heterozygous for a IVS3-109A>G single nucleotide polymorphism (SNP), which is in close proximity to the T158M mutation, allowing us to determine the parental origin of the X chromosome harbouring the mutation and performed XCI studies in duplicate on DNA from the affected individuals and their parents (38Go). We found that in both cases the T158M mutation had arisen on paternally derived X chromosome, and for both patients, there was a tendency for the X-chromosome harbouring the mutation to be inactivated, at least in DNA extracted from peripheral blood samples (mean results for the active allele to inactive allele being 68:32% and 76.5: 23.5%). Although the number of patients studied is small, these preliminary data are consistent with our findings in the mouse. A more systematic examination of a larger cohort of RTT patients with this and other pathogenic mutations would be of value and may shed further light on some of the inconsistencies that have been encountered to date in attempts at phenotype–genotype correlations in RTT.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
XCI bias against Mecp2308 mutant alleles
Mice studies have indicated that XCI patterns in the brain favour cells in which the X-chromosome bearing the wild-type allele is the active X (27Go). It appears that it is not just the proportions of cells expressing either wild-type or mutant alleles that may be important in determining phenotype. Cells which express the mutant allele (and lack MeCP2 protein) may have a repressive effect on the expression of wild-type MeCP2 in surrounding cells in a localized manner (26Go) thus potentially amplifying the effect of cells in which the mutant allele is active. It has been shown that the higher the proportion of cells expressing the wild-type Mecp2 allele the less severe is the mutant phenotype (27Go).

Our results indicate that the number of skewed individuals is higher in the mutant Mecp2308 animals when the mutant allele is inherited from the father. This is consistent with the greater degree of skewing in the Mecp2+/308 animals except that XCI in one animal is skewed in favour of the mutant allele being active in CFB. It is also consistent with the XCI pattern where an active wild-type allele was found in a greater proportion of Purkinje cells in 1-year-old Mecp2308 heterozygous female mice (27Go). Braunschweig et al. (26Go) examined the overall XCI pattern and found that there was a bias in favour of an active wild-type allele in a small number of mice (about five each) of two Mecp2 null lines. In these cases, the mutation would have been maternally derived as males of these lines do not reproduce efficiently. In our study, there may have been a bias in favour of scoring ß-gal+ cells as they are easier to detect than ß-gal– Purkinje cells. The bias, however, would have been present in both wild-type controls and mutant samples. The use of an antibody to MeCP2 in two studies (26Go,27Go) has the advantage of directly assaying expression of the gene at the protein level but has the disadvantage that the XCI pattern cannot be determined in wild-type controls. Our study has enabled us to directly compare wild-type and heterozygous mutant animals. Consistent with our observation of a bias favouring the expression of the wild-type Mecp2 allele when the mutant allele is paternally inherited, further analysis of the data obtained from the 1-year-old Mecp2308 mutant mice revealed that ~70% of these mice, which show a bias for wild-type Mecp2-expressing Purkinje cells, had their mutant allele paternally inherited (J.I. Young and H.Y. Zoghbi, personal communication). Therefore, it appears that there is a trend against the X chromosome bearing the mutant Mecp2308 allele being active especially if the mutation is paternally inherited. The basis of this parent-specific effect of Mecp2 mutation is not known, but it may be potentially related to the downstream action of Mecp2 on imprinted genetic activity (39Go,40Go). Although parental origin of the mutated MECP2 allele has been studied in humans, its impact on XCI or phenotype has not been fully determined. Most MECP2 mutations were, however, of paternal origin (20Go). In our limited analysis of 20 RTT patients with T158M mutation, we have found a tendency of skewed XCI favouring wild-type allele expression in two cases where the mutation was of paternal origin.

Developmental timing of the impact of Mecp2308 mutation
When does the trend to favour cells in which the wild-type allele is on the active X chromosome begin? We examined this question by determining the XCI pattern at E9.5. If the Mecp2 gene had any direct effect on the selection of which X chromosome to inactivate at the time of initiation of XCI, then it would result in a bias that would be evident at E9.5. We found no difference between the XCI patterns in the brains of wild-type and heterozygous mutant Mecp2308 embryos, and therefore, mutant cells must be at a selective disadvantage at a later stage of embryonic or adult life.

There is a bias towards inactivation of the mutant allele of 10% in the Purkinje cells at 20 weeks of age (this study) and at 1 year of age (27Go). Is this a result of cell death as brains age or does it happen during differentiation and migration of the Purkinje cell lineage? We estimated the number of cells at the time of allocation of the Purkinje cell lineage. A difference was detected between the paternally derived mutant Mecp2308 heterozygous animals and their matched controls with mutants allocating 13% less cells to the lineage by computation based on the XCI-inactivation mosaicism model. This is consistent with an average difference of 10% between wild-type and mutant Mecp2308 XCI patterns at 20 weeks of age. It is also possible that there are a reduced population of Purkinje cells in mutant Mecp2308 heterozygotes in addition to the difference in XCI patterns; however, we are not able to determine the total number of Purkinje cells in the adult cerebellum in the current study. Our findings are consistent with parental origin of the Mecp2308 mutant allele having different effects on the number of precursor cells allocated to the Purkinje cell lineage.

It would be of interest to explore these estimates further by studying cell proliferation or apoptosis. However, this is not feasible for a number of reasons. The precursor population arises over 3 days of gestation with Purkinje cell precursors allocated between E11 and E13 (41Go). Thus, in order to study cell proliferation and apoptosis, three cellular markers would need to be optimized: lacZ transgene or Mecp2308 truncated protein antibodies; a precursor population marker and a TUNEL assay. Purkinje cell precursors at the time cannot be distinguished morphologically from other cerebellar neuron precursors such as Golgi cells (42Go). A large number of embryos would need to be studied for each cross because of the large stochastic variation in XCI patterns and to have the statistical power to detect a relatively small population difference (10–13%) between the paternally derived mutant Mecp2308 heterozygotes and the wild-type control XCI patterns, potentially over a 3 day period, which makes detection of small differences unlikely. Only when the progenitors of multiple cerebellar cell lineages were substantially reduced in number with a markedly different XCI pattern in Mecp2308 heterozygous and wild-type control mice, would cell death and proliferation studies be useful.

Purkinje cells are lost in wild-type mice from ~18 months of age onwards. In heterozygous staggerer mice, the process of Purkinje cell loss begins about a year earlier than in wild-type mice (43Go). It is possible that some loss could be occurring over time in heterozygous Mecp2308 mice due to differences in cell viability between cells expressing the mutant Mecp2308 allele and those expressing the wild-type allele. In our study, the spread of percentage of ß-gal– cells in Mecp2308/– was 19–61% and for Mecp2+/308, it was 1–49% and the number of individuals with skewed XCI patterns (≥20:80 mutant cells/wild-type cells) was 5 and 26%, respectively (Table 2). A spread of 18–46% of cells with the mutant Mecp2308 active and with 31% skewed at (≥20:80 mutant/wild-type cells) 1 year of age (27Go). In our study, we compared the spread of heterozygous mutant Mecp2308 percentage of ß-gal– values with those of wild-type, i.e. Mecp2129+/+ 21–64%; Mecp2+/129+ 18–53%, which revealed a skew in values presumably due to ease of scoring blue ß-gal+ cells. No such comparison was possible in study by Young and Zoghbi (27Go) but it would appear from the data available that there was no further change in the XCI pattern between 5 months of age and 1 year. The two studies do agree that there is a real difference between heterozygous mutant Mecp2308 XCI patterns and those of wild-type. Therefore, it is most likely that the impact of Mecp2 deficiency occurs during lineage allocation.

Human mutant lymphocyte cells do not proliferate as well as wild-type cells in culture suggesting that mutant Mecp2 cells are at a disadvantage (44Go). It is possible that cells in which the Mecp2308 allele is active may be selected against during the earliest stages of Purkinje cell lineage allocation and differentiation giving rise to a tendency for wild-type cells to predominate in the mature cerebellum. Therefore, it is most likely that the difference that exists in the XCI patterns of wild-type and Mecp2+/308 mice occurs during Purkinje cell lineage allocation.

In summary, our study shows that Mecp2 does not have any effect on the primary pattern of XCI established in early embryogenesis. A significant difference in XCI patterns can be detected at postnatal week 20 in the Purkinje cells when the mutant Mecp2308 allele is paternally inherited. This raises interesting questions about the role of parental origin of the mutation in the pattern of XCI and therefore in the phenotype. The estimate of a reduced number of Purkinje cell progenitors in Mecp2+/308 mice suggest that this difference has its origin at the time of allocation to the Purkinje cell lineage.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
SUPPLEMENTARY MATERIAL
 REFERENCES
 
Mice and genetic crosses
Two strains of mice were used in this study: Mecp2308 mice on a pure 129/SvEv background (2Go) and D4X mice maintained as a closed-bred stock on a mixed background. D4X mice are derived from crosses of H253 [(C57BL6xDBA/2)F1] mice, which express an X-linked Hmgcr–nls–lacZ transgene [TgN (HMG–lacZ) 253Tan] (45Go) with D4 [(129R1xICR)F1] mice with an X-linked pBA–CMV–EGFP transgene [TgN(GFPX)4 Nagy (pBA–CMV–EGFP)] (46Go). Crosses of Mecp2308 and D4X mice will therefore always generate heterozygous mice in which the X-chromosome carrying the Mecp2308 allele is of 129/SvEv origin and the X-chromosome bearing the lacZ transgene and the EGFP transgene are of mixed genetic background derived from three original stocks. Reciprocal crosses of Mecp2308 and D4X mice were performed to produce heterozygous females that inherit the Mecp2 mutation from either parent. First, Mecp2308 heterozygous females were crossed to male D4X mice to produce female offspring carrying the Mecp2 mutation on the maternally derived X chromosome and the lacZ and EGFP transgenes on the paternally derived X chromosome. These mice were designated as Mecp2308/+. The wild-type female littermates were used as controls. Secondly, D4X homozygous females were crossed to Mecp2308 males to produce females with a paternally derived Mecp2308 mutation and the lacZ and EGFP transgenes on the maternally derived X chromosome. These mice were designated as Mecp2+/308. Crosses of homozygous D4X females and 129/SvEv males were performed to produce wild-type female and male controls. The expression of the X-linked EGFP was used for the genotyping of the heterozygous mice in the breeding stock and for detecting the heterozygous E9.5 embryos obtained from the experimental crosses. The X-linked lacZ transgene was used as the reporter of X-chromosome activity in histological analysis.

Sample collection and genotyping
Offspring from the aforementioned crosses were collected for the analysis of X-linked transgene expression at two time points: E9.5 embryos and adults at postnatal 20 weeks. Yolk sacs of embryos and tail tips collected from 3-week-old mice were used for genotyping by PCR. The PCR primers were 5'-AACGGGGTAGAAAGCCTG-3', 5'-TGATGGGGTCCTCAGAGC-3' and 5'-ATGCTCCAGACTGCCTTG-3' and cycling conditions were 94°C for 3 min, 35 cycles of 94°C for 30 s, 60°C for 30 s and 72°C for 30 s ending with 72°C for 3 min. A 396 bp band is produced from the wild-type allele and a 318 bp band from the mutant Mecp2308 allele. The final reaction buffer consisted of 15 mM Tris–HCl pH 8.0, 50 mM KCl, 200 µM of each dNTP, 1.5 mM MgCl2, 0.2 µM of each primer, 2.5 U Taq polymerase (Roche) and 1 µl of yolk sac DNA in a 25 µl reaction.

Cultured fibroblasts
Postmortem tail biopsies from 20-week-old mice were minced and cultured on gelatinized plates in medium containing 10% foetal calf serum, 1x penicillin/streptomycin in Dulbecco's minimal essential medium until outgrowths of fibroblasts appeared in the culture. The cells were washed in phosphate-buffered saline (PBS) and lightly fixed in 4% paraformaldehyde (PFA) for 2.5 min. Cells were washed twice for 1–2 min in PBS and stained overnight at 37°C in X-Gal solution [5 mM ferrocyanide, 5 mM ferricyanide, 2 mM MgCl2, 0.1% Tween-20 (Sigma), 1 mg/ml X-gal in PBS without Ca2+ and Mg2+]. Cells were counterstained with nuclear fast red (Certistain, Merck). The stained cell cultures were then photographed and the number of blue- and red-stained cells was scored on a magnified digital image (Supplementary Material, Fig. S2). An average of 679 cells (range 100–2529) was counted for each tail culture.

Histology
E9.5
Embryos from the E9.5 time point were collected and placed in PBS. DNA was extracted from yolk sacs (47Go) and embryos were sexed using Zfy primers (48Go) and genotyped by PCR as described earlier. Embryos were fixed in 4% PFA for 7 min and stained in 1 mg/ml X-gal solution (49Go) overnight at 37°C in the dark. Embryos were post-fixed in 4% PFA, embedded in paraffin wax, sagittally sectioned and counterstained with nuclear fast red. Photographs of histological sections were taken using an Olympus BX50 microscope and SPOT digital camera (Diagnostic Instruments) controlled with SPOT version 3.5 software. Contrast and brightness were adjusted using Photoshop 7 (Adobe). Cells are scored on high magnification digital images of the histological sections that were cut serially at 8 µm, which is the average nuclear diameter. This, together with the nuclear-localized X-gal and fast red staining would ensure that every nucleus in the section can be visualized with minimal masking or overlapping. The scoring was not affected by the cell or nuclear density. Five regions of the embryonic brain were analysed: telencephalon, diencephalon, dorsal midbrain, ventral midbrain and basal part of the upper rhombencephalon. An average of 345 cells per brain region of each embryo and 20 embryos of each genotype were analysed (Table 1).

Postnatal 20 weeks
The cerebellum was dissected from 20-week-old mice, washed briefly in PBS then fixed for 90 min in 4% PFA, 0.2% glutaraldehyde at 4°C. After fixation, the cerebellum was washed three times in PBS and soaked in 30% sucrose overnight. Specimens were embedded by immersion in OTC compound (Tissue-Tek, Pro Sci Tech) in cryomolds (Tissue-Tek) then frozen in isopentane over liquid nitrogen and stored at –80°C. Frozen sections were cut at 10–12 µm at –20°C, placed on StarFrost glass slides (ProSciTech) and stained overnight in 1 mg/ml X-gal solution (49Go) at 37°C in the dark. After counterstaining with nuclear fast red sections were mounted in Canada Balsam (in Xylene, Merck) and coverslipped. At least 200 Purkinje cells were counted in randomly selected sections of cerebellar folia from each sample (Table 2).

Statistical analysis
The data were first analysed for normal distribution using normal probability plots (KaleidaGraph version 3.6 2003 Synergy Software). In view of the non-normal distribution of the data, a non-parametric Wilcoxon–Mann–Whitney rank sum test was performed to determine whether a significant difference existed between the scores of wild-type and heterozygous mutant Mecp2308 female specimens. A linear regression analysis was performed on paired CFB and purkinje cell percentage of ß-gal– (lacZ negative) cells using Microsoft Office Excel 2003 11.6113.5703. An estimation of the number of precursors of the Purkinje cells was made based on a model of XCI mosaicism outlined by McMahon et al. (35Go) for each of the four groups of mice, i.e. maternal and paternal wild-type 129/SvEv alleles; maternal and paternal mutant Mecp2308 alleles. Initially, a two-way analysis of variance (ANOVA) was performed (Microsoft Office Excel 2003) on data obtained by scoring the relative abundance of lacZ-expressing and non-expressing Purkinje cells in three randomly selected and non-adjacent sections of the week 20 brain samples (Table 2). So that the variance of the mean values between animals and between samples could be determined. The results of the two way ANOVA were then used to calculate the theoretical number of Purkinje cell progenitors present at the time of lineage specification (35Go).

Examination of XCI patterns in RTT patient samples
The parental origin of the c.473C>T (T158M) mutation was determined by PCR amplification and sequencing of the genomic DNA surrounding the T158M mutation, using the oligonucleotides 5'-TGGAGAGACTGAGCACCGTA-3' and 5'-CTTCCCAGGACTTTTCTCCA-3'. This PCR fragment was cloned into pGEMTeasy (Promega) for those individuals with both a SNP and the T158M mutation, and sequencing of individual clones was performed to determine whether the two variations were in cis or trans. Sequencing of parental DNA was used to determine the origin of the SNP and therefore the origin of the T158M mutation.

The pattern of XCI in RTT patients was determined by examining the androgen receptor locus using previously described methods (38Go). Genotyping the parents at the androgen receptor locus was performed to ascertain which of the parental X chromosomes was potentially skewed and/or inactivated.


   SUPPLEMENTARY MATERIAL
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
REFERENCES
 
Supplementary Material is available at HMG Online.


   ACKNOWLEDGEMENTS
 
We thank Huda Zoghbi for sharing the Mecp2308 mutant mice and unpublished data, and her input on our work and this paper, Mehtap Baserdem, Irma Villaflor and Andrew Titmus for care of mice, Andrew Hayen for statistical advice, Helen Leonard and Mark Davis for the provision of patient DNA samples and Peter Rowe and David Loebel for comments on the manuscript. The animal experiments were approved by the CMRI-CHW Animal Ethics Committee (ACEC project 156). Our work was supported by the National Health and Medical Research Council (NHMRC) of Australia and Mr James Fairfax. G.P. is a NHMRC Dora Lush scholar and a Western Sydney Genetic Program RTT Scholar. S.W. is a University of Sydney UPA Scholar and a Western Sydney Genetics Program RTT Scholar. P.P.L.T. is a NHMRC Senior Principal Research Fellow.

Conflict of Interest statement. None declared.


   FOOTNOTES
 
{dagger} Present address: Department of Biological Chemistry, University of California, Los Angeles, CA 90095, USA. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 

  1. Amir, R.E., Van den Veyver, T.B., Wan, M., Tran, C.Q., Francke, U. and Zoghbi, H.Y. (1999) Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat. Genet., 23, 185–188.[CrossRef][ISI][Medline]

  2. Shahbazian, M., Young, J., Yuva-Paylor, L., Spencer, C., Antalffy, B., Noebels, J., Armstrong, D., Paylor, R. and Zoghbi, H. (2002) Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron, 35, 243–254.[CrossRef][ISI][Medline]

  3. Ellaway, C. and Christodoulou, J. (2001) Rett syndrome: clinical characteristics and recent genetic advances. Disabil. Rehabil., 23, 98–106.[CrossRef][ISI][Medline]

  4. Adler, D.A., Quaderi, N.A., Brown, S.D., Chapman, V.M., Moore, J., Tate, P. and Disteche, C.M. (1995) The X-linked methylated DNA binding protein, Mecp2, is subject to X inactivation in the mouse. Mamm. Genome, 6, 491–492.[CrossRef][ISI][Medline]

  5. D'Esposito, M., Quaderi, N.A., Ciccodicola, A., Bruni, P., Esposito, T., D'Urso, M. and Brown, S.D. (1996) Isolation, physical mapping, and northern analysis of the X-linked human gene encoding methyl CpG-binding protein, MECP2. Mamm. Genome, 7, 533–535.[CrossRef][ISI][Medline]

  6. Clerc, P. and Avner, P. (2003) Multiple elements within the Xic regulate random X inactivation in mice. Semin. Cell Dev. Biol., 14, 85–92.[CrossRef][ISI][Medline]

  7. Takagi, N. (2001) The role of X-chromosome inactivation in the manifestation of Rett syndrome. Brain Dev., 23 (Suppl. 1), S182–S185.

  8. Busque, L., Mio, R., Mattioli, J., Brais, E., Blais, N., Lalonde, Y., Maragh, M. and Gilliland, D.G. (1996) Nonrandom X-inactivation patterns in normal females: lyonization ratios vary with age. Blood, 88, 59–65.[Abstract/Free Full Text]

  9. El Kassar, N., Hetet, G., Briere, J. and Grandchamp, B. (1998) X-chromosome inactivation in healthy females: incidence of excessive lyonization with age and comparison of assays involving DNA methylation and transcript polymorphisms. Clin. Chem., 44, 61–67.[Abstract/Free Full Text]

  10. Percec, I., Plenge, R.M., Nadeau, J.H., Bartolomei, M.S. and Willard, H.F. (2002) Autosomal dominant mutations affecting X inactivation choice in the mouse. Science, 296, 1136–1139.[Abstract/Free Full Text]

  11. Pegoraro, E., Whitaker, J., Mowery-Rushton, P., Surti, U., Lanasa, M. and Hoffman, E.P. (1997) Familial skewed X inactivation: a molecular trait associated with high spontaneous-abortion rate maps to Xq28. Am. J. Hum. Genet., 61, 160–170.[ISI][Medline]

  12. Villard, L., Levy, N., Xiang, F., Kpebe, A., Labelle, V., Chevillard, C., Zhang, Z., Schwartz, C.E., Tardieu, M., Chelly, J., Anvret, M. and Fontes, M. (2001) Segregation of a totally skewed pattern of X chromosome inactivation in four familial cases of Rett syndrome without MECP2 mutation: implications for the disease. J. Med. Genet., 38, 435–442.[Abstract/Free Full Text]

  13. Racchi, O., Mangerini, R., Rapezzi, D., Rolfo, M., Gaetani, G.F. and Ferraris, A.M. (1998) X chromosome inactivation patterns in normal females. Blood Cells Mol. Dis., 24, 439–447.[CrossRef][ISI][Medline]

  14. Hoffbuhr, K., Devaney, J.M., LaFleur, B., Sirianni, N., Scacheri, C., Giron, J., Schuette, J., Innis, J., Marino, M., Philippart, M. et al. (2001) MeCP2 mutations in children with and without the phenotype of Rett syndrome. Neurology, 56, 1486–1495.[Abstract/Free Full Text]

  15. Shahbazian, M.D., Sun, Y. and Zoghbi, H.Y. (2002) Balanced X chromosome inactivation patterns in the Rett syndrome brain. Am. J. Med. Genet., 111, 164–168.[CrossRef][ISI][Medline]

  16. Hammer, S., Dorrani, N., Hartiala, J., Stein, S. and Schanen, N.C. (2003) Rett syndrome in a 47XXX patient with a de novo MECP2 mutation. Am. J. Med. Genet., 122A, 223–226.

  17. Plenge, R.M., Stevenson, R.A., Lubs, H.A., Schwartz, C.E. and Willard, H.F. (2002) Skewed X-chromosome inactivation is a common feature of X-linked mental retardation disorders. Am. J. Hum. Genet., 71, 168–173.[CrossRef][ISI][Medline]

  18. Dotti, M.T., Orrico, A., de Stefano, N., Battisti, C., Sicurelli, F., Severi, S., Lam, C.W., Galli, L., Sorrentino, V. and Federico, A. (2002) A Rett syndrome MECP2 mutation that causes mental retardation in men. Neurology, 58, 226–230.[Abstract/Free Full Text]

  19. Thomas, G.H. (1996) High male: female ratio of germ-line mutations: an alternative explanation for postulated gestational lethality in males in X-linked dominant disorders. Am. J. Hum. Genet., 58, 1364–1368.[ISI][Medline]

  20. Trappe, R., Laccone, F., Cobilanschi, J., Meins, M., Huppke, P., Hanefeld, F. and Engel, W. (2001) MECP2 mutations in sporadic cases of Rett syndrome are almost exclusively of paternal origin. Am. J. Hum. Genet., 68, 1093–1101.[CrossRef][ISI][Medline]

  21. Girard, M., Couvert, P., Carrie, A., Tardieu, M., Chelly, J., Beldjord, C. and Bienvenu, T. (2001) Parental origin of de novo MECP2 mutations in Rett syndrome. Eur. J. Hum. Genet., 9, 231–236.[CrossRef][ISI][Medline]

  22. Orrico, A., Lam, C., Galli, L., Dotti, M.T., Hayek, G., Tong, S.F., Poon, P.M., Zappella, M., Federico, A. and Sorrentino, V. (2000) MECP2 mutation in male patients with non-specific X-linked mental retardation. FEBS Lett., 481, 285–288.[CrossRef][ISI][Medline]

  23. Imessaoudene, B., Bonnefont, J.P., Royer, G., Cormier-Daire, V., Lyonnet, S., Lyon, G., Munnich, A. and Amiel, J. (2001) MECP2 mutation in non-fatal, non-progressive encephalopathy in a male. J. Med. Genet., 38, 171–174.[Abstract/Free Full Text]

  24. Chen, R.Z., Akbarian, S., Tudor, M. and Jaenisch, R. (2001) Deficiency of methyl-CpG binding protein-2 in CNS neurons results in a Rett-like phenotype in mice. Nat. Genet., 27, 327–331.[CrossRef][ISI][Medline]

  25. Guy, J., Hendrich, B., Holmes, M., Martin, J.E. and Bird, A. (2001) A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat. Genet., 27, 322–326.[CrossRef][ISI][Medline]

  26. Braunschweig, D., Simcox, T., Samaco, R.C. and LaSalle, J.M. (2004) X-chromosome inactivation ratios affect wild-type MeCP2 expression within mosaic Rett syndrome and Mecp2–/+ mouse brain. Hum. Mol. Genet., 13, 1275–1286.[Abstract/Free Full Text]

  27. Young, J.I. and Zoghbi, H.Y. (2004) X-chromosome inactivation patterns are unbalanced and affect the phenotypic outcome in a mouse model of Rett syndrome. Am. J. Hum. Genet., 74, 511–520.[CrossRef][ISI][Medline]

  28. Tan, S.S., Williams, E.A. and Tam, P.P. (1993) X-chromosome inactivation occurs at different times in different tissues of the post-implantation mouse embryo. Nat. Genet., 3, 170–174.[CrossRef][ISI][Medline]

  29. Pelka, G.J., Watson, C.M., Christodoulou, J. and Tam, P.P. (2005) Distinct expression profiles of Mecp2 transcripts with different lengths of 3'UTR in the brain and visceral organs during mouse development. Genomics, 85, 441–452.[CrossRef][ISI][Medline]

  30. Shahbazian, M.D., Antalffy, B., Armstrong, D.L. and Zoghbi, H.Y. (2002) Insight into Rett syndrome: MeCP2 levels display tissue- and cell-specific differences and correlate with neuronal maturation. Hum. Mol. Genet., 11, 115–124.[Abstract/Free Full Text]

  31. Sotelo, C. (2004) Cellular and genetic regulation of the development of the cerebellar system. Prog. Neurobiol., 72, 295–339.[CrossRef][ISI][Medline]

  32. Avner, P., Prissette, M., Arnaud, D., Courtier, B., Cecchi, C. and Heard, E. (1998) Molecular correlates of the murine Xce locus. Genet. Res., 72, 217–224.[CrossRef][ISI][Medline]

  33. Plenge, R.M., Percec, I., Nadeau, J.H. and Willard, H.F. (2000) Expression-based assay of an X-linked gene to examine effects of the X-controlling element (Xce) locus. Mamm. Genome, 11, 405–408.[CrossRef][ISI][Medline]

  34. Plenge, R.M., Tranebjaerg, L., Jensen, P.K., Schwartz, C. and Willard, H.F. (1999) Evidence that mutations in the X-linked DDP gene cause incompletely penetrant and variable skewed X inactivation. Am. J. Hum. Genet., 64, 759–767.[CrossRef][ISI][Medline]

  35. McMahon, A., Fosten, M. and Monk, M. (1983) X-chromosome inactivation mosaicism in the three germ layers and the germ line of the mouse embryo. J. Embryol. Exp. Morphol., 74, 207–220.[ISI][Medline]

  36. Baader, S.L., Schilling, M.L., Rosengarten, B., Pretsch, W., Teutsch, H.F., Oberdick, J. and Schilling, K. (1996) Purkinje cell lineage and the topographic organization of the cerebellar cortex: a view from X inactivation mosaics. Dev. Biol., 174, 393–406.[CrossRef][ISI][Medline]

  37. Panning, B. and Jaenisch, R. (1996) DNA hypomethylation can activate Xist expression and silence X-linked genes. Genes Dev., 10, 1991–2002.[Abstract/Free Full Text]

  38. Gibson, J., Williamson, S., Arbuckle, S. and Christodoulou, J. (2005) X chromosome inactivation patterns of the brain in Rett syndrome: implications for the disease phenotype. Brain Dev., 27, 266–270.[CrossRef][ISI][Medline]

  39. Makedonski, K., Abuhatzira, L., Kaufman, Y., Razin, A. and Shemer, R. (2005) MeCP2 deficiency in Rett syndrome causes epigenetic aberrations at the PWS/AS imprinting center that affects UBE3A expression. Hum. Mol. Genet., 14, 1049–1058.[Abstract/Free Full Text]

  40. Horike, S., Cai, S., Miyano, M., Cheng, J.F. and Kohwi-Shigematsu, T. (2005) Loss of silent-chromatin looping and impaired imprinting of DLX5 in Rett syndrome. Nat. Genet., 37, 31–40.[CrossRef][ISI][Medline]

  41. Herrup, K. (1986) Cell lineage relationships in the development of the mammalian CNS: role of cell lineage in control of cerebellar Purkinje cell number. Dev. Biol., 115, 148–154.[CrossRef][ISI][Medline]