Skip Navigation


Human Molecular Genetics Advance Access originally published online on April 13, 2006
Human Molecular Genetics 2006 15(11):1769-1782; doi:10.1093/hmg/ddl099
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
15/11/1769    most recent
ddl099v1
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 Liu, J.
Right arrow Articles by Francke, U.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Liu, J.
Right arrow Articles by Francke, U.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Identification of cis-regulatory elements for MECP2 expression

Jinglan Liu1 and Uta Francke1,2,*

1Department of Genetics and 2Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA

* To whom correspondence should be addressed at: Department of Genetics, Stanford University School of Medicine, Beckman Center for Molecular and Genetic Medicine, 279 Campus Drive, Stanford, CA 94305-5323, USA. Tel: +1 6507258089; Fax: +1 6507258112; Email: ufrancke{at}stanford.edu

Received February 17, 2006; Revised March 31, 2006; Accepted April 5, 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Rett syndrome (RTT) is an X-linked dominant disabling neurodevelopmental disorder caused by loss of function mutations in the MECP2 gene, located at Xq28, which encodes a multifunctional protein. MECP2 expression is regulated in a developmental stage and cell-type-specific manner. The need for tightly controlled MeCP2 levels in brain is strongly suggested by neurologically abnormal phenotypes of mouse models with mild overexpression and by mental retardation in human males with MECP2 duplication. We set out to identify long-range cis-regulatory sequences that differentially regulate MECP2 transcription and, when mutated, may contribute to the pathogenesis of RTT, autism or X-linked mental retardation. By inter-species sequence comparisons, we detected 27 highly conserved non-coding DNA sequences within a 210 kb region covering MECP2. We functionally confirmed four enhancer and two silencer elements by performing luciferase reporter assays in four different human cell lines. The transcription factor binding capability of the identified regulatory elements was tested by gel shift assays. To locate the human MECP2 core promoter, we dissected the promoter region by reporter assays with deletion constructs. We then used chromosome conformation capture methods to document long-range interactions of three enhancers and two silencers with the MECP2 promoter. Acting over distances of up to 130 kb, these elements may influence chromatin configurations and regulate MECP2 transcription. Our study has defined the ‘MECP2 functional expression module’ and identified enhancer and silencer elements that are likely to be responsible for the tissue-specific, developmental stage-specific or splice-variant-specific control of MeCP2 protein expression.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Rett syndrome (RTT) (OMIM 372150) is an X-linked dominant disabling neurodevelopmental disorder caused by mutations in the MECP2 gene that encodes a methyl-CpG binding protein and is located on Xq28 (1Go). Most cases are sporadic due to new mutations. The reported incidence ranges from 1 in 10 000 to 1 in 15 000 female births. Rare males affected with RTT are either X-chromosomal aneuploids or somatic mosaics. Males hemizygous for MECP2 mutations may have a more severe disorder, called congenital encephalopathy, or X-linked mental retardation, based on the type of mutation (reviewed in 2Go). In classic RTT, developmental delay becomes apparent at 6–18 months of age following an early period of grossly normal development. Major clinical features include loss of motor and verbal skills, mental retardation, stereotypic hand movements, seizures, growth deficiency and autonomic system dysfunctions. Although 70–80% of RTT females have mutations in the coding regions of exons 2–4 of the MECP2 gene, other cases might result from mutations in non-coding regions (3Go), or large deletions (4Go,5Go). Variant forms of RTT with infantile-onset of seizures may be caused by mutations in another gene, STK9/CDKL5 (6Go,7Go). Mild overexpression of wild-type MeCP2 protein induced neurodevelopmental abnormalities in transgenic mice (8Go,9Go), and duplication of the MECP2 gene causes mental retardation in human males (10Go), indicating the need for tightly controlled MeCP2 levels in vivo.

MECP2 is expressed in many tissues, but expression is highest in the brain. During development, MECP2 expression is very low or absent in immature neurons, then it increases during neuronal maturation and is highest in post-mitotic post-migratory neurons. There it remains high throughout adult life (11Go,12Go). This developmental-time-specific expression profile indicates that MeCP2 may function primarily in neuronal maturation and dendritic arborization and maintenance, rather than in neurogenesis or migration.

Originally identified as a methyl-CpG binding protein (13Go), MeCP2's primary role was thought to form a link between DNA methylation, histone acetylation and co-repressor molecules (14Go). Evidence is now accumulating for a more expanded role of MeCP2 as a multifunctional protein involved in the modulation of chromatin structure by different mechanisms. For example, MeCP2 also binds to unmethylated DNA and affects chromatin condensation by a methylation- and HDAC-independent mechanism (15Go). It recruits histone-methyl transferase activity that leads to histone H3 lysine 9 methylation, a hallmark of repressive chromatin (16Go). Furthermore, MeCP2 contains an RNA-binding RG repeat region and binds RNA or double-stranded methyl CpG-containing DNA in a mutually exclusive manner in vitro (17Go). When the RNA-binding protein YB1 was identified as a MeCP2-interacting protein, a novel function for MeCP2 in regulating alternative splicing was proposed (18Go). The mechanisms that regulate the cell and developmental stage-specific MECP2 expression patterns are not known (19Go).

At another level of transcriptional regulation, alternative splicing generates two transcripts that differ at the protein level. The initially reported transcript uses the initiator ATG codon in exon 2, whereas the more recently identified alternative form skips exon 2 and uses an ATG in exon 1 as initiator codon. This predominant isoform (MeCP2 e1) contains 24 amino acids encoded by exon 1 and lacks the nine amino acids encoded by exon 2 (3Go,20Go). The 3'-UTR contains different polyadenylation sites that give rise to two main mRNA variants (21Go). The ~10 kb transcript predominates during intrauterine life, whereas the short ~2 kb transcript is more abundant in adult brain. All these different levels of regulation make MECP2 an ideal candidate for the study of cis-regulatory elements.

Cis-regulatory transcriptional elements, such as enhancers and silencers, control the spatio-temporal expression of their cognate gene by specifically recruiting transcription activating or silencing factors that bind to the promoter region, whereas the core promoter alone only supports a basal level of transcription (22Go). Inter-species sequence comparisons have led to the discovery of common mammalian, as well as primate-specific, functional non-coding elements in the human genome (23Go). Mutations in such non-coding conserved sequences can lead to human disease, e.g. point mutations in an enhancer residing in a non-coding region that is conserved between human, mouse and Fugu rubripes cause a form of preaxial polydactyly (24Go). This enhancer lies within the intronic region of a different gene and regulates the topology of expression of sonic hedgehog (Shh) in limbs, from a distance of 1 Mb away. Recently, a whole genome comparison of human and Fugu revealed ~1400 highly conserved non-coding sequences, almost all of which are located in the vicinity of genes that either regulate vertebrate development or function as transcriptional regulators (25Go). These findings suggest that highly conserved genetic information coordinates the developmental processes of vertebrates by finely adjusting the transcription levels of specific genes.

The 75 kb human MECP2 genomic region is characterized by three salient features: a very large intron 2 (~60 kb), a 3'-UTR of 8.5 kb with highly conserved regions and different polyadenylation sites and a 40 kb intergenic region separating MECP2 from the nearest upstream gene. We hypothesized that these regions may harbor long-range cis-regulatory sequences that may differentially regulate MECP2 transcription in tissue-specific, developmental-stage-specific or protein-isoform (splice variant)-specific manners and that mutations at the regulatory elements could contribute to the pathogenesis of RTT and related disorders.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Identification of 27 non-coding DNA sequences that are conserved between human and mouse in the MECP2 genomic region
We compared a 210 kb human genomic region on Xq28, including the non-coding parts of MECP2 and the ~100 kb downstream and ~40 kb upstream intergenic regions, with the corresponding mouse sequence by using ECR browser (http://ecrbrowser.dcode.org/). We identified 27 conserved DNA sequences that met the criteria of at least 100 bases in length and at least 70% identity. Thirteen conserved non-coding sequences are located in the intragenic regions of MECP2, and most of those are at least 90% identical with the corresponding mouse sequences. In addition, nine conserved elements are located downstream and five upstream of MECP2 (Fig. 1). We designed primers to amplify the 27 fragments by PCR. The amplification products (F1–F27) and the conserved non-coding regions they represent are listed in Table 1.


Figure 0991
View larger version (65K):
[in this window]
[in a new window]
  		Figure 1. Human–mouse conserved segments in the Xq28 genomic region around MECP2/Mecp2.The plot, derived from the DCODE Comparative Genomics Center ECR genome browser (http://ecrbrowser.dcode.org), shows four contiguous panels from Xq28 in the orientation centromere to telomere. The baseline represents the human genomic sequence. Top: the genomic structure and transcriptional orientation are indicated for the genes HCFC1, IRAK1, MECP2 and OPNILW, and for the transcript CXorf12. Exons 1–4 of the MECP2 gene are identified. Red horizontal bars denote evolutionary conserved region (ECR) tracks. Green horizontal bars indicate the location of repetitive elements. Peaks represent sequence conservation between human and mouse. Color codes: blue, coding exon; yellow, UTR; pink, intron; red, intergenic element. ECR fragments F1–F27 were selected based on a threshold of 100 bp minimum length with at least 70% identity.

 

View this table:
[in this window]
[in a new window]
  		Table 1. List of 27 conserved non-coding sequences located within or near the human MECP2 genomic region
 
Identification of four enhancer and two silencer elements in the 210 kb region encompassing MECP2
To assess their functional significance by luciferase reporter assays, we cloned the 27 amplified conserved non-coding fragments into different pGL3 reporter vectors (26Go). The pGL3-enhancer vector that lacks an internal promoter was used to test three fragments for promoter activity: F1 located in the MECP2 promoter region, and F14 and F15 located in the IRAK1 promoter region. The other 24 fragments were cloned into a pGL3-promoter vector designed to test exogenous sequences for enhancer or silencer activity as this vector lacks an internal enhancer. All cloned plasmids were sequenced to exclude any with point mutations acquired during the PCR and cloning process.

To detect possible enhancer, repressor or promoter activities of the cloned fragments, we co-transfected them into SK-N-SH neuroblastoma cells using pTK plasmids as internal control. In this dual luciferase reporter assay, seven fragments, F11, F16, F17, F18, F19, F20 and F21, significantly increased luciferase expression levels (P<0.05), whereas four sequences, F3, F13, F24 and F26, reduced luciferase expression from the pGL3-promoter present in the vector (Supplementary Material, Fig. S1 and Table S1). To examine the putative repressor activity more stringently, fragments F3, F13, F24 and F26 were re-cloned into the pGL3-control vector that contains both internal promoter and enhancer elements. In these vectors, F3 and F13 were able to reduce the expression of the luciferase reporter protein, whereas F24 and F26 showed no significant inhibitory activity (Supplementary Material, Table S2). Therefore, only F3 and F13 were confirmed as potential transcriptional silencers in neuroblastoma cells.

Interestingly, F18 and F19, showing the most dramatic transcriptional enhancement activities, and F20 (Supplementary Material, Fig. S1 and Table S1) are located just upstream of the recently identified gene HCFC1. Although the DCODE program had originally identified this region as intergenic, the three sequences, which overlap each other, are likely to harbor the promoter for HCFC1 and were, therefore, excluded from further studies.

To evaluate tissue specificity of the potential regulatory sequence elements, we transfected three additional human cell lines of different tissue origins, astrocytoma (CRL1718), fibrosarcoma (HT1080) and epitheloid carcinoma (HeLa), with the potential cis-regulatory elements. Although enhancers F11, F16, F17 and F21 were all active in SK-N-SH cells, the data were statistically significant only for F11 in HT1080 and HeLa and for F16 in CRL1718 (Fig. 2A and Table 2) (Supplementary Material, Table S2). When silencers F3, F13 and F24 were tested, F3 functioned in all cell lines and silencer F13 only functioned in the neuroblastoma cell line, indicating neuronal-specific roles for these elements. The silencing function of F24 was limited to CRL 1718 astrocytoma cells (Fig. 2B and Table 2) (Supplementary Material, Table S2).


Figure 0992
View larger version (17K):
[in this window]
[in a new window]
  		Figure 2. Transcriptional activity of conserved non-coding genomic sequences as detected in pGL3 reporter plasmids. The pGL3-promoter vector was used as cloning vector to assay enhancer activity, whereas the pGL3-control vector was used to assay repressor activity. The constructs were transiently transfected into four different human cell lines. (A) Fragments F11, F16, F17 and F21 have enhancer activity. They increase reporter plasmid expression, compared with the enhancerless pGL3-promoter vector, but to variable levels in different human cell lines. (B) Fragments F3, F13 and F24 have silencer activity. They reduce the reporter expression when transiently transfected into human cell lines. F3 universally inhibits reporter expression in all four cell lines, whereas F13 specifically reduces the transcription activity in SK-N-SH cells. F24 exhibits silencer function only in CRL1718 astrocytoma cells. Displayed results are from three independent experiments. The complete initial reporter assay results for all 27 fragments are shown in Supplementary Material, Figure S1 and Table S1. The complete statistical data for Figure 2 can be found in Supplementary Material, Table S2.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Identification of enhancers and silencers among the candidate sequences in four different human cell lines by reporter assays
 
Characterization of the MECP2 core promoter
In the luciferase reporter assay, the conserved fragment F1, located immediate upstream of MECP2 exon 1 and spanning nucleotides –1071 to +9, showed a high level of promoter activity (Supplementary Material, Fig. S1 and Table S1). To delineate the functional promoter elements, we created serial deletions of the F1 segment, representing nucleotides 89142–90221 of the AF030876 [GenBank] sequence (gi:22830571) (Supplementary Material, Fig. S2A). DNA fragments were amplified by PCR, sequenced and cloned into pGL3-basic vectors to test their promoter activities in human neuroblastoma, fibrosarcoma, astrocytoma and HeLa cell lines. The core promoter should be able to induce basal transcriptional activity in all cells regardless of their tissue origin (22Go). The proximal region from –179 to –309 uniformly triggered luciferase reporter expression in all four cell lines, indicating that it harbors core promoter elements for the human MECP2 gene (P<0.05) (Fig. 3) (Supplementary Material, Fig. S2B and Table S3). Region –309 to –370 reduced luciferase expression by 2-, 3.5-, 18- and 3-fold in SK-N-SH, HT1080, CRL1718, and HeLa cells, respectively. This result suggests the presence of an embedded silencer element between –309 and –370 that is predominantly active in the astrocytoma CRL1718 cells. Region –847 to –1071 increased reporter gene expression by 2–3-fold in all cell types, indicating that it harbors a universal enhancer element. In addition, a weak negative regulatory element may be located between –553 and –681 that is active in SK-N-SH and HT1080 cells, but not in CRL1718 and HeLa cells. In summary, we have identified –179 to –309 (nt 89330–89459 on gi:22830571) as the region containing the MECP2 core promoter.


Figure 0993
View larger version (16K):
[in this window]
[in a new window]
  		Figure 3. Identification of MECP2 core promoter by 5' end serial deletions of the ~1 kb MECP2 promoter sequence. The construct called ‘Full –1071/+9’ encompasses the MECP2 promoter from upstream position –1071 to position +9 within exon 1 cloned into the pGL3-basic vector. SD1–SD7 include serial 5' end deletions of the full-length promoter construct. Transcription activities produced by the truncated fragments were tested in four different cell lines: SK-N-SH, HT1080, CRL1718 and HeLa cells. The results were compared with the activity of the enhancer- and promoter-containing pGL3-control vector set at 100% for each cell line. The proximal segment –179 to –309 supports and segment –309 to –370 inhibits basal transcription. Segment –553 to –681 suppresses transcription only in SK-N-SH and HT1080 cells, but not in CRL1718 and HeLa cells. The pGL3-basic vector was used for cloning, pRL-TK plasmids were co-transfected as internal control for transfection efficiency. Three independent experiments were performed in quadruplicate. The data and statistics are shown in Supplementary Material, Table S3.

 
Detection of transcription factor binding capability of the identified regulatory elements
To test the ability of the four enhancer and two silencer elements to form complexes with nuclear proteins, we performed electrophoretic mobility shift assays (EMSA) using crude nuclear extracts from SK-N-SH, H1080, CRL1718 and HeLa cells. To generate probes for labeling, we PCR-amplified 39 overlapping DNA segments, each 100–150 bp in length, that cover the six regulatory elements previously identified in the reporter assays. A 150 bp DNA fragment that did not produce band shifts in preliminary EMSA experiments was used as non-specific control probe. A competition assay was performed with unlabeled specific and non-specific probes to distinguish the specific protein-binding capability for each of the 39 segments. As an example, in Figure 4, the labeled fragment S8 produced two shifted bands when incubated with SK-N-SH nuclear extract (lane 7), whereas the non-specific probe generated no specific banding pattern (lane 6). Adding unlabeled non-specific DNA at 100x or 10x molar excess did not affect the shifted banding patterns (lanes 1 and 2). In contrast, increasing the amount of unlabelled competitor S8 by 10-, 50- and 100- fold progressively reduced the binding signals to zero (lanes 5, 4 and 3). These results demonstrate that S8 is able to specifically bind to nuclear proteins from SK-N-SH cells. The complete results for all 39 sub-fragments, tested in four cell lines, are summarized in Table 3. In brief, we discovered that all six regulatory elements harbor regions that are able to interact with nuclear proteins, presumably transcriptional activator or repressor proteins. No tissue-specific information could be gleaned from these results, because no distinctive band shift pattern was observed in particular cell lines. Notably, none of the silencer fragments generated band shifts in HeLa cells (Table 3).


Figure 0994
View larger version (91K):
[in this window]
[in a new window]
  		Figure 4. Competition EMSA reveals that segment S8, within enhancer F11, binds to nuclear proteins. Crude nuclear protein extract from SK-N-SH cells was incubated with labeled S8, as specific probe (lane 7), or with a similar-sized labeled DNA fragment as non-specific probe (lane 6). The two bands associated with labeled S8 (marked with arrows) were not changed by incubation with 100x or 10x excess of unlabeled non-specific probe (lanes 1 and 2). The dosage-dependent competition with 100x, 50x or 10x molar excess of unlabeled specific probe, however, revealed the specificity of the association between S8 and both protein complexes (lanes 3–5). Lanes 8 and 9 are controls lacking nuclear extract.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Summary of EMSA results in four different cell lines
 
Long-range interactions of three enhancers and two silencers with the MECP2 promoter
Spatial and temporal regulation of gene transcription requires the interaction between certain cis-regulatory elements and the cognate gene promoter (22Go). To test whether the identified elements regulate MECP2 transcription by interacting with the promoter, we performed chromosome conformation capture (3C) assays (27Go). Chromatin from formaldehyde fixed SK-N-SH cells was solubilized and digested with the palindromic restriction enzyme Sau3AI. Before ligation, the digested chromatin was diluted to <2.5 ng/µl to ensure ‘intramolecular’ ligation of restriction fragments that were located near each other in vivo. The purified ligated restriction fragments were used as templates for PCR amplifications of specific ligation products between the MECP2 promoter region and each of the six regulatory elements. The primers were designed towards to the Sau3AI sites. The genomic localizations of the cis-regulatory elements and primers are shown in Figure 5A. In brief, two genomic sequence contigs AF030876 [GenBank] (gi:22830571) and U52112 [GenBank] (gi:22773272) cover the promoter and the cis-regulatory elements and encompass four known genes MECP2, IRAK1, CXorf12 and HCFC1. On the telomeric side of MECP2, the flanking locus OPN1LW is located at the border of our screening region. We were able to amplify specific PCR fragments with primers corresponding to the MECP2 promoter region and to some, but not all, cis-regulatory elements (Fig. 5B). DNA sequence tracings across the Sau3AI ligation sites in these chimeric PCR products are provided in Supplementary Material, Figure S3.


Figure 0995
Figure 0995
View larger version (33K):
[in this window]
[in a new window]
  		Figure 5. Interaction between MECP2 promoter and long-range cis-regulatory elements is demonstrated by chromosome conformation capture (3C) assay. (A) Schematic representation of the 210 kb genomic region with localization of sequenced genomic clones (horizontal black bars), gene loci, and MECP2 promoter (P), enhancers F21, F17, F16 and F11, and silencers F13 and F3. Hatched boxes represent known genes including exons, introns and UTRs, vertical arrows point to the location of each cis-regulatory element. On the expandad scale below, Sau3AI digestion sites are indicated as black bars. 3C PCR primers for amplification when paired with a promoter primer are identified by numbers. Horizontal arrows point to the direction of amplification. Primers shown in bold did amplify 3C ligation products. (B) PCR amplification products from 3C analysis in neuroblastoma nuclei with primers pairs identified by roman numerals: I, 8414/8422; II, 8414/8432; III, 8414/8434; IV, 8414/8452; V, 8414/8454; VI, 8414/8461; VII, 8414/8463; VIII, 8414/8467. Lane 1, product of 3C experiments (3C sample); lane 2, 3C sample without ligase; lane 3, 3C sample without crosslinking; lane 4, genomic DNA; lane 5, blank; M, marker lanes with 100 bp ladder.

 
We discovered that three of the enhancers, F11, F17 and F21, but not F16, interact with the promoter region of MECP2. F11 is located in the 3'-UTR of MECP2, F17 is in the downstream intergenic region between IRAK1 and CXorf12, and F21 is further downstream in intron 1 of HCFC1. Their distances from the MECP2 promoter are 74, 114 and 130 kb, respectively. The enhancer fragment F16 is 97 kb downstream from the promoter and also located in the intergenic region between IRAK1 and CXorf12, 17 kb upstream of F17, but we were unable to demonstrate interaction of F16 with the MECP2 promoter. Because in the reporter assays (Fig. 2A) (Supplementary Material, Fig. S1 and Table S1) F16 showed robust enhancer activity in the neuroblastoma cells, we suspect that F16 may be an enhancer for another gene. Both negative regulatory elements F3 and F13 were shown to interact with the MECP2 promoter. F3 is located within MECP2 intron 1, ~5 kb from the promoter, and F13 is in the MECP2 3'-UTR, 76 kb from the promoter and 2 kb downstream from the enhancer F11.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Although expression of MECP2 is highly regulated during brain development in neuronal cell lineages while expression is repressed in glia cells, the elements that contribute to this exquisite regulation are completely unknown. By using interspecies sequence comparisons and reporter plasmid transfection assays, we have identified four enhancers, F11, F16, F17 and F21, and two silencers, F3 and F13, out of 27 highly conserved non-coding sequences within a 210 kb genomic region surrounding the human MECP2 locus. Gel shift assays verified their ability to bind nuclear protein(s) in vivo. 3C assays detected long-distance interactions between each of five elements and the MECP2 promoter, strongly suggesting involvement of these elements in regulating MECP2 transcription. Only F16 with high enhancer activity failed in the 3C assay. F16 is located in the 27 kb intergenic region between IRAK1 and CXorf12, distant from any known coding sequences. In our reporter assays, F16 induced one of the highest levels of luciferase expression among all tested sequences. This element may act as an enhancer only in mature neurons or in a distinct neuronal subtype, but not in neuroblasts. Alternatively, F16 could be a strong enhancer for a gene other than MECP2 or a promoter for an unidentified gene embedded in this 27 kb interval. F18, F19 and F20, also showing high level enhancer activities, are located just upstream of HCFC1 exon 1, and we suspect that they represent HCFC1 promoter segments. We included them in our initial screening studies because HCFC1 had not yet been identified in that region in the genome databases.

The 1 kb sequence upstream of the human MECP2 transcription start site was identified as having promoter activity in our study, and this region covers the corresponding mouse Mecp2 promoter segment (–677/+56) which is sufficient to drive Mecp2 expression in neurons, but not in other cells, during mouse development (28Go). By testing serial deletion constructs, we localized the core promoter region of human MECP2 between –179 and –309 bp upstream of the transcription start site. This core promoter supports basal gene expression from reporter plasmids in all four human cell lines with different tissue origins. In addition, we identified two positive regulatory elements between nucleotide positions –681 and –847, and –847 and –1071, respectively, as well as two negative regulatory elements between –309 and –370, and –553 and –681. Adachi et al. (28Go) carried out 5' RACE and identified a different transcriptional start site for mouse Mecp2 located 47 nt upstream of the human start site. They demonstrated neuronal-specific promoter activity of a 19-nucleotide mouse sequence (–63 to –45) (28Go). This fragment corresponds to –110 to –92 of the human MECP2 promoter region sequence and is not conserved (Supplementary Material, Fig. S2B). Sequence similarities in the promoter regions of human and mouse MECP2/Mecp2 genes range between 50 and 80%, which is relatively low compared with other conserved non-coding sequences in our study. Considering that the binding consensus sequence for transcription factors is usually less than 10 nucleotides (http://www.genomatix.de), the less identical sequences between human and mouse MECP2/ Mecp2 promoters could lead to divergent recruitment of different sets of transcription factors in each species. Our study suggests that the promoter structure of human MECP2 is unique and, therefore, transcriptional regulation based on it may also be different from that in mouse. Many developmental genes have been documented to have significantly different expression patterns in human and mouse, such as WNT7A/wnt7a, WNT5/wnt5a, CAPN3 and others (29Go).

Cis-regulatory elements, such as enhancers and silencers, function to precisely regulate gene expression at the right time, in the right cells and at the correct level. They can act from various distances, several  kb to 1 Mb away, or even from locations in introns of unrelated genes (30Go). Among the cis-elements, we have identified for MECP2, only the silencer F3 is close to the promoter, residing in MECP2 intron 1 and ~5 kb downstream from the promoter. The other four are located long distances away, the enhancer F11 and the silencer F13 are in the MECP2 3'-UTR region, 74 and 76 kb distal to the promoter. The enhancer F17 is 114 kb downstream in the intergenic region between IRAK1 and CXorf12, and the enhancer F21 is 130 kb downstream and located in the first intron of HCFC1. Both enhancers were positive when tested for interaction with the MECP2 promoter by 3C analyses. This apparent interaction is unlikely to be caused by ‘random collisions’ (27Go), given the distances and the fact that several intervening fragments were negative in this assay.

We analyzed the putative cis-regulatory elements for sequence variants in the human population by searching BioMart (http://www.ensembl.org/Multi/martview) and the NCBI SNP bank (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=snp&cmd=search&term). The entire 1 kb promoter fragment contains a single SNP (rs7050893). Among the enhancers, only F17 has SNPs (rs12835942, rs5986946, rs5945173). Element F13 that functioned as a repressor in neuroblastoma cells, but not in other cell types, is quite polymorphic with five SNPs (rs3177341, rs7066787, rs3027912, rs3027913 and rs3027914). Located at the distal 3'-UTR of MECP2, F13 covers two polyadenylation signals and one polyadenylation site. To assess any functional role of these sequence variants requires further experimentation.

Developmentally regulated genes are more likely to have cis-regulatory elements in conserved non-coding regions than genes with other functions, such as structural proteins or enzymes (31Go). High levels of conservation of cis-regulatory elements are observed mainly in genes controlling various developmental stages such as genes encoding transcriptional regulators (23Go,32Go). MeCP2 is a universal transcriptional repressor when binding to methyl-CpG, but also plays a chromatin structural role (15Go), binds to matrix attachment regions (33Go), to DNA methyltransferase 1 (34Go) and to RNA splicing factors (18Go) and complexes with RNA in vitro (17Go). These multiple functions of MeCP2 may require an elaborate transcriptional control system.

The mechanism of long-distance controlling is still under investigation, but the looping of chromatin across a 40 Mb large interval to the transcription factory has been well documented (35Go). Recently, differential expression of the oppositely imprinted genes Igf2 and H19 was demonstrated to result from interchangeable interactions between various cis- regulatory elements over a 90 kb genomic region, followed by differential CpG methylation, which determines either the ‘active’ or the ‘inactive’ chromatin loop to be formed, and thus controls the epigenetic switch of the expression of two genes at the same locus (36Go,37Go).

We analyzed sequences of the MECP2 fragments that we identified to have promoter, enhancer and silencer activities for potential transcription factor binding sites through online prediction programs such as MatInspector at Genomatix (http://www.genomatix.de) (38Go). From the long list of corresponding proteins that bind to those sequences, several are involved in neurodevelopment (Table 4). Numerous predicted binding sites exist in the promoter and enhancer regions for members of the POU domain family of CNS-specific transcription factors. The distal promoter region also contains binding sites for neurogenins 1 and 3, which belong to a family of bHLH transcription factors involved in the determination of neural precursor cells in the neuroectoderm. Neurogenin 1 inhibits the differentiation of neural stem cells into astrocytes (39Go).


View this table:
[in this window]
[in a new window]
  		Table 4. Predicted transcription factor binding sites
 
General transcription factor Sp1 binding sites are most abundant in the proximal promoter fragments and in an EMSA-positive fragment of enhancer F21. Sp1 contains zinc-finger domains that bind GC or GT boxes, and MECP2 is a known in vivo target of Sp1. The expression of Mecp2 is 2-fold lower in heterozygous and hardly detectable in homozygous Sp1 knockout embryos (40Go). MYT1 (myelin transcription factor 1), a member of a neural-specific zinc-finger transcription factor family, acts in the specification of neurons and modulates the proliferation and differentiation of oligodendrocytes, the cells that form myelin. A screen for interacting proteins in yeast revealed that Myt1 interacted with the co-repressor Sin3B (41Go). In mammalian cells, MYT1 complexes also included HDAC1 and HDAC2, suggesting that Myt1 can act as a repressor by affecting chromatin structure. MYT1-binding sites are present in the MECP2 promoter, specifically between –553 and –681, where a negative element was identified in our reporter assays, as well as in enhancers F11 and F21 and silencer F13. We speculate that MYT1 could be part of a complex that keeps MECP2 expression low in oligodendrocytes. Other members of the complex are suggested by the presence of binding sites for the transcriptional repressors NFIL3 and BCL6 in the same EMSA-positive sub-fragment of silencer F13. In addition, REST that inhibits expression of neuronal genes in non-neuronal tissues has a predicted binding site at another sub-fragment of F13. Given that no neural-specific transcription factor binding sites were identified in the sequences of the other silencing element, F3, the results for F13 that is located at the end of the 3'-UTR strongly suggest that this region, possibly in interaction with the promoter, may be important for the tissue-specific regulation of MECP2 expression.

Loss of function mutations cause classic RTT and variant phenotypes in human females, congenital encephalopathy in human males and a distinct neurological disorder with early death in male mutant mice (reviewed in 2Go). In contrast, MeCP2 overexpression in transgenic mice produces a distinct progressive early-onset neurological phenotype (8Go,9Go). Even a 2-fold overexpression was pathogenic indicating that the MeCP2 protein level has to be tightly monitored in vivo. Recently, different sizes of intrachromosomal duplications at Xq28 have been discovered in familial cases of severely mentally retarded males by array-CGH and quantitative PCR methods (10Go,42Go). The duplicated regions vary from 0.4 to 0.8 Mb in size and encompass MECP2 and several other genes in its vicinity, including the mental retardation associated genes SLC6A8 and L1CAM, whereas the minimal common duplication region is <450 kb and only harbors one gene, MECP2. The distal border of the minimal region lies 300 kb telomeric to MECP2 and the proximal border just 5' of L1CAM. The enhancers and silencers we have characterized are localized exactly within the minimal duplication region covered by the same two overlapping sequences AF030876 [GenBank] (gi:22830571) and U52112 [GenBank] (gi:22773272) as reported (10Go). As cis-regulatory elements could function at long distances and may not be assigned to a particular gene, the elements identified in our study could conceivably also regulate the transcription of adjacent neurodevelopmental genes such as SLC6A8 and L1CAM. The duplication of this region could result in more complicated disorganization at the transcriptional level, which may lead to either gain or loss of function of neighboring genes. The corresponding phenotypes may be similar to those generated by coding region mutations or could be quite different due to the partial, tissue-specific loss or gain of expression (10Go).

Disruptions of long-distance cis-regulatory relationships cause several human diseases. Interestingly, most of the affected genes are key developmental regulators, and in most cases, the disrupted regions are located at a long distance in introns of functionally irrelevant genes (32Go,43Go). For instance, reduced expression of PAX6 causes Type II Aniridia with hypoplastic iris (OMIM 106210 [OMIM] ). The role of a regulatory domain, located 200 kb downstream within the ubiquitously expressed gene ELP4, was initially identified by the study of human translocations. Transgenic mouse model studies later revealed that this region contains retina and lens enhancers and could completely rescue lethality in homozygous Pax6 knockout mice and eye abnormalities in heterozygotes (44Go,45Go). Similar loss of function phenotypes associated with disruption of long-range interaction of control elements was also reported for TWIST in Saethre–Chotzen syndrome (OMIM 101400 [OMIM] ) (46Go), POU3F4 in X-linked deafness type 3 (OMIM 304400 [OMIM] ) (47Go), MAF in congenital cataracts and ocular anterior segment anomalies (48Go,49Go), SOX9 in autosomal sex reversal (50Go), forkhead family genes FOXC2 in lymphedema distichiasis (OMIM 153400 [OMIM] ) (51Go) and FOXL2 in blepharophimosis–ptosis–epicanthus inversus syndrome (OMIM 110100 [OMIM] ) (52Go) and others. Therefore, Kleinjan and van Heyningen (43Go) proposed an alternative definition for genes as ‘functional expression modules’, which ‘encompass both the transcribed regions and their cis-regulatory control systems and function appropriately in different cell types within the context of the local chromatin architecture’. Our study has defined the ‘MECP2 functional expression module’ that is likely to be required for the stringent control of MeCP2 protein levels during neuronal maturation and in post-mitotic neurons. Further research is required to define the cell-type and developmental-stage-specific characters of the cis-regulatory elements and to identify the transcription factors that bind to them. This information will be required before gene therapy for RTT or other MECP2 mutation related diseases could be seriously considered.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
SUPPLEMENTARY MATERIAL
 REFERENCES
 
Cell lines and culture conditions
Human neuroblastoma cell line SK-N-SH, astrocytoma cell line CRL1718, fibrosarcoma cell line HT1080 and epitheloid carcinoma cell line HeLa were obtained from ATCC (American Type Culture Collection, Manassas, VA, USA). SK-N-SH and CRL1718 cells were grown in RPMI 1640 supplemented with 10% fetal bovine serum, 100 U of penicillin–streptomycin at 37°C with 5% CO2, HT1080 and HeLa cells were grown under the same conditions but in DMEM.

Use of online bioinformatics tools
For the identification of conserved non-coding elements by human-mouse sequence alignments, we used DCODE ECR browsers (http://ecrbrowser.dcode.org/), corresponding to the human May 2004 (hg17) assembly and the mouse May 2004 (mm5) assembly (53Go). Non-coding intergenic and intronic genomic sequences containing a minimum of 100 bases with at least 70% identity were chosen. For the alignment of human and mouse MECP2/Mecp2 promoter sequences, we used DCODE zPicture (http://zpicture.dcode.org).

To identify SNPs in functional cis-regulatory elements, we searched the Ensemble Data mining tool BioMart (http://www.ensembl.org/Multi/martview) and the NCBI SNP bank (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=snp&cmd=search&term).

We used the MatInspector at Genomatix (http://www.genomatix.de/index.html) to identify candidate transcription factors that could bind to the EMSA-positive cis-regulatory elements and to select those which may regulate gene expression during nervous system development.

Construction of luciferase reporter plasmids
Primers were designed to amplify 27 DNA fragments ranging from 500 to 1200 bp in size. To amplify putative promoter elements, NheI and XhoI restriction sites or HindIII and XhoI restriction sites were included at the 5' end of the forward and reverse primers, respectively, to maintain the promoter orientation while cloning into the luciferase test vector. Fragments to be tested for enhancer or repressor activity had KpnI and MluI restriction sites included at the 5' end of the forward and reverse amplification primers, respectively, for cloning of the PCR fragments into luciferase reporter vectors. The fragments were inserted in the sense orientation of the MECP2 transcript. Primer sequences and PCR conditions are available on request. PCR products were extracted from agarose gel slices by using the QIAquick gel extraction kit (Qiagen) and digested with the corresponding enzymes to generate sticky ends for cloning. Digested products were ligated upstream of the firefly luciferase gene into pGL3 reporter vectors (Promega Technical Manual: pGL3-basic Cat. no. E1751; pGL3-control Cat. no. E1741; pGL3-enhancer Cat. no. 1771; pGL3-promoter Cat. no. 1761). The putative promoter elements were ligated to pGL3-enhancer vector, whereas suspected enhancer or silencer elements were ligated into pGL3-promoter and pGL3-control vector, respectively. The pGL3-basic vector was used for the promoter truncation assays.

Transfection and luciferase assay
When cells were ~80% confluent, 50 ng of each experimental pGL3 luciferase plasmid and 5 ng of the Renilla luciferase-containing pRL-TK control plasmid (Promega) were co-transfected into SK-N-SH, CRL1718, HT1080 and HeLa cells by using the FuGENE6 Lipofectamine Reagent (Roche) at a DNA:FuGENE ratio of 1:6. For each transfection assay, the pGL3-control vector containing the internal promoter and enhancer elements was used as positive control and the pGL3-basic vector with neither internal promoter nor internal enhancer element was used as negative control. Transfections of the empty pGL3-promoter and pGL3-enhancer vectors also served as controls. After 24 h, cell lysates were prepared from each transfected culture and were transferred to 96-well white microtiter plates. The firefly and Renilla luciferase activities were analyzed in a 96-well plate luminometer (Perkin Elmer Wallace) according to manufacturer's protocol in the Dual Luciferase Kit (Promega). Each fragment was tested in quadruplicate in each of four cell lines, and each experiment was repeated three or four times. To control for transfection efficiency, the ratio of firefly luciferase signal to Renilla signal was calculated for each transfected sample. To determine the strength of the potential cis-regulatory activity of each DNA fragment, the signal ratio for the negative control (pGL3-basic vector) for each cell line was subtracted. The resulting value was then divided by the positive control ratio (pGL3-control vector) to get the normalized value. The normalized value for each DNA fragment represents its activity expressed as percentage of the positive control. Finally, for each cell type, a paired Student's t-test was performed. We compared the activities of reporter constructs containing cloned DNA fragments with those of the corresponding empty vectors to calculate P-values.

Electrophoretic mobility shift assay
Six candidate non-coding regions were selected for EMSA studies. A total of 39 100–150 bp DNA fragments were synthesized by PCR and sequenced to exclude possible point mutations. Experimental conditions were optimized as described (22Go). Briefly, nuclear proteins were extracted from SK-N-SH, CRL1718, HT1080 and HeLa cells using the Nuclear Extract Kit (Active Motif). Their concentrations were measured by the Bradford assay following manufacturer's instruction (Bio-Rad). Each DNA fragment was end-labeled with 32{gamma}-ATP. Labeled (20 000 c.p.m.) probe was incubated at 4°C for 20 min with 10 µg nuclear protein, 8 µg dI–dC polymers, and 1x binding buffer. One specific unlabeled probe and one similar-sized non-specific unlabeled probe were used for the competition assays. The samples were electrophoresed in 4.5% non-denaturing polyacrylamide gel for 3 h at 200 V and 4°C, and then the gel was exposed to autoradiographic film (Kodak) for 16–24 h.

3C assay
Experimental procedures were performed according to published methods with modifications (37Go,54Go). Briefly, 4x107 neuroblastoma SK-N-SH cells were fixed with formaldehyde at the final concentration of 2% for 10 min at room temperature. Glycine (2 M) was added to a final concentration of 0.125 M to stop the cross-linking. Cells were trypsinized for 15 min at 37°C and harvested by scraping. After washing with 2x cold PBS, the cells were centrifuged at 2800g for 15 min and re-suspended in ice-cold lysis buffer [10 mM Tris (pH 8.0), 10 mM NaCl, 1% NP-40] with freshly added 0.1 mM PMSF and 1:500 protease inhibitor cocktail (Roche) and incubated at 4°C on a magnetic stirrer plate for 8 h. The nuclei were harvested by centrifuging at 1430g for 15 min, re-suspended in 1.2xNEB buffer 4 (New England Biolabs) containing 0.3% SDS and incubated at 37°C for 1 h with shaking. Triton X-100 was added to a final concentration of 1.8% to sequester the SDS, and the samples were incubated for 1 h at 37°C with shaking. The nuclei were counted by hemacytometry and aliquoted at 1x106 each. One aliquot was digested with 600 U Sau3AI at 37°C overnight with shaking. The restriction enzyme was inactivated by adding SDS to 1.6% and heating the sample to 65°C for 20 min. One-tenth of the digested chromatin was added to ligation buffer for a total volume of 800 µl. The final chromatin concentration was diluted to ~2.5 ng/µl to favor intramolecular ligation, and Triton X-100 was added to 1%, followed by incubation at 37°C for 1 h. After that 30 Weiss units of T4 DNA ligase were added and ligation was performed at 16°C for 4 h, followed by 30 min at room temperature. Samples were incubated at 65°C overnight with 100 µg/ml proteinase K followed by 0.4 µg/ml RNaseA treatment at 37°C for 30 min before phenol extraction and ethanol precipitation.

MECP2 promoter, enhancer and silencer specific PCR primers were designed to flank Sau3AI sites with an orientation that will amplify potential ligated chimeric sequences. The sizes of the predicted PCR products varied from 200 to 800 bp. Nested PCR assays were performed to ensure the specificity of the amplifications. After the second round of PCR amplification, specific bands were purified from agarose gel and sequenced. All PCRs were conducted in 25 µl reaction volumes under the same conditions: denaturation at 95°C for 5 min, followed by 35 cycles of amplification (95°C at 45 s, 58°C at 45 s and 72°C at 90 s). Primer sequences are available on request. Four independent experiments were carried out, and the final PCR products were sequenced.


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


   ACKNOWLEDGEMENTS
 
The authors are grateful to Dr Shelly Force Aldred for helpful discussions about reporter assays and statistical analysis, to the Myers lab for help with the luminometer measurements and to Dr Jianqun Ling for his valuable suggestions regarding 3C methods. This work was supported in part by grants from the NIH and the International Rett Syndrome Association.

Conflict of Interest statement. The authors declare that they have no conflict of interest.


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

  1. Amir, R.E., Van den Veyver, I.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. Francke, U. (2006) Mechanism of disease: neurogenetics of MeCP2 deficiency. Nat. Clin. Pract. Neurol., 2, 212–221.[Medline]

  3. Mnatzakanian, G.N., Lohi, H., Munteanu, I., Alfred, S.E., Yamada, T., MacLeod, P.J., Jones, J.R., Scherer, S.W., Schanen, N.C., Friez, M.J. et al. (2004) A previously unidentified MECP2 open reading frame defines a new protein isoform relevant to Rett syndrome. Nat. Genet., 36, 339–341.[CrossRef][ISI][Medline]

  4. Archer, H.L., Whatley, S.D., Evans, J.C., Ravine, D., Huppke, P., Bunyan, D., Kerr, A.M., Kerr, B., Sweeney, E. and Davies, S.J. et al. (2006) Gross rearrangements of the MECP2 gene are found in both classical and atypical Rett syndrome. J. Med. Genet., 43, 451–456.[Abstract/Free Full Text]

  5. Shi, J., Shibayama, A., Liu, Q., Nguyen, V.Q., Feng, J., Santos, M., Temudo, T., Maciel, P. and Sommer, S.S. (2005) Detection of heterozygous deletions and duplications in the MECP2 gene in Rett syndrome by robust dosage PCR (RD–PCR). Hum. Mutat., 25, 505.[Medline]

  6. Weaving, L.S., Christodoulou, J., Williamson, S.L., Friend, K.L., McKenzie, O.L., Archer, H., Evans, J., Clarke, A., Pelka, G.J., Tam, P.P. et al. (2004) Mutations of CDKL5 cause a severe neurodevelopmental disorder with infantile spasms and mental retardation. Am. J. Hum. Genet., 75, 1079–1093.[CrossRef][ISI][Medline]

  7. Scala, E., Ariani, F., Mari, F., Caselli, R., Pescucci, C., Longo, I., Meloni, I., Giachino, D., Bruttini, M., Hayek, G. et al. (2005) CDKL5/STK9 is mutated in Rett syndrome variant with infantile spasms. J. Med. Genet., 42, 103–107.[Abstract/Free Full Text]

  8. Collins, A.L., Levenson, J.M., Vilaythong, A.P., Richman, R., Armstrong, D.L., Noebels, J.L., David Sweatt, J. and Zoghbi, H.Y. (2004) Mild overexpression of MeCP2 causes a progressive neurological disorder in mice. Hum. Mol. Genet., 13, 2679–2689.[Abstract/Free Full Text]

  9. Luikenhuis, S., Giacometti, E., Beard, C.F. and Jaenisch, R. (2004) Expression of MeCP2 in postmitotic neurons rescues Rett syndrome in mice. Proc. Natl Acad. Sci. USA, 101, 6033–6038.[Abstract/Free Full Text]

  10. Van Esch, H., Bauters, M., Ignatius, J., Jansen, M., Raynaud, M., Hollanders, K., Lugtenberg, D., Bienvenu, T., Jensen, L.R., Gecz, J. et al. (2005) Duplication of the MECP2 region is a frequent cause of severe mental retardation and progressive neurological symptoms in males. Am. J. Hum. Genet., 77, 442–453.[CrossRef][ISI][Medline]

  11. Kishi, N. and Macklis, J.D. (2004) MECP2 is progressively expressed in post-migratory neurons and is involved in neuronal maturation rather than cell fate decisions. Mol. Cell. Neurosci., 27, 306–321.[ISI][Medline]

  12. Matarazzo, V., Cohen, D., Palmer, A.M., Simpson, P.J., Khokhar, B., Pan, S.J. and Ronnett, G.V. (2004) The transcriptional repressor Mecp2 regulates terminal neuronal differentiation. Mol. Cell. Neurosci., 27, 44–58.[Medline]

  13. Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I., Jeppesen, P., Klein, F. and Bird, A. (1992) Purification, sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA. Cell, 69, 905–914.[CrossRef][ISI][Medline]

  14. Nan, X., Campoy, F.J. and Bird, A. (1997) MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin. Cell, 88, 471–481.[CrossRef][ISI][Medline]

  15. Georgel, P.T., Horowitz-Scherer, R.A., Adkins, N., Woodcock, C.L., Wade, P.A. and Hansen, J.C. (2003) Chromatin compaction by human MeCP2. Assembly of novel secondary chromatin structures in the absence of DNA methylation. J. Biol. Chem., 278, 32181–32188.[Abstract/Free Full Text]

  16. Fuks, F., Hurd, P.J., Deplus, R. and Kouzarides, T. (2003) The DNA methyltransferases associate with HP1 and the SUV39H1 histone methyltransferase. Nucleic Acids Res., 31, 2305–2312.[Abstract/Free Full Text]

  17. Jeffery, L. and Nakielny, S. (2004) Components of the DNA methylation system of chromatin control are RNA-binding proteins. J. Biol. Chem., 279, 49479–49487.[Abstract/Free Full Text]

  18. Young, J.I., Hong, E.P., Castle, J.C., Crespo-Barreto, J., Bowman, A.B., Rose, M.F., Kang, D., Richman, R., Johnson, J.M., Berget, S. et al. (2005) Regulation of RNA splicing by the methylation-dependent transcriptional repressor methyl-CpG binding protein 2. Proc. Natl Acad. Sci. USA, 102, 17551–17558.[Abstract/Free Full Text]

  19. Miltenberger-Miltenyi, G. and Laccone, F. (2003) Mutations and polymorphisms in the human methyl CpG-binding protein MECP2. Hum. Mutat., 22, 107–115.[CrossRef][ISI][Medline]

  20. Kriaucionis, S. and Bird, A. (2004) The major form of MeCP2 has a novel N-terminus generated by alternative splicing. Nucleic Acids Res., 32, 1818–1823.[Abstract/Free Full Text]

  21. Coy, J.F., Sedlacek, Z., Bachner, D., Delius, H. and Poustka, A. (1999) A complex pattern of evolutionary conservation and alternative polyadenylation within the long 3'-untranslated region of the methyl- CpG-binding protein 2 gene (MeCP2) suggests a regulatory role in gene expression. Hum. Mol. Genet., 8, 1253–1262.[Abstract/Free Full Text]

  22. Carey, M. and Stephen, S.T. (1999) Transriptional Regulation in Eukaryotes. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

  23. Bejerano, G., Pheasant, M., Makunin, I., Stephen, S., Kent, W.J., Mattick, J.S. and Haussler, D. (2004) Ultraconserved elements in the human genome. Science, 304, 1321–1325.[Abstract/Free Full Text]

  24. Lettice, L.A., Heaney, S.J., Purdie, L.A., Li, L., de Beer, P., Oostra, B.A., Goode, D., Elgar, G., Hill, R.E. and de Graaff, E. (2003) A long-range Shh enhancer regulates expression in the developing limb and fin and is associated with preaxial polydactyly. Hum. Mol. Genet., 12, 1725–1735.[Abstract/Free Full Text]

  25. Woolfe, A., Goodson, M., Goode, D.K., Snell, P., McEwen, G.K., Vavouri, T., Smith, S.F., North, P., Callaway, H., Kelly, K. et al. (2005) Highly conserved non-coding sequences are associated with vertebrate development. PLoS Biol., 3, e7.[CrossRef][Medline]

  26. Trinklein, N.D., Aldred, S.J., Saldanha, A.J. and Myers, R.M. (2003) Identification and functional analysis of human transcriptional promoters. Genome Res., 13, 308–312.[Abstract/Free Full Text]

  27. Dekker, J. (2006) The three ‘C’ s of chromosome conformation capture: controls, controls, controls. Nat. Methods, 3, 17–21.[CrossRef][ISI][Medline]

  28. Adachi, M., Keefer, E.W. and Jones, F.S. (2005) A segment of the Mecp2 promoter Is sufficient to drive expression in neurons. Hum. Mol. Genet., 14, 3709–3722.[Abstract/Free Full Text]

  29. Fougerousse, F., Bullen, P., Herasse, M., Lindsay, S., Richard, I., Wilson, D., Suel, L., Durand, M., Robson, S., Abitbol, M. et al. (2000) Human-mouse differences in the embryonic expression patterns of developmental control genes and disease genes. Hum. Mol. Genet., 9, 165–173.[Abstract/Free Full Text]

  30. Howard, M.L. and Davidson, E.H. (2004) cis-Regulatory control circuits in development. Dev. Biol., 271, 109–118.[CrossRef][ISI][Medline]

  31. Boffelli, D., Nobrega, M.A. and Rubin, E.M. (2004) Comparative genomics at the vertebrate extremes. Nat. Rev. Genet., 5, 456–465.[CrossRef][ISI][Medline]

  32. Dickmeis, T. and Muller, F. (2005) The identification and functional characterisation of conserved regulatory elements in developmental genes. Brief Funct. Genomic Proteomic, 3, 332–350.[Abstract/Free Full Text]

  33. Koch, C. and Stratling, W.H. (2004) DNA binding of methyl-CpG-binding protein MeCP2 in human MCF7 cells. Biochemistry, 43, 5011–5021.[CrossRef][Medline]

  34. Kimura, H. and Shiota, K. (2003) Methyl-CpG-binding protein, MeCP2, is a target molecule for maintenance DNA methyltransferase, Dnmt1. J. Biol. Chem., 278, 4806–4812.[Abstract/Free Full Text]

  35. Osborne, C.S., Chakalova, L., Brown, K.E., Carter, D., Horton, A., Debrand, E., Goyenechea, B., Mitchell, J.A., Lopes, S., Reik, W. et al. (2004) Active genes dynamically colocalize to shared sites of ongoing transcription. Nat. Genet., 36, 1065–1071.[CrossRef][ISI][Medline]

  36. Kato, Y. and Sasaki, H. (2005) Imprinting and looping: epigenetic marks control interactions between regulatory elements. Bioessays, 27, 1–4.[CrossRef][ISI][Medline]

  37. Murrell, A., Heesons, S. and Reik, W. (2004) Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 in parent-specific chromatin loops. Nat. Genet., 36, 889–893.[CrossRef][ISI][Medline]