Human Molecular Genetics, 2005, Vol. 14, No. 6
doi:10.1093/hmg/ddi073
Advance Access published on February 2, 2005
785–797
Homologous pairing of 15q11– 13 imprinted
domains in brain is developmentally regulated
but deficient in Rett and autism samples
Karen N. Thatcher, Sailaja Peddada, Dag H. Yasui and Janine M. LaSalle*
Received December 17, 2004; Revised and Accepted January 24, 2005
Rett syndrome (RTT), caused by mutations in MECP2 (encoding methyl CpG binding protein 2), and
Angelman syndrome (AS), caused by maternal deficiency of chromosome 15q11 – 13, are autism-spectrum
neurodevelopmental disorders. MeCP2 is a transcriptional repressor of methylated genes, but MECP2
mutation does not directly affect the imprinted expression of genes within 15q11 – 13. We tested a potential
role for MeCP2 in the homologous pairing of imprinted 15q11 – 13 alleles in human brain tissue and differentiated neurons by fluorescence in situ hybridization (FISH). FISH analysis of control cerebral samples demonstrated a significant increase in homologous pairing specific to chromosome 15 from infant to juvenile brain
samples. Significant and specific deficiencies in the percentage of paired chromosome 15 alleles were
observed in RTT, AS and autism brain samples when compared with normal controls. SH-SY5Y neuroblastoma cells also showed a significant and specific increase in the percentage of chromosome 15q11 – 13
paired alleles following induced differentiation in vitro. Transfection with a methylated oligonucleotide
decoy specifically blocked binding of MeCP2 to the SNURF/SNRPN promoter within 15q11 – 13 and significantly lowered the percentage of paired 15q11 –13 alleles in SH-SY5Y cells. These combined results suggest
a role for MeCP2 in chromosome organization in the developing brain and provide a potential mechanistic
association between several related neurodevelopmental disorders.
INTRODUCTION
Autism is a complex genetic disorder involving multiple
chromosomal loci and environmental influences (1,2). Several
autism-spectrum neurodevelopmental disorders with known
genetic causes can be useful in understanding the multiple
genetic and epigenetic pathways involved in the etiology of
autism. Rett syndrome (RTT) is an X-linked neurodevelopmental disorder in females caused by mutation in MECP2
on Xq28 (3). Similar to autism, onset of symptoms are
delayed until 6 –18 months of age and include severe mental
retardation with absence of speech, stereotypic hand movements and epileptic seizures. Both autism and RTT fall
under the heading of pervasive developmental disorders
(PDD). Although mutations have been found in a few cases
of autism, several large studies have ruled out mutations in
the coding region of MECP2 as a significant genetic cause
of autism (4 – 8). Angelman syndrome (AS) is clinically very
similar to RTT (except for the lack of regression in skills)
and severe mental retardation, epilepsy and autistic features
combine with a happy disposition and inappropriate laughter
(9). AS is caused by maternal deficiency of chromosome
15q11 – 13 by maternal deletion, paternal disomy, maternal
UBE3A mutation or maternal methylation defects. Prader–
Willi syndrome (PWS) is a distinct neurodevelopmental
disorder caused by paternal deficiency of 15q11 –13. In one
study, 2% of AS patients had mutations in MECP2 (10),
and another recent study found 15q11 –13 rearrangements in
5% of RTT patients (11). Interestingly, 1 – 2% of autism
patients have been described with maternal duplications of
15q11 – 13 by interstitial duplications or marker chromosomes
(12). In addition, several recent studies have demonstrated
linkage of autism to polymorphisms within 15q11 – 13 near
UBE3A (13), GABRB3 (14) and GABRG3 (15). We have
recently demonstrated deficiencies in MeCP2 expression in
autism, AS and PWS postmortem brain samples arranged on
*To whom correspondence should be addressed at: Medical Microbiology and Immunology, One Shields Avenue, Davis, CA 95616, USA. Tel:
þ1 5307547598; Fax: þ1 5307528692; Email: [email protected]
Human Molecular Genetics, Vol. 14, No. 6 # Oxford University Press 2005; all rights reserved
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Medical Microbiology and Immunology and Rowe Program in Human Genetics, School of Medicine,
University of California, Davis, CA, USA
786
Human Molecular Genetics, 2005, Vol. 14, No. 6
RESULTS
Developmental changes in homologous pairing of
15q11 – 13 domains
In order to investigate the possibility that homologous association of imprinted 15q11 –13 domains may occur in brain
during normal development, we devised a fluorescence in
situ hybridization (FISH) protocol for detection of chromosome 15 in sections from human postmortem brain. A probe
specific to the pericentromeric region of chromosome 15
(CEP 15-SpectrumGreen, or D15Z1) was used as it is physically close to 15q11 –13 and previously showed significant
transient association in cycling lymphocytes (33). As a
control, another pericentromeric probe specific for chromosome 11 (CEP 11-SpectrumOrange) was used that did not
show homologous pairing in brain. In addition, a probe for a
single copy locus within 15q11–13 (GABRB3-SpectrumOrange,
5 Mb from CEP15) was chosen to directly examine 15q11–13
pairing. A control single copy probe was chosen from
another acrocentric chromosome (LSI 22-SpectrumGreen,
from BCR-ABL on chromosome 22, also 5 Mb from the pericentromeric heterochromatin), because acrocentric effects of
nuclear organization were expected.
In order to accurately compare FISH signals from multiple
human brain samples and control for slide-to-slide variability,
we used a tissue microarray approach (34,35). Sections from
a previously described tissue microarray containing triplicate
600 mm cores of frontal cortex samples from 28 different
controls and patients with neurodevelopmental disorders
(16) were hybridized with the FISH probe combinations
described earlier and then counterstained with DAPI (blue
fluorescence). Representative images are shown in Figure 1A
and B. Because tissue sections can result in incomplete
nuclei that may be missing FISH signals, three different possible FISH patterns were scored per nucleus and are shown to
the right of the graph in Figure 2. Nuclei with only one FISH
signal (‘one spot’) could be the result of either a missing
FISH signal due to sectioning or two overlapping FISH
signals due to pairing of alleles (Fig. 1B). In contrast,
‘paired’ nuclei were those scored as having two closely
spaced but discernable FISH spots per nucleus, whereas
‘unpaired’ nuclei showed two FISH signals per nucleus
.2 mm apart, based on the threshold set in a previous
study taking actual distance measurements (33). The percentage of nuclei (mean + SEM) for each of the three scoring
categories is graphed in Figure 2. A significant increase
was observed in the percentage of ‘one spot’ nuclei (white
bars) in juvenile/adult when compared with infant control
samples for CEP 15, but not CEP 11, suggesting a specific
change in the organization of chromosome 15 not likely
due simply to lost FISH spots. In addition, a specific and significant increase in ‘one spot’ nuclei was observed for the
15q11 – 13-specific GABRB3 probe but not for the control
LSI 22 probe in juvenile/adult when compared with infant
brain samples. A larger number of ‘paired’ nuclei were
observed for LSI 22 than for CEP 11 control probes, demonstrating an expected effect on organization of acrocentric
chromosomes around the nucleolus (36). A significant developmental-specific change in paired alleles was not observed
with LSI 22, however, suggesting that the developmental
increase in ‘one spot’ nuclei observed with GABRB3 was
not simply due to changes in acrocentric organization.
These combined results suggest a specific pairing of
15q11 – 13 alleles during postnatal brain development that
correlates with the timing of increased MeCP2 expression
(24,37).
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
a tissue microarray (16). Furthermore, UBE3A and GABRB3
expression defects were observed in RTT and autism brain
samples and Mecp2-deficient mouse brain (17). Together,
these observations suggest phenotypic and genetic overlap
among autism, RTT and AS.
MECP2 encodes methyl CpG-binding protein 2 that binds
to methylated CpG sites within nuclear heterochromatin
(18,19) and is predicted to be a transcriptional repressor of
methylated genes through its interaction with molecules such
as Sin3A, histone deacetlyase (HDAC) (20), DNA methyltransferase (DNMT1) (21) and histone methyltransferase
(22). As the 15q11 – 13 locus is subject to parental imprinting
and characterized by allele-specific methylation and transcription, the first hypothesis was that MeCP2 was essential for the
repression of the methylated imprinted genes within 15q11–13.
We have previously disproven this simple hypothesis,
however, by demonstrating that several imprinted genes within
15q11 –13 and 11p15 retained monoallelic expression in
MECP2-mutant lymphocyte clones and Rett brain samples
(23) as well as Mecp2-deficient mouse brain (17). MeCP2 is
most highly expressed in the nuclei of large mature neurons
within the CNS (24 – 26). Because changes in nuclear heterochromatin and chromosome positions accompany activation of
neurons (27), we now investigated whether MeCP2 may be
involved in the organization of chromosomes within neuronal
nuclei and thus have a more indirect effect on gene expression
within 15q11 –13.
Specific organization of homologous chromosomes has
previously been observed in three-dimensional reconstruction
studies of neuronal nuclei, with 9q12 and 1q12 showing
association around the nucleolus (28,29). In other studies on
human brain, chromosome 1 and 17 showed evidence for
somatic pairing (30,31). In addition, dynamic changes in the
position and clustering of centromeres in Purkinje neurons
occur during early postnatal development, associated with
the nucleolus (32). Homologous association of 15q11 –13
domains has been previously observed during late S-phase in
lymphocytes (33) but has not been previously examined
in neurons or brain tissue. In this report, we demonstrate
evidence for significant increased homologous pairing of
15q11 –13 domains during normal postnatal brain development in human brain. In addition, we demonstrate that brain
samples from several related neurodevelopmental disorders
show deficiencies in homologous pairing specifically for
15q11 –13. We further implicate MeCP2 in the mechanism
of homologous pairing by specifically blocking its binding
to endogenous chromatin and demonstrating a significant
reduction in homologous pairing of 15q11 – 13 domains in cultured neuroblastoma cells. These results open up many new
areas of investigation for understanding the roles of MeCP2
and 15q11 –13 during normal neuronal maturation and the
pathogenesis of several neurodevelopmental disorders.
Human Molecular Genetics, 2005, Vol. 14, No. 6
787
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Figure 1. Representative images of FISH signals obtained for these studies from human postmortem brain samples (A, B) or SH-SY5Y neuroblastoma cell lines
(C– F). (A) A section of cerebral cortex from a 1-month-old infant brain hybridized with probes specific for pericentromeric regions of chromosome 15 (CEP 15,
green) and chromosome 11 (CEP 11, red). (B) The same probe combination as in (A) hybridized to a 2.5-year-old juvenile cerebral section showing an increase
in the number of nuclei showing one green spot while still maintaining two red spots (yellow arrows). Nuclei were counterstained with DAPI (blue). Partial
nuclei without detectable FISH spots were not scored. (C) SH-SY5Y neuroblastoma cells prior to differentiation show growth in clusters without long
axonal projections. Hybridization with a CEP 15 (green) and a 15q11–13 specific probe (GABRB3, red) resulted in the majority of nuclei (blue) showing
two unpaired spots of each color. (D) Following PMA-induced differentiation for 3 days, SH-SY5Y cells show long axonal projections (white arrows) and
pairing of both CEP 15 (green) and GABRB3 (red) signals. Two different fields are shown because of the lower density of differentiated SH-SY5Y cells
in culture. (E) Undifferentiated and (F) 3 day PMA differentiated SH-SY5Y nuclei hybridized with control FISH probes CEP 11 (red) and LSI 22 (green)
do not show evidence for increased pairing of signals following differentiation. For images in (C –F), phase contrast images (grayscale) were overlayed on
three-color FISH images to demonstrate projections.
788
Human Molecular Genetics, 2005, Vol. 14, No. 6
Defects in homologous pairing of 15q11 –13 observed in
several neurodevelopmental disorders
The tissue microarray used for the analysis of multiple brain
samples in Figure 2 also contained samples of age- and
region-matched cerebrum from patients with RTT (n ¼ 6),
autism (n ¼ 5), AS (n ¼ 1), PWS (n ¼ 2) and PDD (n ¼ 1).
The nuclei within each sample were scored in an identical
fashion to the control samples described previously and the
results are graphed in Figure 3. In order to test the significance
of changes in the FISH patterns, samples with similar diagnoses were grouped and compared with normal controls
(mean 18.1 years). AS and PWS samples were grouped
together because of the paucity of samples in these categories
and the expected loss of homologous pairing for both disorders (33). The most significant changes in FISH patterns
from control samples were observed using the GABRB3
15q11 –13-specific probe, as both RTT and autism samples
showed a significant increase in the percentage of ‘unpaired’
nuclei and a corresponding decrease in the percentage of
‘one spot’ nuclei. The PWS 865 and AS 293 samples
graphed for GABRB3 are monosomic for 15q11– 13 because
of deletions and therefore were expected to show primarily
‘one spot’ nuclei. A PWS uniparental disomy sample (PWS
1290) was also scored and showed defects in GABRB3
pairing (Supplementary Material, Table S1). In contrast,
control acrocentric probe LSI 22 showed no significant
changes in the percentage of ‘one spot’ or ‘paired’ nuclei,
suggesting a specific loss of 15q11 –13 pairing in these neurodevelopmental disorder samples. Furthermore, the results
obtained with pericentromeric probe CEP 15 showed significant increases in the percentage of ‘unpaired’ alleles in
RTT, autism and PWS/AS samples, whereas those with the
control CEP 11 probe showed no significant differences.
These results suggest defects in homologous pairing of
15q11 – 13 alleles in several related neurodevelopmental
disorders.
Although the result of identifying defects in 15q11 – 13
organization in patient samples with different genetic
mutations was somewhat unexpected, our previous investigation of MeCP2 expression levels on the same tissue microarray identified multiple abnormalities in MeCP2/MECP2
expression in all of the neurodevelopmental disorder
samples (16). We therefore hypothesized that MeCP2 may
play an important or necessary role in the organization of
15q11 – 13 alleles.
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Figure 2. Specific developmental change in the number of chromosome 15 FISH signals per nucleus in control human brain tissue. A tissue microarray containing frontal cortex samples from 28 different controls and patients with neurodevelopmental disorders has been previously described (16). Replicate section of the
tissue microarray were each hybridized with either CEP 15 (green) and CEP 11 (red) or LSI 22 (green) and GABRB3 (red) and counterstained with DAPI (blue).
Signals of each probe were scored in each nucleus as one of the three patterns shown at right: ‘one spot’ nuclei showed only one spot, ‘paired’ nuclei showed two
closely spaced but discernable spots and ‘unpaired’ nuclei showed two distinct spots. Control samples were categorized as infant (1–56 days) or juvenile to adult
(2.5–36 years), and the total number of individuals and nuclei scored are shown below the graph. The percentage of nuclei in each of the three categories (mean
+ SEM) of different samples and replicate hybridizations are graphed. A significant increase was observed in the percentage of ‘one spot’ nuclei (white bars)
following infancy in control samples for CEP 15 and GABRB3, but not CEP 11 or LSI 22. A corresponding significant decrease in the percentage of ‘unpaired’
nuclei (dark gray bars) was observed for both chromosome 15 specific probes, but not control probes. P , 0.0005 by t-test. Data for scoring of individual
brain samples is shown in Supplementary Material, Table S1 and Figure S1. Five additional infant brain samples are also included in Supplementary Material,
Figure S1.
Human Molecular Genetics, 2005, Vol. 14, No. 6
789
Increased homologous pairing of chromosome
15 following induced differentiation of SH-SY5Y
neuroblastoma cells
Because the problem of incomplete nuclei was inherent to use
of brain tissue, we sought an independent cell culture system
to investigate the changes in the organization of 15q11 –13
domains during neuronal differentiation and to directly test
the necessity of MeCP2 in the process. SH-SY5Y neuroblastoma cells were selected because they can be induced to
undergo differentiation within 3 days using PMA, resulting
in a morphologic change in the extension of axonal projections
and the increased expression of neuron specific enolase (38). In
addition, SH-SY5Y cells are diploid for most chromosomes,
including those sampled by our FISH probes (39). Nuclei
showing more than two spots because of replication or
aneuploidy were infrequent and excluded from scoring.
Representative images are shown in Figure 1D–F, with
projections indicated by arrows. MeCP2 expression is
significantly increased 24–72 h following PMA treatment
(40), making it a good model for the developmental
maturation stage characteristic of neurons expressing high
levels of MeCP2 in the developing brain (37).
SH-SY5Y neuroblastoma cells were cultured on glass
slides, fixed either before (untreated) or 72 h following differentiation with PMA (PMA treated) and hybridized with the
same FISH probes described previously for brain tissue (see
representative images in Fig. 1C –F). Because fixation of
cells results in whole nuclei, FISH patterns were scored as
simply paired (one spot or two spots ,2 mm apart) or
unpaired. The results, graphed in Figure 4, demonstrate a significant increase in the percentage of paired alleles following
PMA-induced differentiation for nuclei hybridized with CEP
15 and GABRB3 probes, but not control CEP 11 or LSI 22
probes. These results provide additional support to the conclusion that chromosome 15q11–13 alleles show increased
pairing during neuronal differentiation. As MeCP2 expression
is upregulated by 48 h following differentiation in SH-SY5Y
cells (40) (data not shown), the developmental regulation
would coincide with that of the 15q11–13 homologous pairing.
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Figure 3. Significant differences in the number of chromosome 15 specific FISH signals in brain samples from patients with neurodevelopmental disorders. Brain
samples from patients with neurodevelopmental disorders were grouped according to diagnosis and results of the scoring of FISH signals described in Figure 2
were compared with juvenile and adult control samples for significance. The total number of nuclei scored and number of brain samples in each category are
shown below the graph. The percentage of nuclei in each of the three categories (mean + SEM) of different samples and replicate hybridizations is shown. Only
samples with .90% hybridization efficiency were scored, resulting in slight differences among the total number of samples scored for each probe. Significant
increases in the percentage of ‘unpaired’ nuclei were observed for in all three categories of neurodevelopmental disorders (RTT, autism and PWS/AS) for CEP
15 and GABRB3 (with the exception of PWS/AS) when compared with control samples. Both PWS and AS samples graphed for GABRB3 are monosomic for
15q11– 13 and, therefore, served as a control for non-specific hybridization signals (,5%). In contrast to the differences in FISH spot signals observed for
chromosome 15 specific probes, no significant differences were observed in the FISH patterns using either of the control probes CEP 11 or LSI 22. The
average nuclear area (73.8 + 2.1 mm2) was not significantly different between different sample categories on the tissue microarray as determined from laser
scanning cytometry data collected in previous analyses (17). P , 0.05, P , 0.005 and P , 0.0005 by t-test. Data for scoring of individual brain
samples is shown in Supplementary Material, Table S1.
790
Human Molecular Genetics, 2005, Vol. 14, No. 6
Transfected methylated decoys for MeCP2 partially block
neuronal differentiation and homologous pairing
To directly test the role of MeCP2 in the nuclear organization
of 15q11– 13 alleles, we sought to temporarily disrupt the
function of MeCP2 in the SH-SY5Y system. We chose to
use an oligonucleotide decoy approach because of the
success of these systems for blocking targets of transcription
factors (41). Two different double-stranded phosphorothioate
oligodeoxynucleotides were obtained from GeneDetect.com:
the wild-type MeCP2 binding sequence (MDWT) contained
two CpG methylation sites, whereas the mutant sequence
(MDMT) was identical except for the replacement of CG
with AT and therefore served as a specificity and transfection
control. FITC-labeled oligonucleotides showed a high transfection efficiency by quantitative laser scanning cytometry
and stability within SH-SY5Y cells for 72 h (data not
shown). For the results graphed in Figure 5, SH-SY5Y cells
were transfected with MDWT or MDMT 12 h prior to PMA
addition, then fixed 72 h later. The analysis of FISH signals
from three experimental replicates demonstrated a significant
difference between MDWT and MDMT transfected cells in
the percentage of nuclei showing paired alleles of GABRB3,
but not CEP 15, CEP 11 or LSI 22. These results suggest
that MeCP2 has a specific effect on the organization of
15q11 –13 alleles but is not involved in the nucleolar organization of CEP 15.
In order to confirm that the MDWT transfection specifically
blocked the binding of MeCP2 to endogenous methylated CpG
Figure 5. Transfection of methylated MeCP2 decoy into SH-SY5Y cells prior
to differentiation significantly reduces the homologous pairing of GABRB3
alleles. SH-SY5Y neuroblastoma cells were cultured on glass slide chambers
and transfected with either MDWT (containing two methylated CpGs) or
MDMT (without CpGs) oligonucleotide decoys. Twelve hours following
transfection, PMA was added to induce differentiation and fixed after 72 h
in culture. The percentage of paired nuclei were scored and the results
shown are the mean +SEM of nine replicate scorings of seven replicate
hybridizations from four replicate transfections, each with 100– 300 nuclei
scored. A significant decrease in the percentage of paired alleles of
GABRB3 was observed for cells transfected with the methylated MeCP2
decoy (MDWT transfected, light gray bars) when compared with the control
(MDMT transfected, dark gray bars). In contrast, no significant difference
was observed using CEP 15 or either of the control probes. P , 0.05, by
t-test.
targets within 15q11 –13, we performed chromatin immunoprecipitation (ChIP) on SH-SY5Y cells transfected with
MDWT or MDMT and differentiated for 48 h as well as
untreated and PMA treated but untransfected controls. Two
different antibodies reactive to the C-terminal epitope of
MeCP2 were used to IP endogenous fragments bound to
MeCP2 and the resulting DNA samples were assayed by
PCR. Representative results are shown in Figure 6A for the
methylated SNURF/SNRPN promoter within the 15q11 –13
imprinting control region (42) and an expressed and unmethylated housekeeping control gene (GAPDH ). The untransfected
controls showed a high recovery of the SNURF/SNRPN promoter region from total input DNA by anti-MeCP2 ChIP
with both antibodies. An increase in MeCP2 binding following
differentiation was also apparent, concordant with the increase
in MeCP2 expression by 48 h (40). In contrast, differentiated
SH-SY5Y cells transfected with MDWT showed a 10-fold
reduction of SNRPN/SNURF containing fragments precipitated with anti-MeCP2 when compared with MDMT transfected controls. As expected, GAPDH sequences were not
recovered at a detectable level following ChIP with antiMeCP2. To confirm the reproducibility of the ChIP results,
semi-quantitative PCR results from four separate experiments
graphed in Figure 6B show a significant decrease in the
amount of SNURF/SNRPN sequences (as a ratio of input
DNA) immunoprecipitated from differentiated SH-SY5Y
cells transfected with the MDWT decoy when compared
with the MDMT control (P , 0.005). To determine whether
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Figure 4. SH-SY5Y neuroblastoma cells show an increase in the percentage of
15q11–13 paired alleles following induced differentiation. SH-SY5Y neuroblastoma cells were cultured on glass slide chambers and fixed before
(untreated) or 72 h following differentiation with 16 nM PMA (PMA
treated), then hybridized with either CEP 15 (green) and GABRB3 (red) or
CEP 11 (red) and LSI 22 (green). Representative images are shown in
Figure 1C –F. Nuclei were scored as either ‘unpaired’ or ‘paired’ for simplicity, but one experiment scored for ‘one spots’ is shown in Supplementary
Material, Figure S2. The percentage of ‘paired’ nuclei is graphed for untreated
(white bars) and PMA treated (gray bars) as the mean + SEM of three to four
replicate experiments, each with 100–500 nuclei scored. A significant increase
in the percentage of nuclei showing pairing of CEP 15 and GABRB3 was
observed following SH-SY5Y differentiation. In contrast, no significant differences were observed using control probes CEP 11 or LSI 22. P , 0.05, by
t-test.
Human Molecular Genetics, 2005, Vol. 14, No. 6
791
the effect of MDWT transfection was specific to MeCP2
binding, antibodies to all other known methyl binding
domain (MBD)-containing proteins (MBD1, MBD2, MBD3
and MBD4) were used on the same chromatin preparations
in the ChIP assay. The results of two separate experiments
with each antibody are graphed in Figure 6B. MBD1 and
MBD2 showed an increase in binding to the SNURF/SNRPN
promoter following differentiation and a slight non-specific
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
Figure 6. ChIP analysis of MeCP2 bound to the SNURF/SNRPN promoter demonstrates that the methylated decoy specifically blocked binding of MeCP2 to an
endogenous site within 15q11–13. (A) Representative gel of PCR products using primers specific for the SNURF/SNRPN promoter following ChIP with two
different C-terminal MeCP2 reactive antibodies (Aves chicken or Upstate rabbit). The total DNA isolated from chromatin prior to IP (Input) was used as a positive control. The PCR fragment was highly represented in differentiated SH-SY5Y cells or cells transfected with control decoy (MDMT) but not in chromatin
isolated from cells transfected with methylated MeCP2 decoy (MDWT). In contrast, the GAPDH promoter was not detectable following ChIP with anti-MeCP2.
Additional PCR experiments performed with or without predigestion with a methyl-sensitive enzyme demonstrated that MeCP2 was bound to the methylated
maternal allele, as expected (Supplementary Material, Fig. S3). (B) The mean + SEM of four separate ChIP experiments using anti-MeCP2 (both antibodies) are
graphed as a ratio of the immunoprecipitated to the ‘input’ band intensities. For ChIP performed with anti-MeCP2, the SNURF/SNRPN promoter sequence was
significantly lower (P , 0.005) in MDWT transfected cells (light gray bars) when compared with MDMT transfected cells (hatched bars), demonstrating that
binding of MeCP2 to an endogenous methylated site within 15q11–13 was specifically blocked by the MDWT decoy. Other MBD proteins (mean + SEM of
two replicates each) did not show significant differences between MDWT and MDMT transfected cells in the recovery of the SNURF/SNRPN promoter or were
undetectable (no bar) by ChIP.
792
Human Molecular Genetics, 2005, Vol. 14, No. 6
blocking effect with transfection of both MDWT and MDMT,
suggesting a non-methyl-specific effect of the decoy transfection. MBD3 and MBD4 showed a decrease in binding to the
SNURF/SNRPN promoter following differentiation and therefore no significant changes due to decoy transfection. These
results demonstrate that transfection with the methylated
decoy resulted in specific decreased binding of MeCP2 to an
endogenous methylated target within the imprinting control
region of 15q11 –13 that could potentially explain the
reduced homologous pairing of this region.
Despite the identification of several candidate genes for neurodevelopmental disorders, the precise role of the gene products
in the development or differentiation of neurons remains
uncharacterized. MeCP2 has been predicted to be part of an
epigenetic pathway of gene expression during neuronal maturation (1), but evidence is lacking for a direct role of MeCP2
in silencing imprinted genes within 15q11 – 13. Maternal
15q11 –13 deficiency causes AS, whereas maternal duplication of 15q11 – 13 is observed in autism, implying that
correct parental chromosome dosage is important for normal
brain development. Recent reports of trans effects of
mutations in imprinted regions (43 –45) prompted us to
examine the possibility of homologous interactions of
15q11 –13 domains in human brain. In this study, a novel
investigation of the organization of homologous 15q11 –13
alleles has revealed developmental changes in normal brain
development and defects in RTT, AS/PWS and autism. In
addition, the novel implication of MeCP2 in the process of
homologous 15q11 – 13 pairing provides an important molecular link between these different neurodevelopmental disorders.
The organization of chromosomes in interphase nuclei has
been predicted to be a potentially important mechanism of
regulating gene expression during cellular differentiation,
especially in neurons (30,46 – 48). Homologous pairing has
been previously observed in human brain for pericentromeric
regions of chromosomes 1, 8 and 17 (30,31,48). Homologous
associations of human 15q11 – 13 domains were previously
observed in lymphocytes but restricted to the late S-phase of
the cell cycle and cells with a biparental contribution of
15q11 –13 (33). 15q11– 13 pairing was also observed at a
high frequency in cycling fibroblast and NT2 neuronal cultures
(49). Homologous pairing of the syntenic region on distal
murine chromosome 7 in mouse fibroblasts was also observed
exclusively in S-phase (50). Our finding of homologous
pairing of 15q11 –13 domains in postnatal human brain
samples containing mostly nuclei in G0 suggests that the
association of certain homologous regions can occur both
during the cell cycle and during the cellular differentiation.
As postmitotic neurons undergo substantial changes in
synapse formation and dendritic branching during postnatal
development, changes in nuclear organization may reflect
dynamic transcriptional patterns.
In addition to the developmental changes in organization of
15q11 –13 homologs in human brain, we demonstrate significant defects in homologous pairing in several related neurodevelopmental disorders. The defects in CEP 15 pairing of
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
DISCUSSION
PWS and AS patients with paternal or maternal deletions of
15q11 – 13 were expected as this result had been observed
previously in lymphocytes (33). Maternal duplications of
15q11 – 13 in 1– 2% of autism patients (12) also suggest that
mismatching of 15q11 –13 homologous pairing may cause
an autistic phenotype. The novel demonstration of defects in
homologous pairing in patient brain samples without detectable cytogenetic abnormalities in 15q11 –13, however,
suggests that RTT and autism may share defects in an overlapping pathway that regulates homologous pairing of 15q11 – 13
regions in brain. We have recently demonstrated defects in the
level of MeCP2 expression in the same neurodevelopmental
samples from the same tissue microarray (16). Two of the
RTT patients have truncation mutations in MECP2, but the
remaining two RTT and all five autism samples have no
detectable mutation in MECP2 (including 1 kb of promoter
and exon 1), but significant differences in MeCP2 protein
expression (16). Interestingly, two samples (PDD 144 and
AUT 732) shown to have increased MeCP2 expression compared with controls also showed reduced 15q11– 13 pairing
(Supplementary Material, Table S1). As higher MeCP2
expression in transgenic mice also causes a severe neurologic
phenotype (51,52), perhaps a precise level of MeCP2 binding
is required for nuclear organization of 15q11 – 13 in brain. On
the basis of these combined results, we hypothesized that
MeCP2 may be directly involved in the homologous pairing
of 15q11 – 13 domains in the postnatal brain.
The analysis of brain samples has several potential limitations for the analysis of homologous pairing by FISH. First,
as FISH is performed on sectioned tissue, not all nuclei are
complete and ‘one spots’ may represent absent signals in
addition to paired alleles and the developmental changes
could simply be due to increased nuclear area. The lack of a
significant increase during development in ‘one spots’ of
control probes CEP 11 and LSI 22, however, argues against
this trivial explanation and suggests a specific homologous
pairing of chromosome 15 during postnatal brain development. A second potential problem could be the recently
reported aneuploidy of normal postnatal neurons (53). Our
results are not consistent with monosomy of chromosome 15
being an explanation for our results because there was an
increase in the percentage of both ‘paired’ and ‘one spot’
nuclei for CEP 15 during development. The use of an in
vitro system for inducing neuronal maturational differentiation
and elevated MeCP2 expression was essential for confirming
that the changes in homologous pairing of chromosome 15
could be experimentally induced. Our results demonstrate
that significant increases in the percentage of nuclei showing
chromosome 15 paired alleles is observed within 72 h following differentiation of SH-SY5Y neuroblastoma cells. The
GABRB3 probe signals were closer than the CEP 15 or LSI
22 probe signals, as evident from the increased number of
‘one spots’ in these intact nuclei (Supplementary Material,
Fig. S2). The FISH experiments, however, do not provide evidence for an actual physical contact between alleles but
instead indicate non-random nuclear organization that could
come from two alleles sharing a ‘transcription factory’ (54)
or ‘chromatin hub’ (55).
Although MeCP2 acts as a transcriptional repressor of
methylated gene constructs (20,56), a paucity of methylated
Human Molecular Genetics, 2005, Vol. 14, No. 6
upstream of the MeCP2 binding SNRPN promoter containing
the recently duplicated maternal imprinting conrol region
(72) that is not conserved in mouse chromosome 7B4 but is
highly conserved in the chimpanzee (UCSC human genome
browser, Chimpanzee Genome Sequence Consortium) (73)
could explain the discrepancy between species. Interestingly,
Mecp2 deficiency in mouse results in reduced expression of
both UBE3a/Ube3a and GABRB3/Gabrb3, although less
significantly than that observed in human RTT or autism
brain (17). Perhaps MeCP2 can act in both cis as a longrange regulator of chromatin and trans when 15q11-13
homologs are close. This possibility is supported by the
formation of oligomeric chromatin suprastructures by
MeCP2 in vitro (65) but remains to be directly investigated.
The observation that MeCP2 decoy specifically blocked the
association of GABRB3 but not CEP15 would also support a
specific role for MeCP2 in 15q11–13 rather than nucleolar
organization.
Although the mechanism of homologous pairing of
imprinted 15q11 –13 domains has not been fully characterized, at least one part of the pathway must involve an allelediscrimination step. Allele-specific methylation patterns are
found throughout 15q11 –13 (42,74,75), with the most stable
methylated sites at the 50 end of the maternal SNURF/
SNRPN within the imprinting control region (76). Our ChIP
results demonstrate that MeCP2 binds to this methylated
CpG island and binding is increased following SH-SY5Y
differentiation. The reduced 15q11 –13 pairing in brain
samples from patients with RTT and autism may therefore
be due to the defects in MeCP2 expression in these samples.
As MeCP2 has been shown to be involved in silent chromatin
looping for both Igf2/H19 (77) and Dlx5/Dlx6 (61) imprinted
domains, perhaps loss of MeCP2 binding to the imprinting
control region and other differentially methylated regions of
15q11 – 13 results in abnormal loop structures and nuclear
mislocation of both alleles of UBE3A and GABRB3. We
have recently described expression defects of UBE3A and
GABRB3 in autism and RTT brain samples, consistent with
this model (17). Although the possibility remains that the
15q11 – 13 nuclear organization changes observed here may
be unrelated to the expression changes of UBE3A and
GABRB3 or to the disease etiology, the model of chromatin
looping and transcriptional activity is a testable one. Clearly,
much additional work lies ahead in understanding the mechanism of homologous pairing of chromosomes in postnatal brain
as well as the downstream effects on gene expression within
the 15q11– 13 region. Our results suggest, however, that
the pathogenic mechanisms of overlapping human autismspectrum disorders with different underlying genetic causes
may intersect at the developmentally controlled organization
of oppositely imprinted 15q11 –13 domains in the postnatal
brain.
MATERIALS AND METHODS
Tissue culture
SH-SY5Y neuroblastoma cells (ATCC) were grown in complete minimal essential media with 15% fetal calf serum.
Cells were seeded onto two-chamber glass slides treated
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
genes showing increased expression in mutant brain has
been identified by genome-wide expression profiling (57,58).
Brain derived neurotrophic factor (BDNF ), whose activitydependent transcriptional activation is regulated by methylation (59), shows significantly increased basal transcription
in Mecp2-null cultured neurons (60). Recently, two genes
within an imprinted domain (Dlx5 and Dlx6 ) have demonstrated to exhibit increased transcription in Mecp2-deficient
brain (61). MeCP2 has been predicted to have additional
roles, including HDAC-independent transcriptional repression
(62), association with WW-domain splicing factors (63),
matrix attachment activity (64), chromatin compaction
activity (65) and silent chromatin looping (61). In this
report, we demonstrate another potential role for MeCP2 in
long-range interactions of an imprinted chromosomal region
essential for normal brain development. By blocking the
binding of MeCP2 by a methylated decoy approach, we
demonstrate a significant defect in homologous pairing when
compared with the transfection control, suggesting that
MeCP2 is involved in the pathway of 15q11 –13 allele
pairing. ChIP assays confirmed that binding of MeCP2 to
the SNURF/SNRPN promoter within 15q11 –13 was significantly reduced by the methylated decoy.
Recently, a second isoform of MeCP2 has been described
that arises from alternative splicing of exon 2 and results in
a change in the N-terminus (66,67). Because our ChIP experiments utilized C-terminal reactive antibodies for MeCP2,
both MeCP2 isoforms were most likely precipitated by
ChIP. An unexpected non-specific effect of increasing the
homologous pairing was observed following transfection of
either MDWT or MDMT decoys using the single copy
probes to acrocentric chromosomes (GABRB3 and LSI 22).
This result could be because of the subtle effects of the
decoys or transfection on the binding of MBD1 or MBD2 to
endogenous CpG sites (Fig. 6B), as MBD2 has been shown
to repress rRNA transcription and could influence the nucleolar organization (68). These effects were neither methylationspecific by ChIP nor 15q11 – 13-specific by FISH, suggesting
that MDMT was essential for controlling the non-specific
effects of the transfection and decoy approach. Although our
results strongly implicate MeCP2 in the process of homologous pairing of 15q11 – 13 domains, we cannot exclude the
possibility of additional methylation-specific or non-specific
effects of the decoy transfection explaining our results. The
relatively subtle effect of the MeCP2 decoy on 15q11 –13
pairing in SH-SY5Y cells suggests that factors in addition to
MeCP2 may be important in the interaction.
Mecp2-deficient mouse models of RTT recapitulate the disorder but with a milder phenotype, as hemizygous male mice
are more similar to heterozygous female RTT patients in onset
and severity (69,70). Interestingly but unfortunately, no evidence for homologous pairing of the 15q11 – 13 syntenic
region in mouse (7qB4) was observed in either wild-type or
Mecp2-deficient mouse brain at any developmental stage
(Supplementary Material, Fig. S4 and Table S2). The most
likely explanation for the discrepancy is that mouse 7qB4 is
not adjacent to ribosomal DNA (rDNA) genes as it is for acrocentric chromosome 15 in human, as the placement of the
15q11 –13 domain close to rDNA genes occurred during
primate evolution (71). Alternately, the 1 Mb region
793
794
Human Molecular Genetics, 2005, Vol. 14, No. 6
with poly-D -lysine and grown on slides until 30 – 50% confluent. Cells were fixed either before (untreated) or 72 h after the
addition of 16 nM PMA (PMA treated) for 15 min in Histochoice (Ameresco) then washed in 1 PBS/0.5% Tween for
5 min and stored in 70% ethanol at 2208C.
all of fluorescence within the section. Scoring of FISH
signals was perfomed manually and results are averages of
scoring performed both blinded (in which the individual
scoring did not know the identity of the samples) and
unblinded, as no evidence for bias was observed (Supplementary Material, Table S1).
Brain tissue microarray
FISH
Slides were dehydrated in 70, 90 and 100% ethanol (10 min
each) and then dried at 508C. A probe mixture containing
1 ml each probe (Vysis, Inc.), 2 ml ddH2O and 7 ml LSI/
WCP buffer (Vysis, Inc.) was warmed to 378C, then added
to the slide, coverslipped and sealed with rubber cement.
Probe and cells were simultaneously denatured at 808C for
1.5 min (for SH-SY5Y cells) or 858C for 2 min (for tissue
micoarray slides) on a slide cycler (Hybaid). Slides were incubated overnight at 378C, then washed in 50% formamide/
50% 2 SSC thrice for 5 min, 0.5 SSC for 5 min and
0.5 SSC/0.1% IGEPAL for 5 min, all at 468C and pH 7.6.
To the slides 250 mg/ml RNase was added, coverslipped and
incubated at 378C for 30 min, then 5 min in 1 PBS and air
dried. Slides were mounted with 5 mg/ml DAPI in Vectrashield (Vector Laboratories), coverslipped and sealed with
nailpolish. Mouse BAC clones for Gabrb3 (RP23-24D4),
Snrpn (RP24-275J20) and Ptgs1 (RP23-274M8) were
labeled with biotin or digoxigenin by nick translation and
detected as described previously (33).
Fluorescence microscopy
Slides were analyzed on an Axioplan 2 fluorescence microscope (Carl Zeiss, Inc., NY, USA) equipped with a Sensys
CCD camera (Photometrics, Tucson, AZ, USA), appropriate
fluorescent filter sets, and automated xyz stage controls. The
microscope and peripherals were controlled by a Macintosh
running IPLab Spectrum (Scanalytics, Vienna, VA, USA)
software with Multiprobe, Zeissmover and 3D extensions.
Images were captured for blue, green and red filters at one
edge of the specimen, then repeated at 0.4 mm sections
through the depth of the tissue. Each image stack was digitally
deconvolved to remove out-of-focus light using HazeBuster
software (Vaytek, Fairfield, IA, USA). Following haze
removal, image stacks for each fluorophore were merged
and stacked to create a two-dimensional image representing
MeCP2 decoy transfections
MeCP2 decoy and control decoy were obtained commercially
(GeneDetect.com). Both mutant (50 -TAATCTAGTCTAGAC
TAGATTA-30 ) and wild-type (50 -TAATCCGGTCTAGACC
GGATTA-30 ) double-stranded phosphorothioate oligodeoxynucleotides were treated with HpaII methylase overnight to
methylate the CpG sites. The methylase-treated decoys,
MDWT (MeCP2 decoy wild-type) and MDMT (MeCP2
decoy mutant control), were digested with HpaII and analyzed
by PAGE to confirm methylation.
SH-SY5Y cells were grown on two-chamber glass slides
treated with poly-D -lysine and transfected with decoy
mixture (100 ml per chamber): 92 ml serum free media, 3 ml
Fugene 6 (Roche) and 5 ml methylated decoy (containing
1 mM of either MDWT or MDMT) which was incubated at
room temperature for 30 min before addition to slides.
Twelve hours after transfection, cells were treated with
16 nM PMA and fixed 72 h later, as described previously.
Chromatin immunoprecipitation
Chromatin was prepared from SH-SY5Y cells and purified by
urea gradient centrifugation as described previously (79,80).
Immunoprecipitation, reverse crosslinking and PCR amplification were performed as described previously (81) with some
modifications. For each experiment, 150– 200 mg of chromatin
was digested into 5 kb fragments with Sac1 (New England
Biolab) and precleared first by incubation with appropriate
agarose beads (PrecipHen agarose, Aves labs or protein A/G
agarose, Pierce) alone, then with appropriate preimmune
serum (preabsorbed IgY, rabbit IgG, mouse IgG) followed
by agarose beads. Precleared chromatin was divided (30 mg
per tube) and incubated overnight with 5 mg of either
C-terminal anti-MeCP2 (raised in chicken to C-terminal
peptide N-RPNREEPVDSRTPVTERVS-C, Aves Labs) or
preabsorbed IgY as a control for non-specific binding;
C-terminal anti-MeCP2 (rabbit commercial, Upstate); antiMBD1 (Affinity bioreagents), anti-MBD4 (Imgenex), or
rabbit IgG control; anti-MBD2 and anti-MBD3 (Imgenex) or
mouse IgG control. Antibody incubations were followed by
additional incubation for 4– 6 h with 40 ml of agarose beads.
Equal amounts of precleared chromatin were processed
without IP as total input control. Immunoprecipitates collected
by centrifugation were washed, then digested with 50 mg/ml
DNase free RNaseA for 30 min at 378C, followed by SDS/
proteinase K digestion and subjected to phenol/chloroform
extraction before ethanol precipitation with glycogen. Onetwentieth of the DNA from each IP reaction was PCR amplified
in reactions containing 2.5 U of TaKaRa LA Taq (TaKaRa),
1 GC buffer I or II, dNTP mix (2.5 mM each) and
0.2 mM primers of either Pr 291 and Pr 292 (50 -actgccatagcc
tcctcgcctc-30 and 50 -cttgctgttgtgccgttctgcc-30 ) specific to the
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
The paraffin-embedded tissue microarray described previously
(16) containing triplicate frontal cortex (Brodman area 9,
layers III –V) was sectioned at 5 mm onto glass slides.
Slides were baked overnight at 558C, then placed in four
5 min washes with xylene, then two 5 min washes with
100% ethanol and then 1 h at 958C in antigen retrieval solution (DAKO). Slides were then post-fixed in Histochoice
for 90 min and then washed 5 min in 1 PBS. A mouse
brain tissue microarray was constructed containing multiple
Mecp2 þ/y, 2/y, þ/þ, 2/þ cerebrum samples (developmental
ages shown in Supplementary Material, Table S2) obtained
by mating Mecp2 tm.1.Bird/þ females (Jackson Labs) to
C57BL/6 wild-type males as described previously (78).
Human Molecular Genetics, 2005, Vol. 14, No. 6
SNURF/SNRPN promoter within the 15q11–13 imprinting control region or Pr 279 and Pr 280 (50 -ccaatctcagtcccttccccc-30 and
50 -gtttctctccgcccgtcttc-30 ) specific to the GAPDH promoter
region using one cycle of 958C for 5 min, 30– 35 cycles of
958C for 30 s, 608C for 30 s, 728C for 30 s, with a final
cycle of 728C for 7 min. PCR products were resolved by
agarose gel electrophoresis, stained with Sybr Gold (molecular
probes) and intensities of the PCR bands were quantified using
GelExpert software (Nucleotech).
Supplementary Material is available at HMG Online.
ACKNOWLEDGEMENTS
The authors thank K. Ehmsen, D. Braunschweig and
R. Samaco for technical assistance and M. Lalande for critical
reading of the manuscript. This work was supported in part by
the U.C. Davis MIND Institute, the Rett Syndrome Research
Foundation and the NIH (1R01HD/NS41462). Human tissue
samples were generously provided by the Autism Tissue
Program, the University of Maryland Brain and Tissue Bank
for Developmental Disorders (supported by NIH N01-HD-13138), Harvard Brain Tissue Resource Center (supported in
part by PHS MH/NS 31862) and M. Lalande.
REFERENCES
1. Zoghbi, H.Y. (2003) Postnatal neurodevelopmental disorders: meeting at
the synapse? Science, 302, 826–830.
2. Volkmar, F.R. and Pauls, D. (2003) Autism. Lancet, 362, 1133–1141.
3. 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.
4. Lam, C., Yeung, W., Ko, C., Poon, P., Tong, S., Chan, K., Lo, I., Chan, L.,
Hui, J., Wong, V. et al. (2000) Spectrum of mutations in the MECP2 gene
in patients with infantile autism and Rett syndrome. J. Med. Genet., 37,
E41.
5. Beyer, K.S., Blasi, F., Bacchelli, E., Klauck, S.M., Maestrini, E.,
Poustka, A. and International Molecular Genetic Study Of Autism
Consortium. (2002) Mutation analysis of the coding sequence of the
MECP2 gene in infantile autism. Hum. Genet., 111, 305–309.
6. Lobo-Menendez, F., Sossey-Alaoui, K., Bell, J.M., Copeland-Yates, S.A.,
Plank, S.M., Sanford, S.O., Skinner, C., Simensen, R.J., Schroer, R.J. and
Michaelis, R.C. (2003) Absence of MeCP2 mutations in patients from the
South Carolina autism project. Am. J. Med. Genet., 117B, 97–101.
7. Carney, R.M., Wolpert, C.M., Ravan, S.A., Shahbazian, M., AshleyKoch, A., Cuccaro, M.L., Vance, J.M. and Pericak-Vance, M.A. (2003)
Identification of MeCP2 mutations in a series of females with autistic
disorder. Pediatr. Neurol., 28, 205–211.
8. Zappella, M., Meloni, I., Longo, I., Canitano, R., Hayek, G.,
Rosaia, L., Mari, F. and Renieri, A. (2003) Study of MECP2 gene in
Rett syndrome variants and autistic girls. Am. J. Med. Genet., 119B,
102 –107.
9. Lalande, M. (1996) Parental imprinting and human disease. Ann. Rev.
Genet., 30, 173– 195.
10. Watson, P., Black, G., Ramsden, S., Barrow, M., Super, M., Kerr, B. and
Clayton-Smith, J. (2001) Angelman syndrome phenotype associated with
mutations in MECP2, a gene encoding a methyl CpG binding protein.
J. Med. Genet., 38, 224 –228.
11. Longo, I., Russo, L., Meloni, I., Ricci, I., Ariani, F., Pescucci, C.,
Giordano, C.T., Canitano, R., Hayek, G., Zappella, M. et al. (2004) Three
Rett patients with both MECP2 mutation and 15q11–13 rearrangements.
Eur. J. Hum. Genet., 12, 682 –685.
12. Schroer, R.J., Phelan, M.C., Michaelis, R.C., Crawford, E.C.,
Skinner, S.A., Cuccaro, M., Simensen, R.J., Bishop, J., Skinner, C.,
Fender, D. et al. (1998) Autism and maternally derived aberrations of
chromosome 15q. Am. J. Med. Genet., 76, 327 –336.
13. Nurmi, E.L., Bradford, Y., Chen, Y.-h., Hall, J., Arnone, B.,
Gardiner, M.B., Hutcheson, H.B., Gilbert, J.R., Pericak-Vance, M.A.,
Copeland-Yates, S.A. et al. (2001) Linkage disequilibrium at the
Angelman syndrome gene UBE3A in autism families. Genomics, 77,
105–113.
14. Buxbaum, J.D., Silverman, J.M., Smith, C.J., Greenberg, D.A.,
Kilifarski, M., Reichert, J., Cook, E.H., Fang, Y., Song, C.Y. and
Vitale, R. (2002) Association between a GABRB3 polymorphism and
autism. Mol. Psychiatry, 7, 311 –316.
15. Menold, M.M., Shao, Y., Wolpert, C.M., Donnelly, S.L., Raiford, K.L.,
Martin, E.R., Ravan, S.A., Abramson, R.K., Wright, H.H., Delong, G.R.
et al. (2001) Association analysis of chromosome 15 gabaa receptor
subunit genes in autistic disorder. J. Neurogenet., 15, 245– 259.
16. Samaco, R.C., Nagarajan, R.P., Braunschweig, D. and LaSalle, J.M.
(2004) Multiple pathways regulate MeCP2 expression in normal brain
development and exhibit defects in autism-spectrum disorders. Hum. Mol.
Genet., 13, 629 –639.
17. Samaco, R.C., Hogart, A. and LaSalle, J.M. (2005) Epigenetic overlap in
autism-spectrum neurodevelopmental disorders: MECP2 deficiency
causes reduced expression of UBE3A and GABRB3. Hum. Mol. Genet.,
14, 483–492.
18. 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.
19. Nan, X., Tate, P., Li, E. and Bird, A. (1996) DNA methylation specifies
chromosomal localization of MeCP2. Mol. Cell. Biol., 16, 414–421.
20. Jones, P.L., Veenstra, G.J., Wade, P.A., Vermaak, D., Kass, S.U.,
Landsberger, N., Strouboulis, J. and Wolffe, A.P. (1998) Methylated DNA
and MeCP2 recruit histone deacetylase to repress transcription. Nat.
Genet., 19, 187 –191.
21. 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.
22. Fuks, F., Hurd, P.J., Wolf, D., Nan, X., Bird, A.P. and Kouzarides, T.
(2003) The methyl-CpG-binding protein MeCP2 links DNA methylation
to histone methylation. J. Biol. Chem., 278, 4035–4040.
23. Balmer, D., Arredondo, J., Samaco, R.C. and LaSalle, J.M. (2002)
MECP2 mutations in Rett syndrome adversely affect lymphocyte growth,
but do not affect imprinted gene expression in blood or brain. Hum.
Genet., 110, 545 –552.
24. 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.
25. LaSalle, J., Goldstine, J., Balmer, D. and Greco, C. (2001) Quantitative
localization of heterologous methyl-CpG-binding protein 2 (MeCP2)
expression phenotypes in normal and Rett syndrome brain by laser
scanning cytometry. Hum. Mol. Genet., 10, 1729–1740.
26. Akbarian, S., Chen, R.Z., Gribnau, J., Rasmussen, T.P., Fong, H.,
Jaenisch, R. and Jones, E.G. (2001) Expression pattern of the Rett
syndrome gene MeCP2 in primate prefrontal cortex. Neurobiol. Dis.,
8, 784–791.
27. Holowacz, T. and De Boni, U. (1991) Arrangement of kinetochore
proteins and satellite DNA in neuronal interphase nuclei: changes induced
by gamma-aminobutyric acid (GABA). Exp. Cell Res., 197, 36 –42.
28. Manuelidis, L. and Borden, J. (1988) Reproducible compartmentalization
of individual chromosome domains in human CNS cells revealed by in
situ hybridization and three-dimensional reconstruction. Chromosoma,
96, 397–410.
29. Manuelidis, L. (1990) A view of interphase chromosomes. Science,
250, 1533–1540.
30. Arnoldus, E.P.J., Peters, A.C.B., Bots, G.T.A.M., Raap, A.K. and
van der Ploeg, M. (1989) Somatic pairing of chromosome 1 centromeres
in interphase nuclei of human cerebellum. Hum. Genet., 83, 231 –234.
31. Arnoldus, E.P., Noordermeer, I.A., Peters, A.C., Raap, A.K. and
van der Ploeg, M. (1991) Interphase cytogenetics reveals somatic pairing
of centromeres in normal human brain tissue, but no chromosome loss.
Cytogenet. Cell. Genet., 56, 214 –216.
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
SUPPLEMENTARY MATERIAL
795
796
Human Molecular Genetics, 2005, Vol. 14, No. 6
54. Cook, P.R. (1997) The transcriptional basis of chromosome pairing.
J. Cell Sci., 110, 1033–1040.
55. Tolhuis, B., Palstra, R.J., Splinter, E., Grosveld, F. and de Laat, W. (2002)
Looping and interaction between hypersensitive sites in the active
beta-globin locus. Mol. Cell, 10, 1453–1465.
56. 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.
57. Tudor, M., Akbarian, S., Chen, R.Z. and Jaenisch, R. (2002)
Transcriptional profiling of a mouse model for Rett syndrome reveals
subtle transcriptional changes in the brain. Proc. Natl Acad. Sci. USA, 99,
15536–15541.
58. Colantuoni, C., Jeon, O.H., Hyder, K., Chenchik, A., Khimani, A.H.,
Narayanan, V., Hoffman, E.P., Kaufmann, W.E., Naidu, S. and Pevsner, J.
(2001) Gene expression profiling in postmortem Rett syndrome brain:
differential gene expression and patient classification. Neurobiol. Dis., 8,
847–865.
59. Martinowich, K., Hattori, D., Wu, H., Fouse, S., He, F., Hu, Y.,
Fan, G. and Sun, Y.E. (2003) DNA methylation-related chromatin
remodeling in activity-dependent BDNF gene regulation. Science, 302,
890–893.
60. Chen, W.G., Chang, Q., Lin, Y., Meissner, A., West, A.E., Griffith, E.C.,
Jaenisch, R. and Greenberg, M.E. (2003) Derepression of BDNF
transcription involves calcium-dependent phosphorylation of MeCP2.
Science, 302, 885 –889.
61. 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.
62. Yu, F., Thiesen, J. and Stratling, W.H. (2000) Histone deacetylaseindependent transcriptional repression by methyl-CpG-binding
protein 2. Nucleic Acids Res., 28, 2201–2206.
63. Buschdorf, J.P. and Stratling, W.H. (2004) A WW domain binding
region in methyl-CpG-binding protein MeCP2: impact on Rett syndrome.
J. Mol. Med., 82, 135 –143.
64. Weitzel, J.M., Buhrmester, H. and Stratling, W.H. (1997) Chicken MARbinding protein ARBP is homologous to rat methyl-CpG-binding protein
MeCP2. Mol. Cell. Biol., 17, 5656–5666.
65. 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.
66. 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.
67. 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.
68. Ghoshal, K., Majumder, S., Datta, J., Motiwala, T., Bai, S., Sharma, S.M.,
Frankel, W. and Jacob, S.T. (2004) Role of human ribosomal RNA
(rRNA) promoter methylation and of methyl-CpG-binding protein MBD2
in the suppression of rRNA gene expression. J. Biol. Chem., 279,
6783–6793.
69. 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.
70. 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.
71. Muller, S. and Wienberg, J. (2001) “Bar-coding” primate chromosomes:
molecular cytogenetic screening for the ancestral hominoid karyotype.
Hum. Genet., 109, 85–94.
72. Farber, C., Dittrich, B., Buiting, K. and Horsthemke, B. (1999) The
chromosome 15 imprinting centre (IC) region has undergone multiple
duplication events and contains an upstream exon of SNRPN that is
deleted in all Angelman syndrome patients with an IC microdeletion.
Hum. Mol. Genet., 8, 337– 343.
73. Waterston, R.H., Lindblad-Toh, K., Birney, E., Rogers, J., Abril, J.F.,
Agarwal, P., Agarwala, R., Ainscough, R., Alexandersson, M., An, P.
et al. (2002) Initial sequencing and comparative analysis of the mouse
genome. Nature, 420, 520–562.
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
32. Martou, G. and De Boni, U. (2000) Nuclear topology of murine, cerebellar
Purkinje neurons: changes as a function of development. Exp. Cell Res.,
256, 131–139.
33. LaSalle, J. and Lalande, M. (1996) Homologous association of oppositely
imprinted chromosomal domains. Science, 272, 725–728.
34. Kallioniemi, O.P., Wagner, U., Kononen, J. and Sauter, G. (2001) Tissue
microarray technology for high-throughput molecular profiling of cancer.
Hum. Mol. Genet., 10, 657–662.
35. Kononen, J., Bubendorf, L., Kallioniemi, A., Barlund, M., Schraml, P.,
Leighton, S., Torhorst, J., Mihatsch, M.J., Sauter, G. and
Kallioniemi, O.P. (1998) Tissue microarrays for high-throughput
molecular profiling of tumor specimens. Nat. Med., 4, 844 –847.
36. Haaf, T. and Schmid, M. (1991) Chromosome topology in mammalian
interphase nuclei. Exp. Cell Res., 192, 325–332.
37. Balmer, D., Goldstine, J., Rao, Y.M. and LaSalle, J.M. (2003) Elevated
methyl-CpG-binding protein 2 expression is acquired during postnatal
human brain development and is correlated with alternative
polyadenylation. J. Mol. Med., 81, 61–68.
38. Pahlman, S., Hoehner, J.C., Nanberg, E., Hedborg, F., Fagerstrom, S.,
Gestblom, C., Johansson, I., Larsson, U., Lavenius, E., Ortoft, E. et al.
(1995) Differentiation and survival influences of growth factors in human
neuroblastoma. Eur. J. Cancer, 4, 453 –458.
39. Spengler, B.A., Biedler, J.L. and Ross, R.A. (2002) A corrected karyotype
for the SH-SY5Y human neuroblastoma cell line. Cancer Genet.
Cytogenet., 138, 177 –178.
40. Jung, B.P., Jugloff, D.G., Zhang, G., Logan, R., Brown, S. and
Eubanks, J.H. (2003) The expression of methyl CpG binding factor
MeCP2 correlates with cellular differentiation in the developing rat brain
and in cultured cells. J. Neurobiol., 55, 86 –96.
41. Morishita, R., Higaki, J., Tomita, N. and Ogihara, T. (1998) Application
of transcription factor “decoy” strategy as means of gene therapy and
study of gene expression in cardiovascular disease. Circ. Res., 82,
1023–1028.
42. Buiting, K., Saitoh, S., Gross, S., Dittrich, B., Schwartz, S., Nicholls, R.D.
and Horsthemke, B. (1995) Inherited microdeletions in the Angelman and
Prader–Willi syndromes define an imprinting centre on human
chromosome 15. Nat. Genet., 9, 395–400.
43. Georges, M., Charlier, C. and Cockett, N. (2003) The callipyge locus:
evidence for the trans interaction of reciprocally imprinted genes. Trends
Genet., 19, 248 –252.
44. Herman, H., Lu, M., Anggraini, M., Sikora, A., Chang, Y., Yoon, B.J. and
Soloway, P.D. (2003) Trans allele methylation and paramutation-like
effects in mice. Nat. Genet., 34, 199–202.
45. Tsai, T.F., Bressler, J., Jiang, Y.H. and Beaudet, A.L. (2003) Disruption of
the genomic imprint in trans with homologous recombination at Snrpn in
ES cells. Genesis, 37, 151–161.
46. Manuelidis, L. (1985) Individual interphase chromosome domains
revealed by hybridization. Hum. Genet., 71, 288–293.
47. De Boni, U. (1994) The interphase nucleus as a dynamic structure. Int.
Rev. Cytol., 150, 149 –171.
48. Dalrymple, S.J., Herath, J.F., Borell, T.J., Moertel, C.A. and Jenkins, R.B.
(1994) Correlation of cytogenetic and fluorescence in situ (FISH)
studies in normal and gliotic brain. J. Neuropathol. Exp. Neurol.,
53, 448 –456.
49. Herzing, L.B.K., Cook, E.H. and Ledbetter, D.H. (2002) Allele-specific
expression analysis by RNA-FISH demonstrates preferential maternal
expression of UBE3A and imprint maintenance within 15q11–q13
duplications. Hum. Mol. Genet., 11, 1707–1718.
50. Riesselmann, L. and Haaf, T. (1999) Preferential S-phase pairing of the
imprinted region on distal mouse chromosome 7. Cytogenet. Cell. Genet.,
86, 39 –42.
51. Collins, A.L., Levenson, J.M., Vilaythong, A.P., Richman, R.,
Armstrong, D.L., Noebels, J.L., Sweatt, J.D. and Zoghbi, H.Y. (2004)
Mild overexpression of MeCP2 causes a progressive neurological disorder
in mice. Hum. Mol. Genet., 21, 2679–2689.
52. 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.
53. Rehen, S.K., McConnell, M.J., Kaushal, D., Kingsbury, M.A., Yang, A.H.
and Chun, J. (2001) Chromosomal variation in neurons of the developing
and adult mammalian nervous system. Proc. Natl Acad. Sci. USA, 98,
13361–13366.
Human Molecular Genetics, 2005, Vol. 14, No. 6
78.
79.
80.
81.
and H19 into parent-specific chromatin loops. Nat. Genet., 36,
889–893.
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 Mecp22/þ mouse brain.
Hum. Mol. Genet., 78, 1275–1286.
de Belle, I., Cai, S. and Kohwi-Shigematsu, T. (1998) The genomic
sequences bound to special AT-rich sequence-binding protein 1 (SATB1)
in vivo in Jurkat T cells are tightly associated with the nuclear matrix
at the bases of the chromatin loops. J. Cell Biol., 141, 335 –348.
Orlando, V. and Paro, R. (1993) Mapping Polycomb-repressed domains
in the bithorax complex using in vivo formaldehyde cross-linked
chromatin. Cell, 75, 1187–1198.
Yasui, D., Miyano, M., Cai, S., Varga-Weisz, P. and
Kohwi-Shigematsu, T. (2002) SATB1 targets chromatin remodelling to
regulate genes over long distances. Nature, 419, 641–645.
Downloaded from http://hmg.oxfordjournals.org at University of California, Davis on May 11, 2010
74. Driscoll, D.J., Waters, M.F., Williams, C.A., Zori, R.T., Glenn, C.C.,
Avidano, K.M. and Nicholls, R.D. (1992) A DNA methylation
imprint, determined by the sex of the parent, distinguishes
the Angelman and Prader–Willi syndromes. Genomics,
13, 917 –924.
75. Dittrich, B., Robinson, W.P., Knoblauch, H., Buiting, K., Schmidt, K.,
Gillessen-Kaesbach, G. and Horsthemke, B. (1992) Molecular diagnosis
of the Prader–Willi and Angelman syndromes by detection of parentof-origin specific DNA methylation in 15q11–13. Hum. Genet., 90,
313 –315.
76. LaSalle, J.M., Ritchie, R.J., Glatt, H. and Lalande, M. (1998) Clonal
heterogeneity at allelic methylation sites diagnostic for
Prader–Williand Angelman syndromes. Proc. Natl Acad. Sci. USA,
95, 1675–1680.
77. Murrell, A., Heeson, S. and Reik, W. (2004) Interaction between
differentially methylated regions partitions the imprinted genes Igf2
797