Human Molecular Genetics, 2020, Vol. 29, No.

R1 R43–R51

doi: 10.1093/hmg/ddaa175
Advance Access Publication Date: 7 August 2020
Invited Review Article

INVITED REVIEW ARTICLE

The phenomenal epigenome in neurodevelopmental

disorders
Ummi Ciptasari1 and Hans van Bokhoven1,2,*
1 Department of Human Genetics, Donders Institute for Brain, Cognition and Behavior, Radboud university
medical center, 6500 HB Nijmegen, The Netherlands and 2 Department of Cognitive Neuroscience, Donders
Institute for Brain, Cognition and Behavior, Radboud university medical center, 6500 HB Nijmegen, The
Netherlands
*To whom correspondence should be addressed at: Department of Human Genetics 855, Donders Institute for Brain, Cognition and Behavior, Radboud
university medical center, PO Box 9101, 6500 HB Nijmegen, The Netherlands. Tel: +31 24 3614017; Email: [Link]@[Link]

Abstract
Disruption of chromatin structure due to epimutations is a leading genetic etiology of neurodevelopmental disorders,
collectively known as chromatinopathies. We show that there is an increasing level of convergence from the high diversity
of genes that are affected by mutations to the molecular networks and pathways involving the respective proteins, the
disrupted cellular and subcellular processes, and their consequence for higher order cellular network function. This
convergence is ultimately ref lected by specific phenotypic features shared across the various chromatinopathies. Based on
these observations, we propose that the commonly disrupted molecular and cellular anomalies might provide a rational
target for the development of symptomatic interventions for defined groups of genetically distinct neurodevelopmental
disorders.

Consequently, epigenetics is considered a rational mediator of

Introduction
gene × environment interactions. Finally, epigenetic modifica-
Almost all cells in our body contain the same DNA, yet clearly a tions are required for memory formation and other forms of
large variety of different cell types can be distinguished by their learning, suggesting that epigenetic modifications might act as
morphology, function and RNA expression profiles (1). The differ- a cellular memory (3,4). In view of this, it is not surprising that
ences between these cell types are controlled by chemical modi- disruptions of the epigenetic machinery are often underlying
fications and structural organization of the chromatin, popularly mental disorders alluded to as chromatinopathies (5).
alluded to as ‘epigenetic mechanisms’. Epigenetic modifications
guide the patterns of gene expression and are installed along
Epimutations in genes causing
pre-defined developmental lineages from the zygote to the full
set of differentiated cells and organs (2). Such epigenetic modi-
neurodevelopmental disorders
fications can be transmitted between successive generations of Our review is confined to monogenetic neurodevelopmental
cells to maintain cell identity. In addition, epigenetic modifica- disorders (NDDs) caused by highly penetrant secondary epimu-
tions can also be introduced by external triggers, e.g. exposure tations causing cognitive impairment as the major and consis-
to toxins, nutrients and stress, which can have life-long effects. tent feature. Previously, we and others have made compilations

Received: July 8, 2020. Revised: July 16, 2020. Accepted: July 30, 2020
© The Author(s) 2020. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ([Link]
which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

R43
 R44 Human Molecular Genetics, 2020, Vol. 29, No. R1

of such disorders and respective genes carrying epimutations lation, biotinylation, O-GlcNAcylation and DNA methylation).
(6–11). These overviews vary because of the rapid increase of Members of the superfamily II helicases involved in adenosine
new NDD genes. In addition, the definition ‘epigenetic genes’ triphosphate (ATP)-dependent nucleosome translocation are
is variably used in drafting these overviews. A useful dynamic prominent in the group of remodelers. These helicases cover
overview is maintained by the group of H. Bjornsson at [Link] five conserved protein complexes: (1) SWI/SNF (Switch/Sucrose
[Link] (12). Non-Fermentable) [a.k.a. BRG1-associated factors (BAF)], (2)
For this review we have defined chromatinopathy genes imitation switch, (3) chromatin helicase DNA binding (CHD),
based on their ability to chemically modify or remodel chromatin (4) INO80/SWR1 and (5) ATRX.
structure. This essentially encompasses a list of genes whose
corresponding protein products have a catalytic function in
installing post-translational histone modifications and DNA
Chromatinopathy genes cluster in epigenetic
modifications (writers), removing such modifications (erasers)
modules
or have chromatin remodeling activity (remodelers). The readers
of chromatin structures are confined to proteins that have a The functional interactions across chromatinopathy genes pro-
reader domain for binding to specific chromatin modifications vide a first step to picture how epigenetic dysregulation leads
and chromatin architectural proteins. To select these genes, to neurodevelopmental defects. To uncover possible genotype–
we first identified all human genes that satisfied these criteria phenotype network associations, hierarchical clustering of func-
by querying the Gene Ontology (GO) database ([Link] tional interactions with the 130 genes was performed using
[Link]/amigo/search/ontology; release May 2020) STRING (17). Using stringent connectivity based on experimental
with the search term ‘chromatin organization’ (GO:0006325). evidence, several chromatinopathy genes are not clearly con-
Chromatin organization is defined as any process that results nected to others. However, abundant direct interactions between
in the specification, formation or maintenance of the physical chromatinopathy genes catch the eye with major epigenetic
structure of eukaryotic chromatin and together with its child regulator complexes such as complex of proteins associated
terms (Supplementary Material, Table S1) seems to cover all with Set1 (COMPASS), polycomb recessive complex 2 (PRC2),
chromatin modulating processes. In total 771 of 19 858 human nucleosome remodeling deacetylase (NuRD) and BAF complex as
genes in the GO database are classified with the GO term converging hubs in the network. These complexes are connected
‘chromatin organization’. This group was complemented by to one another by other chromatinopathy proteins, such as β-
the selection of 295 genes from the epigenetic machinery catenin (CTNNB1) and transcriptional activators/HAT p300 and
database, which harbors a wider collection of stringently cAMP response element-binding protein (CREB) binding protein
selected chromatin readers, resulting in a collection of 872 (CREBBP/CBP).
genes (12) (Supplementary Material, Table S2). Next, we selected The BAF complex, or the mammalian SWI/SNF complex,
all genes for which causative variants have been reported to emerged as the strongest hub across chromatinopathy proteins
cause an NDD, using the SysID database ([Link] (18). The complex has the ability to mobilize nucleosomes and
[Link]/). SysID comprises 1333 verified NDD genes (March 2020) is composed by core ATPase proteins and unique components
for which causative monogenic variants have been reported dependent on the cell type in which they function. BAF
in multiple unrelated individuals with a neurodevelopmental complexes seem to have a genome-wide control on both
phenotype (13). Overlay of the SysID genes with the chromatin active H3K4me2 and repressive H3K27me3 marks during
control genes revealed 121 SysID genes with the GO annotation late cortical development by directly interacting with the
chromatin organization. In line with previous notions, this corresponding demethylases (19). Thus, inactivation of BAF
analysis shows that genes with an epigenetic function are complexes leads to H3K27me3-linked silencing (via KDM6A/B)
enriched in the list of NDD genes compared to the list containing of neuronal differentiation-related genes with concurrent
all human genes by a factor of 2.33 (P < 0.0001) (8,14), which H3K4me2-mediated activation (via KDM1A) of proliferation-
is similar to the enrichment of synapse-related genes [1354 associated genes via derepression of WNT signaling. Mutations
genes with GO term synapse, 210 in SysID; enrichment of 2,31 in genes encoding various subunits of BAF complex often lead
(P < 0.0001)].To be sure we did not miss a major number of to what is known as SWI/SNF-related intellectual disability
genes implicated in NDDs, we visually inspected the SysID (ID) or BAFopathies, including Coffin–Siris syndrome (CSS),
database, which revealed ADNP, YY1 and RBPJ as additional Nicolaides–Baraitser syndrome and other ID disorders, which
genes. Furthermore, a query of other disease gene databases, share phenotypic features of developmental delay, ID, coarse
such as the Developmental Disorders Genotype-to-Phenotype facial features and phalangeal abnormalities (20,21). The
database (DDG2P) table from the ‘Deciphering Developmental BAFopathies represent a great example of the genetic principle
Disorders’ project and the AutDB autism database (15,16), that mutation of components of the same protein complex or
revealed an additional six high-confidence chromatin-related nodes of the same molecular pathway give rise to disorders
genes that were still ranked as candidates in SysID: DDB2, DDX11, with overlapping phenotypes. A similar genotype–phenotype
GPX4, HDAC4, KMT5B and MORC2, bringing the total overlap to network association has been identified for Kleefstra syndrome,
130 genes (Supplementary Material, Table S3). which is caused by loss-of-function mutations of EHMT1,
The 130 epigenetic NDD genes encompass the writers, encoding euchromatic histone methyltransferase 1 (EHMT1),
erasers, remodelers and readers. Only 18 genes carry bi-allelic a transcriptional repressor via mono- and dimethylation of
variants that cause an NDD phenotype. The majority is either histone 3 lysine 9 (H3K9me1/2). Exome sequencing of individuals
X-linked or has disease-causing variants that manifest in a dom- with a phenotype strongly reminiscent of Kleefstra syndrome
inant pattern that often arise by de novo mutation. This indicates (Kleefstra syndrome spectrum [KSS]) revealed de novo mutations
that chromatinopathy genes are highly dosage sensitive. Histone in several other chromatinopathy genes, including MBD5, KMT2C
methyltransferases (HMTs) are particularly abundant amongst and the BAF complex gene SMARCB1 (22). The differential
the writers, followed by histone acetyltransferases (HATs), and phenotypes of CSS and KSS associated with SMARCB1 might
other rare chromatin modifiers (ubiquitination, phosphory- be explained by the different functional effects of the respective
 Human Molecular Genetics, 2020, Vol. 29, No. R1 R45

variants in addition to the clinical variability typically observed macrocephaly and cortical malformations. Previous integrated
for individual chromatinopathy disorders (21,23,24). Physical analysis of six exome studies uncovered a network of genes
and functional interactions amongst the KSS proteins revealed mutated in autism spectrum disorder (ASD) and ID highlighting
an EHMT1-associated chromatin-modification module, which is proteins converging on β-catenin as a part of the WNT sig-
connected to the BAF, COMPASS and PRC2 complexes (Fig. 1). naling pathway (41). In support of this, a recent study defined
Polycomb recessive complex 1 (PRC1) and PRC2 are both molecular and functional subclasses within a set of 30 autism-
responsible for silencing genes unnecessary for the current related genes (in fact better defined as NDD genes) based on
state of the cells through monoubiquitination of H2AK119 and the in vitro neurogenic potency of pluripotent stem cells carry-
methylation of H3K27, respectively. PRC2 is a highly conserved ing engineered mutations in these genes (42). This identified a
multimeric complex consisting of four core proteins: EZH1/2, class of eight genes with mutations associated with increased
EED, SUZ12 and RBBP4/7, and a variety of accessory proteins proliferation and inhibited neurogenesis (class 1 comprising
that co-regulate PRC2 activity (25). Pathogenic mutations in chromatinopathy genes ASH1L, ASXL3 KDM5B and KMT2C). The
EZH2, SUZ12 and EED are seen in overgrowth syndromes with reduced neurogenesis was strongly correlated with a reduced
variable ID, such as Weaver syndrome and Cohen–Gibson response to WNT signaling. Notably, five of the eight class 1
syndrome (26–28). PRC2 interacts with DNMT3A. Somatic genes are known as regulators of polycomb, which like WNT is
DNMT3A dominant-negative/loss-of-function mutations that critical for coordinating the balance between NPC proliferation
lead to DNA hypomethylation at PRC2-responsive loci are and differentiation (43–48). PRC2 positively regulates WNT/β-
frequently seen in hematological disorders. Similar germline catenin signaling in various cell types through the silencing of
mutations are also encountered in Tatton–Brown–Rahman WNT antagonists, and EZH2 expression maintains NPC prolifer-
syndrome, which is characterized by overgrowth, ID, tall stature, ation and is negatively correlated with neuronal differentiation
distinctive facial appearance and in some cases leukemia (29). (49–51). Accordingly, mutations of EZH2 in Weaver syndrome
Interestingly, DNMT3A gain-of-function mutations in Heyn– are associated with overgrowth and macrocephaly. It should
Sproul–Jackson syndrome are associated with hypermethylation be noted, however, that the study of Cederquist et al. (42) also
at PRC2 loci and pre- and postnatal growth retardation (30). uncovered a smaller class of genes with the opposite neurogenic
Other connections of PRC2 to chromatinopathy proteins effects, which did not have an altered WNT response despite
comprise NSD1, underlying another overgrowth disorder (Sotos the fact that two chromatinopathy genes from this class, KMT2A
syndrome), YY1 (Gabriele–de Vries syndrome) and ATRX (several and CHD8, are also regulators of polycomb. Moreover, mutations
X-linked ID disorders). in these genes have variable growth defects in patients. This
PRC2 cooperatively interacts also with EHMT1 to exert repres- observation reinforces observations that the neurogenic prop-
sive activity by PRC2-installed H3K27me3 marks. The repres- erties of WNT and polycomb complex might be dependent on
sion is countered by the H3K27 demethylase activity of KDM6A the specifics of the progenitor cells and the timing in cortical
(31,32), which is affected by hemizygous loss-of-function muta- development (48).
tions in Kabuki syndrome type 2 (33). KDM6A is an integral regu- Nevertheless, these exciting data reinforce the notion that
lator of macromolecular complexes called COMPASS that consist the balance between proliferation and differentiation controlled
of variable constellations of SET family HMT with H3K4me1/2/3 by WNT signaling is mirrored by a delicate chromatin-based
activity (SETD1A/B and MLL/KMT2 proteins) (34). Germline muta- balance between activation and repression involving compo-
tions in all COMPASS methyltransferases have been implicated nents of PRC2 and COMPASS. Previously, also the BAF complex
in NDDs, which share a range of clinical phenotypes includ- was shown to control WNT signaling. Loss-of-function muta-
ing prominent cognitive impairment, behavioral abnormalities tions affecting individual components in BAFopathies lead to
and features of abnormal craniofacial, skeletal and cardiovas- activation of proliferation-associated genes via derepression of
cular development (22,35–37). Interestingly, KMT2D, which is WNT/β-catenin in various proliferative cell types (52–54). CHD8,
mutated in Kabuki syndrome type 1, interacts with KDM6A and connecting BAF and COMPASS complexes, negatively regulates
CBP/p300 HAT (Rubinstein–Taybi syndrome) to promote tran- β-catenin-targeted gene expression important in cellular prolif-
scriptional activation (38,39). KMT2C was found to be mutated eration and differentiation (55,56). In Chd8 depleted mice, PRC2
in a number of KSS patients, which is somewhat surprising as regulation of WNT signaling appears to be the major altered
KMT2C and EHMT1 have opposing activities as transcriptional pathway (57), and accordingly, an overgrowth phenotype with
activator and repressor, respectively. Many other interactions macrocephaly has been reported in CHD8-related NDD patients
between COMPASS and other chromatinopathy proteins have and animal models (28,58–60).
been uncovered, such as with the chromatin remodeler CHD8,
which activates expression of BRG1-associated BAF complexes
Transcriptional convergence at the synapse in acute
and recruits MLL/KMT2 COMPASS complexes to promote central
nervous system myelination and myelin repair (18,40,41).
neurophysiological processes
Communication between the synapse and the nucleus in both
directions is critical for neuronal development and plasticity.
Synaptic activity triggers Ca-dependent signaling via signaling
Transcriptional convergence in neurogenesis:
pathways such as PI3K-AKT-mTOR and RAS-MAPK that lead
WNT signaling to activation and repression of genes that control neural con-
Not surprisingly, epimutations in chromatinopathies are asso- nectivity and function. Integration of these activity-dependent
ciated with transcriptional deregulation of genes that are crit- transcriptional processes is mediated by some key transcription
ical for proper neurodevelopment as well as genes that are factors, such as CREB, MEF2 and MECP2, and heavily involves
important for acute neurophysiological processes. Disruption of the regulation of chromatin organization. Well-known direct tar-
genes during neurogenesis is often associated with imbalances gets of these transcriptional complexes include immediate early
in proliferation and differentiation of neural progenitors cells genes such as c-FOS, ARC and BDNF, and disrupted expression
(NPCs), giving rise to structural brain anomalies such as micro/ of these genes is seen in many chromatinopathies (37,61–64).
 R46 Human Molecular Genetics, 2020, Vol. 29, No. R1

Figure 1. Simplified protein–protein interaction network of compiled chromatinopathy genes generated by StringDB (version 11.0, MCL clustering) analysis of
chromatinopathy genes. For clarity, only proteins with interactions are shown. Connecting lines showing medium- and high-confidence experimental interactions
(confidence > 0.400) with line thickness indicates the strength of data support.
 Human Molecular Genetics, 2020, Vol. 29, No. R1 R47

The deregulated expression appears to be a direct consequence parvalbumin (PV) expression together with excessive excitatory
of the causative mutation in these genes, and not just a sec- inputs converging onto PV-expressing inhibitory neurons in the
ondary effect of neuronal pathology. Genome-wide expression somatosensory and motor cortex associated with the sensori-
studies in animal and cell models of chromatinopathies show an motor defects (81). Distinct types of inhibitory neurons appear
enrichment of differentially expressed genes that are associated to contribute differently to phenotypic manifestations. Loss of
with synaptic processes (65–69). For example, direct EHMT1 tar- MECP2 in PV+ neurons gives rise to progressive motor, sensory,
gets that were inappropriately repressed in the Ehmt1+/− mouse memory and social deficits, whereas those lacking MeCP2 in
model for KS were linked to ion channels, synaptic vesicle and somatostatin-positive (SOM+ ) neurons leads to seizures and
calcium signaling, and include BDNF as an important factor in stereotypic behavior. A similar observation was made for Arid1b
the disrupted homeostatic plasticity in these animals (64,70–72). haploinsufficiency, which in PV+ neurons contributed to social
Epigenetic repressors such as EHMT1 and KDM5C interact with and emotional impairments, whereas Mecp2 loss in SOM+ neu-
the RE1 silencing transcription factor (REST) to repress neuronal rons caused stereotypies as well as learning and memory dys-
genes in extraneuronal tissues, and accordingly, loss of REST function (54,82). An interesting observation is that the overall
activity leads to upregulation of synaptic proteins, including network imbalance contributing to the phenotypic regression in
SNAP25, Synapsin1, SCN2A, NMDAR1 and BDNF (73–75). In addi- RTT involves N-methyl-D-aspartate receptor (NMDAR) dysfunc-
tion, EHMT1 appears to be a critical regulator of neuronal speci- tion. Indeed, genetic deletion of the NMDAR subunit GluN2A
fication by regulation of protocadherins, possibly in conjunction in mice or NMDAR inhibition by ketamine prevented RTT phe-
with REST, and olfactory receptors. notypes (83,84). Similarly, EHMT1 deficiency in both mice- and
Recent work using neuronal knockdown of genes from the human-induced pluripotent stem cells-derived cortical excita-
KSS epigenetic module (EHMT1, MBD5, KMT2C and SMARCB1) tory neurons reduced the deposition of repressive H3K9me2
identified commonly dysregulated classes of target genes, marks at the GRIN1 promoter and, encoding an NMDAR sub-
mainly involved in regulating neuronal excitability and synaptic unit, leading to increased GRIN1 expression and altered NMDAR
function (76). However, the deregulation of individual genes function (85). Based on these preclinical observations, modula-
from these classes shows differences in line with the specific tion of NMDAR is hypothesized to be a potential symptomatic
activities of these proteins. Thus, knockdown of KMT2C and treatment for several chromatinopathies, and clinical trials with
SMARCB1 generally leads to downregulation of genes, whereas ketamine are currently ongoing for RTT and Helsmoortel–Van
EHMT1 knockdown yielded more upregulated than downreg- der Aa syndrome.
ulated genes, including genes encoding proteins required for
ion transport and synaptic transmission. MBD5 knockdown
had more subtle effects on the expression of these genes
Convergence as a target for therapy
and in addition showed a more pronounced deregulation of Each of the individual chromatinopathy disorders has an
cell adhesion and nucleosome assembly proteins. Thus, KSS extremely low prevalence, and therefore these disorders are
target genes converge on neuronal excitability and synaptic not very attractive for developing gene-based interventions.
function, albeit in variable patterns for each specific KSS gene. However, the patterns of convergence that we have described
Despite these differences, the cumulative effect of deregulation for chromatinopathies might offer opportunities for developing
of the synapse-related genes appears to result in a remarkably therapeutic strategies that are applicable to a wider group
similar disruption of intrinsic synaptic properties and neuronal of disorders. Such interventions can be directed at any level
network activity. in the convergence pyramid (Fig. 2): protein complexes, gene
expression, signaling pathways and cellular processes such as
synaptic activity. This idea is not restricted to chromatinopathies
Neurophysiological phenotype convergence and convergence can be seen for other NDDs as well. For
An emerging concept of a common neurophysiological mode- example, disrupted mTor signaling is well known to be affected
of-failure is the excitation/inhibition (E/I) imbalance hypothesis, in a large number of different NDDs and is downstream of major
brought forward by Rubenstein and Merzenich in 2003, which signaling pathways, PI3K-AKT and RAS-MAPK, which are often
postulates that excessive excitatory activity relative to inhibitory directly affected by genetic variants in many rare syndromes (86).
ones is a major contributor to autism pathology. This hypothesis Consequently, rapamycin and other mTor modulators are being
was supported by the high comorbidity of epilepsy and autism used in trials for ID and ASD disorders. Similarly, regulators
(77). The E/I imbalance hypothesis has since grown into a broader of E/I balance and NMDAR antagonist like ketamine might be
concept encompassing a large range of perturbations within the considered for treatment of several chromatinopathies. The
excitation or inhibition neuronal circuits for a broader group of downside of such generic interventions is that the expected
NDDs (78–80). E/I imbalance might be caused by alterations in effects are likely restricted to specific symptomatic features. Yet,
glutamatergic excitatory drive, inhibitory connectivity or a com- very often these features are comorbid across a large number
bination thereof. As mentioned, KSS gene-deficient neuronal of NDDs, regardless of the classical categorical classification
networks showed increased neuronal excitability, hyperactive such as ID, ASD, schizophrenia and others. Therefore, such
networks with altered network organization and E/I balance symptomatic interventions might be attractive to consider for
(76). For all KSS genes, the shift in E/I balance was toward many genetically distinct disorders. It is of eminent importance
excitation. For EHMT1, this was associated with reduced num- to further strengthen the patterns of convergence to increase
ber of inhibitory synapses, whereas for SMARCB1, KMT2C and knowledge about the underlying biological mechanisms in order
MBD5, a reduction of both inhibitory and excitatory synapses to develop rational strategies for intervention applicable to larger
was seen (76). groups of disorders caused by different genetic defects.
In Rett syndrome (RTT), the E/I balance can either be shifted
toward excessive excitation or inhibition depending on the brain
area. The importance of a tightly regulated E/I balance is demon-
Supplementary Material
strated by the observation that knockout of Mecp2 show elevated Supplementary material is available at HMG online.
 R48 Human Molecular Genetics, 2020, Vol. 29, No. R1

Figure 2. Schematic representation of convergence of chromatinopathies. Individual chromatinopathy genes encode proteins that interact with each other in chromatin
remodeling. These interactions cluster in chromatin remodeling complexes such as the BAF, PRC2 and COMPASS complexes, which then affect the transcription of their
target genes. Aberrant transcription level of various genes in different chromatinopathies leads to commonly disrupted molecular and cellular processes, such as WNT
signaling and genes involved in acute synaptic function. Disrupted WNT signaling affects proliferation of stem cells and the process of neurogenesis, which can lead
to anomalies in brain development such as micro- and macrocephaly. Synaptic function and activity is an important process in brain circuits development. Imbalance
of excitation and inhibitory (E/I) has emerged as a common neurophysiological mode-of-failure in ASD, ID and NDDs in general. Combination of disruption of multiple
developmental processes leads to commonly shared symptoms, which in different combinations and levels of severity builds onto specific chromatinopathy syndromes
and disorders.
 Human Molecular Genetics, 2020, Vol. 29, No. R1 R49

Acknowledgements Synaptic, transcriptional and chromatin genes disrupted in

autism. Nature, 515, 209–215.
The authors are thankful to Freek Manders and Britt Mossink,
15. Basu, S.N., Kollu, R. and Banerjee-Basu, S. (2009) AutDB: a
who helped in compiling the chromatinopathy genes reported
gene reference resource for autism research. Nucleic Acids
in this review. The authors are also indebted by the continuous
Res., 37, 832–836.
support from their colleagues Tjitske Kleefstra, Nael Nadif Kasri,
16. Firth, H.V., Richards, S.M., Bevan, A.P., Clayton, S., Corpas, M.,
Annette Schenck, Lisenka Vissers and Christian Gilissen.
Rajan, D., Van Vooren, S., Moreau, Y., Pettett, R.M., Carter, N.P.
Conflict of Interest statement. None declared. (2009) DECIPHER: Database of Chromosomal Imbalance and
Phenotype in Humans Using Ensembl Resources. Am. J. Hum.
Genet., 84, 524–533.
Funding 17. von Mering, C., Jensen, L.J., Snel, B., Hooper, S.D., Krupp,
M., Foglierini, M., Jouffre, N., Huynen, M.A., Bork, P. (2005)
U.C. is supported by the ZonMw-E-Rare 3 grant (JTC2018),
STRING: known and predicted protein-protein associations,
IMPACT: Identification of converging Molecular Pathways Across
integrated and transferred across organisms. Nucleic Acids
Chromatinopathies as Targets for Therapy.
Res., 33, 433–437.
18. Ronan, J.L., Wu, W. and Crabtree, G.R. (2013) From neural
development to cognition: unexpected roles for chromatin.
References Nat. Rev. Genet., 14, 347–359.
1. Macosko, E.Z., Basu, A., Satija, R., Nemesh, J., Shekhar, K., 19. Nguyen, H., Kerimoglu, C., Pirouz, M., Pham, L., Kiszka, K.A.,
Goldman, M., Tirosh, I., Bialas, A.R., Kamitaki, N., Martersteck, Sokpor, G., Sakib, M.S., Rosenbusch, J., Teichmann, U., Seong,
E.M. et al. (2015) Highly parallel genome-wide expression R.H. et al. (2018) Epigenetic regulation by BAF complexes
profiling of individual cells using nanoliter droplets. Cell, limits neural stem cell proliferation by suppressing Wnt
161, 1202–1214. signaling in late embryonic development. Stem Cell Rep., 10,
2. Cavalli, G. and Heard, E. (2019) Advances in epigenetics 1734–1750.
link genetics to the environment and disease. Nature, 571, 20. Aref-Eshghi, E., Bend, E.G., Hood, R.L., Schenkel, L.C., Carere,
489–499. D.A., Chakrabarti, R., Nagamani, S., Cheung, S.W., Campeau,
3. Vecsey, C.G., Hawk, J.D., Lattal, K.M., Stein, J.M., Fabian, P.M., Prasad, C. et al. (2018) BAFopathies’ DNA methyla-
S.A., Attner, M.A., Cabrera, S.M., McDonough, C.B., Brindle, tion epi-signatures demonstrate diagnostic utility and func-
P.K., Abel, T., Wood, M.A. (2007) Histone deacetylase tional continuum of Coffin-Siris and Nicolaides-Baraitser
inhibitors enhance memory and synaptic plasticity via syndromes. Nat. Commun., 9, 4885.
CREB: CBP-dependent transcriptional activation. J. Neurosci., 21. Bögershausen, N. and Wollnik, B. (2018) Mutational land-
27, 6128–6140. scapes and phenotypic spectrum of SWI/SNF-related intel-
4. Day, J.J. and Sweatt, J.D. (2011) Epigenetic mechanisms in lectual disability disorders. Front. Mol. Neurosci., 11, 1–18.
cognition. Neuron, 70, 813–829. 22. Kleefstra, T., Kramer, J.M., Neveling, K., Willemsen, M.H., Koe-
5. Haggarty, S.J. and Tsai, L.-H. (2011) Probing the role of HDACs mans, T.S., Vissers, E.E.L.M., Wissink-Lindhout, W., Fenckova,
and mechanisms of chromatin-mediated neuroplasticity. M., van der Akker, W.M.R., Kasri, N.N. et al. (2012) Disruption
Neurobiol. Learn. Mem., 96, 41–52. of an EHMT1-associated chromatin-modification module
6. Kramer, J.M. and van Bokhoven, H. (2009) Genetic and epige- causes intellectual disability. Am. J. Hum. Genet., 91, 73–82.
netic defects in mental retardation. Int. J. Biochem. Cell Biol., 23. Valencia, A.M., Collings, C.K., Dao, H.T., St. Pierre, R., Cheng,
41, 96–107. Y.C., Huang, J., Sun, J., Seo, H.S., Mashtalie, N., Comstock,
7. van Bokhoven, H. (2011) Genetic and epigenetic networks in D.E. et al. (2019) Recurrent SMARCB1 mutations reveal a
intellectual disabilities. Annu. Rev. Genet., 45, 81–104. nucleosome acidic patch interaction site that potentiates
8. Kleefstra, T., Schenck, A., Kramer, J.M., van Bokhoven, H. mSWI/SNF complex chromatin remodeling. Cell, 179, 1–15.
(2014) The genetics of cognitive epigenetics. Neuropharmacol- 24. Kosho, T., Miyake, N. and Carey, J.C. (2014) Coffin-Siris
ogy, 80, 83–94. syndrome and related disorders involving components of
9. Fahrner, J.A. and Bjornsson, H.T. (2014) Mendelian disorders the BAF (mSWI/SNF) complex: historical review and recent
of the epigenetic machinery: tipping the balance of chro- advances using next generation sequencing. Am. J. Med.
matin states. Annu. Rev. Genomics Hum. Genet., 15, 269–293. Genet. Part C Semin. Med. Genet., 166, 241–251.
10. Bjornsson, H.T. (2015) The Mendelian disorders of the epige- 25. Kuzmichev, A., Nishioka, K., Erdjument-Bromage, H., Tempst,
netic machinery. Genome Res., 25, 1473–1481. P., Reinberg, D. (2002) Histone methyltransferase activity
11. Fahrner, J.A. and Bjornsson, H.T. (2019) Mendelian disor- associated with a human multiprotein complex containing
ders of the epigenetic machinery: postnatal malleability and the enhancer of zeste protein. Genes Dev., 16, 2893–2905.
therapeutic prospects. Hum. Mol. Genet., 28, R254–R264. 26. Cyrus, S., Burkardt, D., Weaver, D.D., Gibson, W.T. (2019) PRC2-
12. Boukas, L., Havrilla, J.M., Hickey, P.F., Quinlan, A.R., Bjornsson, complex related dysfunction in overgrowth syndromes: a
H.T., Hansen, K.D. (2019) Coexpression patterns define epi- review of EZH2, EED, and SUZ12 and their syndromic phe-
genetic regulators associated with neurological dysfunction. notypes. Am. J. Med. Genet. Part C Semin. Med. Genet., 181,
Genome Res., 29, 532–542. 519–531.
13. Kochinke, K., Zweier, C., Nijhof, B., Fenckova, M., Cizek, 27. Choufani, S., Gibson, W.T., Turinsky, A.L., Chung, B.H.Y.,
P., Honti, F., Keerthikumar, S., Oortveld, M.A.W., Kleefstra, Wang, T., Garg, K., Vitriolo, A., Cohen, A.S.A., Cyrues, S.,
T., Kramer, J.M. et al. (2016) Systematic phenomics anal- Goodman, S. et al. (2020) DNA methylation signature for
ysis deconvolutes genes mutated in intellectual disability EZH2 functionally classifies sequence variants in three PRC2
into biologically coherent modules. Am. J. Hum. Genet., 98, complex genes. Am. J. Hum. Genet., 106, 596–610.
149–164. 28. Tatton-Brown, K., Loveday, C., Yost, S., Clarke, M., Ramsay, E.,
14. De Rubeis, S., He, X., Goldberg, A.P., Poultney, C.S., Samocha, Zachariou, A., Elliot, A., Wylie, H., Ardissone, A., Rittinger, O.
K., Cicek, A.E., Kou, Y., Liu, L., Fromer, M., Walker, S. et al. (2014) et al. (2017) Mutations in epigenetic regulation genes are a
 R50 Human Molecular Genetics, 2020, Vol. 29, No. R1

major cause of overgrowth with intellectual disability. Am. J. S.Y. et al. (2020) A multiplex human pluripotent stem cell
Hum. Genet., 100, 725–736. platform defines molecular and functional subclasses of
29. Tatton-Brown, K., Seal, S., Ruark, E., Harmer, J., Ramsay, E., autism-related genes. Cell Stem Cell, 27, 35–49.e6.
del Vecchio Duarte, S., Zacharious, A., Hanks, S., O’Brien, E., 43. Zhang, Z., Lei, A., Xu, L., Chen, L., Chen, Y., Zhang, X.,
Aksglaede, L. et al. (2014) Mutations in the DNA methyltrans- Gao, Y., Yang, X., Zhang, M. and Cao, Y. (2017) Similarity in
ferase gene DNMT3A cause an overgrowth syndrome with gene-regulatory networks suggests that cancer cells share
intellectual disability. Nat. Genet., 46, 385–388. characteristics of embryonic neural cells. J. Biol. Chem., 292,
30. Heyn, P., Logan, C.V., Fluteau, A., Challis, R.C., Auchyn- 12842–12859.
nikava, T., Martin, C.A., Marsh, J.A., Taglini, F., Kilanowski, 44. Yu, Y., Chou, R., Shyu, W., Hsieh, S., Wu, C., Chiang, S., Chang,
F., Parry, D.A. et al. (2019) Gain-of-function DNMT3A muta- W., Chen, J., Tseng, Y., Lin, Y. et al. (2013) Smurf2-mediated
tions cause microcephalic dwarfism and hypermethylation degradation of EZH2 enhances neuron differentiation and
of polycomb-regulated regions. Nat. Genet., 51, 96–105. improves functional recovery after ischaemic stroke. EMBO
31. Mozzetta, C., Pontis, J., Fritsch, L., Robin, P., Portoso, M., Mol. Med., 5, 531–547.
Proux, C., Margueron, R., Ait-Si-Ali, S. (2014) The histone H3 45. Zhang, J., Ji, F., Liu, Y., Lei, X., Li, H., Ji, G., Yuan, Z. and Jiao, J.
lysine 9 methyltransferases G9a and GLP regulate polycomb (2014) Ezh2 regulates adult hippocampal neurogenesis and
repressive complex 2-mediated gene silencing. Mol. Cell, 53, memory. J. Neurosci., 34, 5184–5199.
277–289. 46. Lei, A., Chen, L., Zhang, M., Yang, X., Xu, L., Cao, N., Zhang,
32. Ler, L.D., Ghosh, S., Chai, X., Thike, A.A., Heng, H.L., Siew, Z. and Cao, Y. (2019) EZH2 regulates protein stability via
E.Y., Dey, S., Koh, L.K., Lim, J.Q., Lim, W.K. et al. (2017) Loss recruiting USP7 to mediate neuronal gene expression in
of tumor suppressor KDM6A amplifies PRC2-regulated tran- cancer cells. Front. Genet., 10, 1–18.
scriptional repression in bladder cancer and can be targeted 47. Munji, R.N., Choe, Y., Li, G., Siegenthaler, J.A. and Pleasure,
through inhibition of EZH2. Sci. Transl. Med., 9, 984–985. S.J. (2011) Wnt signaling regulates neuronal differentiation of
33. Lederer, D., Grisart, B., Digilio, M.C., Benoit, V., Crespin, cortical intermediate progenitors. J. Neurosci., 31, 1676–1687.
M., Ghariani, S.C., Maystadt, I., Dallapiccola, B., Verellen- 48. Testa, G. (2011) The time of timing: how polycomb proteins
Dumoulin, C. (2012) Deletion of KDM6A, a histone demethy- regulate neurogenesis. Bioessays, 33, 519–528.
lase interacting with MLL2, in three patients with Kabuki 49. Adhikari, A. and Davie, J. (2018) JARID2 and the PRC2 complex
syndrome. Am. J. Hum. Genet., 90, 119–124. regulate skeletal muscle differentiation through regulation
34. Shilatifard, A. (2012) The COMPASS family of histone H3K4 of canonical Wnt signaling. Epigenetics Chromatin, 11, 1–20.
methylases: mechanisms of regulation in development and 50. Cheng, A.S.L., Lau, S.S., Chen, Y., Kondo, Y., Li, M.S., Feng, H.,
disease pathogenesis. Annu. Rev. Biochem., 81, 65–95. Ching, A.K., Cheung, K.F., Wong, H.K., Tong, J.H. (2011) EZH2-
35. Lavery, W.J., Barski, A., Wiley, S., Schorry, E.K., Lindsay, mediated concordant repression of Wnt antagonists pro-
A.W. (2020) KMT2C/D COMPASS complex-associated dis- motes β-catenin-dependent hepatocarcinogenesis. Cancer
eases [KCDCOM-ADs]: an emerging class of congenital reg- Res., 71, 4028–4039.
ulopathies. Clin. Epigenetics, 12, 10. 51. Hwang, W.W., Salinas, R.D., Siu, J.J., Kelley, K.W., Delgado, R.N.,
36. Fagan, R.J. and Dingwall, A.K. (2019) COMPASS ascending: Paredes, M.F., Alvarez-Buylla, A., Oldham, M.C., Lim, D.A. et al.
emerging clues regarding the roles of MLL3/KMT2C and (2014) Distinct and separable roles for EZH2 in neurogenic
MLL2/KMT2D proteins in cancer. Cancer Lett., 458, 56–65. astroglia. Elife, e02439.
37. Koemans, T.S., Kleefstra, T., Chubak, M.C., Stone, M.H., Rejin- 52. Vasileiou, G., Ekici, A.B., Uebe, S., Zweier, C., Hoyer, J., Engels,
ders, M.R.F., de Munnik, S., Willemsen, M.H., Fenckova, M., H., Behrens, J., Reis, A., Hadjihannas, M.V. (2015) Chromatin-
Stumpel, C.T.R.M., Bok, L.A. et al. (2017) Functional conver- remodeling-factor ARID1B represses Wnt/β-catenin signal-
gence of histone methyltransferases EHMT1 and KMT2C ing. Am. J. Hum. Genet., 97, 445–456.
involved in intellectual disability and autism spectrum dis- 53. Wu, J., Keng, V.W., Patmore, D.M., Kendall, J.J., Petel, A.V.,
order. PLoS Genet., 13, e1006864. Jousma, E., Jessen, W.J., Choi, K., Tschida, B.R., Silverstein,
38. Gozdecka, M., Meduri, E., Mazan, M., Tzelepis, K., Dudek, M., K.A.T. et al. (2016) Insertional mutagenesis identifies a
Knights, A.J., Pardo, M., Yu, L., Choudhary, J.S., Metzakopian, E. STAT3/Arid1b/β-catenin pathway driving neurofibroma ini-
et al. (2018) UTX-mediated enhancer and chromatin remod- tiation. Cell Rep., 14, 1979–1990.
eling suppresses myeloid leukemogenesis through noncat- 54. Jung, E.M., Moffat, J.J., Liu, J., Dravid, S.M., Gurumurthy, C.B.,
alytic inverse regulation of ETS and GATA programs. Nat. Kim, W.Y. (2017) Arid1b haploinsufficiency disrupts cortical
Genet., 50, 883–894. interneuron development and mouse behavior. Nat. Neu-
39. Wang, S.P., Tang, Z., Chen, C.W., Shimada, M., Koche, R.P., rosci., 20, 1694–1707.
Wang, L.H., Nakadai, T., Chrameic, A., Krivstov, A.V., Arm- 55. Thompson, B.A., Tremblay, V., Lin, G., Bochar, D.A. (2008)
strong, S.A. and Roeder, R.G. (2017) A UTX-MLL4-p300 tran- CHD8 is an ATP-dependent chromatin remodeling factor
scriptional regulatory network coordinately shapes active that regulates β-catenin target genes. Mol. Cell. Biol., 28,
enhancer landscapes for eliciting transcription. Mol. Cell, 67, 3894–3904.
308–321.e6. 56. Nishiyama, M., Skoultchi, A.I. and Nakayama, K.I. (2012)
40. Zhao, C., Dong, C., Frah, M., Deng, Y., Marie, C., Zhang, F., Xu, Histone H1 recruitment by CHD8 is essential for suppression
L., Ma, Z., Dong, X., Lin, Y. et al. (2018) Dual requirement of of the Wnt-β-catenin signaling pathway. Mol. Cell. Biol., 32,
CHD8 for chromatin landscape establishment and histone 501–512.
methyltransferase recruitment to promote CNS myelination 57. Durak, O., Gao, F., Kaeser-Woo, Y.J., Rueda, R., Martorell,
and repair. Dev. Cell, 45, 753–768.e8. A.J., Nott, A., Liu, C.Y., Watsons, L.A. and Tsai, L.H. (2016)
41. Krumm, N., O’Roak, B.J., Shendure, J. and Eichler, E.E. (2014) Chd8 mediates cortical neurogenesis via transcriptional reg-
A de novo convergence of autism genetics and molecular ulation of cell cycle and Wnt signaling. Nat. Neurosci., 19,
neuroscience. Trends Neurosci., 37, 95–105. 1477–1488.
42. Cederquist, G.Y., Tchieu, J., Callahan, S.J., Rammarine, K., 58. Ostrowski, P.J., Zachariou, A., Loveday, C., Beleza-Meireles,
Ryan, S., Zhang, C., Rittemnhouse, C., Zeltner, N., Chung, A., Bertoli, M., Dean, J., Douglas, A.G.L., Ellis, I., Foster, A.,
 Human Molecular Genetics, 2020, Vol. 29, No. R1 R51

Graham, J.L. et al. (2019) The CHD8 overgrowth syndrome: a H3K4 demethylase SMCX links REST target genes to X-linked
detailed evaluation of an emerging overgrowth phenotype in mental retardation. Nature, 447, 601–605.
27 patients. Am. J. Med. Genet. Part C Semin. Med. Genet., 181, 73. Iacono, G., Dubos, A., Méziane, H., Benevento, M., Habibi, E.,
557–564. Mandoli, A., Riet, F., Selloum, M., Zhou, H. et al. (2018)
59. Sugathan, A., Biagioli, M., Golzio, C., Erdin, S., Blumenthal, I., Increased H3K9 methylation and impaired expression of
Manawalan, P., Ragavendran, A., Brand, H., Lucente, D., Miles, protocadherins are associated with the cognitive dysfunc-
J., et al. (2014) CHD8 regulates neurodevelopmental path- tions of the Kleefstra syndrome. Nucleic Acids Res., 46,
ways associated with autism spectrum disorder in neural 4950–4965.
progenitors. Proc. Natl. Acad. Sci., 111, E4468–E4477. 74. Qureshi, I.A., Gokhan, S. and Mehler, M.F. (2010) REST
60. Scandaglia, M., Lopez-Atalaya, J.P., Medrano-Fernandez, A., and CoREST are transcriptional and epigenetic regu-
Lopez-Cascales, M.T., Del Blanco, B., Lipinski, M., Benito, E., lators of seminal neural fate decisions. Cell Cycle, 9,
Olivares, R., Iwase, S., Shi, Y. and Barco, A. (2016) When size 4477–4486.
matters: CHD8 in autism. Nat. Neurosci., 19, 1430–1432. 75. Lyons, D.B., Magklara, A., Goh, T., Sampath, S.C., Schaefer, A.,
61. Chen, W.G. (2003) Derepression of BDNF transcription Schotta, G. and Lomvardas, S. (2014) Heterochromatin-
involves calcium-dependent phosphorylation of MeCP2. Sci- mediated gene silencing facilitates the diversification of
ence, 302, 885–889. olfactory neurons. Cell Rep., 9, 884–892.
62. Scandaglia, M., Lopez-Atalaya, J.P., Medrano-Fernandez, A. 76. Frega, M., Selten, M., Mossink, B., Keller, J.M., Linda, K.,
et al. (2017) Loss of Kdm5c causes spurious transcription and Moerschen, R., Qu, J., Koerner, P., Jansen, S., Oudakker, A.
prevents the fine-tuning of activity-regulated enhancers in et al. (2020) Distinct pathogenic genes causing intellectual
neurons. Cell Rep., 21, 47–59. disability and autism exhibit a common neuronal network
63. Poeta, L., Padula, A., Attianese, B., Valentino, M., Verillo, L., hyperactivity phenotype. Cell Rep., 30, 173–185.e7.
Filosa, S., Shoubridhe, C., Barra, A., Schwartz, C.E., Chris- 77. Rubenstein, J.L.R. and Merzenich, M.M. (2003) Model of
tensen, J. et al. (2019) Histone demethylase KDM5C is a SAHA- autism: increased ratio of excitation/inhibition in key neural
sensitive central hub at the crossroads of transcriptional systems. Genes Brain Behav., 2, 255–267.
axes involved in multiple neurodevelopmental disorders. 78. Selten, M., van Bokhoven, H. and Nadif Kasri, N. (2018)
Hum. Mol. Genet., 28, 4089–4102. Inhibitory control of the excitatory/inhibitory balance in
64. Schaefer, A., Sampath, S.C., Intrator, A., Min, A., Gertler, T.S., psychiatric disorders. F1000Res., 7, 23.
Surmeier, D.J., Tarakhivsky, A. and Greengard, P. (2009) Con- 79. Mullins, C., Fishell, G. and Tsien, R.W. (2016) Unifying views
trol of cognition and adaptive behavior by the GLP/G9a of autism spectrum disorders: a consideration of autoregu-
epigenetic suppressor complex. Neuron, 64, 678–691. latory feedback loops. Neuron, 89, 1131–1156.
65. Voineagu, I., Wang, X., Johnston, P., Lowe, J.K., Tian, Y., Hor- 80. Nelson, S.B. and Valakh, V. (2015) Excitatory/inhibitory bal-
vath, S., Mill, J., Cantor, R.M., Blencowe, B.J. and Geschwind, ance and circuit homeostasis in autism spectrum disorders.
D.H. (2011) Transcriptomic analysis of autistic brain reveals Neuron, 87, 684–698.
convergent molecular pathology. Nature, 474, 380–386. 81. Morello, N., Schina, R., Pilotto, F., Phillips, M., Melani, R.,
66. Wang, P., Zhao, D., Lachman, H.M. et al. (2018) Enriched Plicato, O., Pizzorusso, T., Pozzo-Miller, L. and Giustetto, M.
expression of genes associated with autism spectrum dis- (2018) Loss of Mecp2 causes atypical synaptic and molecular
orders in human inhibitory neurons. Transl. Psychiatry, 8, plasticity of parvalbumin-expressing interneurons reflect-
13. ing Rett syndrome-like sensorimotor defects. eNeuro, 5,
67. Li, X., Zhang, Y., Wang, L., Lin, Y., Gao, Z., Zhan, X., Huang, Y., ENEURO.0086-18.2018.
Sun, C., Wang, D., Liang, S. and Wu, L. (2019) Integrated anal- 82. Smith, A.L., Jung, E.-M., Jeon, B.T. and Kim, W.Y. (2020)
ysis of brain transcriptome reveals convergent molecular Arid1b haploinsufficiency in parvalbumin- or somatostatin-
pathways in autism spectrum disorder. Front. Psych., 10, 1–8. expressing interneurons leads to distinct ASD-like and ID-
68. Griesi-Oliveira, K., Fogo, M.S., Pinto, B.G.G., Alves, A.Y., Suzuki, like behavior. Sci. Rep., 10, 7834.
A.M., Morales, A.G., Ezquina, S., Sosa, O.J., Sutton, G.J., Senaga- 83. Durand, S., Patrizi, A., Quast, K.B., Hachigian, L., Pavlyuk,
Franze, D.Y. et al. (2020) Transcriptome of iPSC-derived neu- R., Saxena, A., Carninci, P., Hensch, T.K. and Fagiolino, M.
ronal cells reveals a module of co-expressed genes con- (2012) NMDA receptor regulation prevents regression of
sistently associated with autism spectrum disorder. Mol. visual cortical function in the absence of Mecp2. Neuron, 76,
Psychiatry, (online ahead of print). 1078–1090.
69. Barnard, R.A., Pomaville, M.B. and O’Roak, B.J. (2015) Muta- 84. Patrizi, A., Picard, N., Simon, A.J., Gunner, G., Centofante, E.,
tions and modeling of the chromatin remodeler CHD8 define Andrews, N.A. and Fagiolini, M. (2016) Chronic adminis-
an emerging autism etiology. Front. Neurosci., 9, 1–13. tration of the N-methyl-D-aspartate receptor antagonist
70. Benevento, M., Iacono, G., Selten, M., Ba, W., Oudakker, A., ketamine improves Rett syndrome phenotype. Biol. Psychi-
Frega, M., Keller, J., Mancini, R., Lewerissa, E., Kleefstra, T. et al. atry, 79, 755–764.
(2016) Histone methylation by the Kleefstra syndrome pro- 85. Frega, M., Linda, K., Keller, J.M., Gumus-Aksay, G., van Rhijin,
tein EHMT1 mediates homeostatic synaptic scaling. Neuron, J., Negwer, M., Gunneweik, T.K., Foreman, K., Kompier, N.,
91, 341–355. Schoenmaker, C. et al. (2019) Neuronal network dysfunction
71. Ding, N., Zhou, H., Esteve, P.O., Chin, H.G., Kim, S., Xu, X., in a model for Kleefstra syndrome mediated by enhanced
Joseph, S.M., Friez, M.J., Schwartz, C.E., Pradhan, S. and Boyer, NMDAR signaling. Nat. Commun., 10, 4928.
T.G. (2008) Mediator links epigenetic silencing of neuronal 86. Reijnders, M.R.F., Kousi, M., Van Woerden, G.M., Klein, M.,
gene expression with X-linked mental retardation. Mol. Cell, Bralten, J., Mancini, G.M.S., Van Essen, T., Proietti-Onori, M.,
31, 347–359. Smeets, E.E.J., Van Gastel, M. et al. (2017) Variation in a range
72. Tahiliani, M., Mei, P., Fang, R., Leonor, T., Rutenberg, M., of mTOR-related genes associates with intracranial volume
Shimizu, F., Li, J., Rao, A. and Shi, Y. (2007) The histone and intellectual disability. Nat. Commun., 8, 1–12.