University of Groningen
Down & Alzheimer
Dekker, Alain Daniel
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2017
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Dekker, A. D. (2017). Down & Alzheimer: Behavioural biomarkers of a forced marriage. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Epigenetics: the Neglected Key to
Minimize Learning and Memory Deficits
in Down Syndrome
Alain D. Dekkera,b – Peter P. De Deyna,b
and Marianne G. Rotsa
a University of Groningen and University Medical Center Groningen b Institute Born-Bunge, University of Antwerp
Originally published in
Neuroscience and Biobehavioral Reviews (Elsevier), 2014, 45C: 72–84 Updated version published as book chapter in
Neuropsychiatric Disorders and Epigenetics (Academic Press, Elsevier), 2017, 163–184 edited by Yasui, Peedicayil and Grayson
Abstract
Down syndrome (DS), or trisomy 21, is the most common genetic intellectual disability. Although the triplication of chromosome 21 (HSA21) would theoretically lead to a 1.5 fold increase in gene transcription, transcript levels of many genes significantly deviate from this. Epigenetic mechanisms, including DNA methylation and post-translational histone modifications, regulate gene expression and mounting evidence indicates epigenetics to play an important role in learning, memory and intellectual disabilities. Surprisingly, epigenetic marks have been hardly investigated in DS. Importantly, various overexpressed HSA21 proteins affect epigenetic mechanisms and DS individuals are thus likely to present epigenetic aberrations. Excitingly, epigenetic marks are reversible, offering a huge therapeutic potential to alleviate or cure certain genetic deficits. Epigenetic therapy is already used for cancer, and may also provide new avenues for cognition-enhancing treatment in DS. This chapter summarizes the current knowledge on epigenetics in DS and discusses the potential of epigenetic therapy to reverse dysregulated gene expression.
9.1. Introduction:
Epigenetics has been largely neglected in Down syndrome
With an incidence of approximately 1 in 650-1000 live births, Down syndrome (DS) is the most common genetic cause of intellectual disability (Bittles et al., 2007). In 1866, the British physician John Langdon Down described various recurring symptoms of the ‘Mongolian type of idiocy’ that he observed among more than 10% of the children that he treated for cognitive impairment (Down, 1866). The cause of what became known as DS remained unclear for almost a century until the late nineteen fifties when Lejeune et al. discovered its origin: trisomy 21 (Lejeune et al., 1959). Over 95% of DS cases is a whole-chromosome trisomy due to meiotic non-disjunction, i.e. a failed separation of one of the paired chromosomes (Antonarakis et al., 2004; Lubec and Engidawork, 2002). The three copies of the human chromosome 21 (HSA21) lead to various complications, such as the characteristic facial appearance and the intellectual disability associated with impaired linguistic skills and diminished learning and memory capacities (Lott and Dierssen, 2010).
In addition to the congenital intellectual disability, individuals with DS face accelerated aging, including early-onset dementia due to Alzheimer’s disease (AD). By the time they reach 60-70 years of age, 50-70% of the DS population has developed AD compared to 11% of those aged 65+ in the general population (Alzheimer’s Assocation, 2015; Zigman and Lott, 2007). This strongly increased risk for AD in DS has been predominantly attributed to the triplication of the HSA21-encoded amyloid precursor protein (APP) gene. Consequently, this yields higher levels of APP protein and its secretase-splicing product amyloid-β (Aβ), the main constituent of the characteristic extraneuronal amyloid plaques in AD (Ness et al., 2012). Despite the fact that 95% of DS cases is due to a full trisomy 21, the DS population is characterized by an enormous variability in the type and the severity of clinical features (Roper and Reeves, 2006). This phenotypical variability is strikingly illustrated by the observation that the onset of clinical dementia symptoms in DS differs tremendously. Remarkably, 30-50% of the DS individuals do not develop dementia symptomatology, despite the full-blown AD-like neuropathology that is present in practically all DS individuals aged 40 years and older (Mann, 1988; Wisniewski et al., 1985; Zigman and Lott, 2007).
The complete DNA sequence of HSA21 was elucidated in 2000 (Hattori et al., 2000). Since then, many researchers have investigated the overexpressed protein-encoding genes and their effects on learning and memory. Despite increased understanding of the possible underlying genetic mechanisms, explaining the aforementioned variability among the DS population remains a scientific challenge (Jiang et al., 2013; Prandini et al., 2007). Although the triplication of HSA21 would theoretically lead to a 1.5 fold increase in gene transcription, gene expression studies suggested otherwise. For instance, analysis of HSA21 gene expression in DS lymphoblastoid cells showed that only 22% of the analysed genes had expression levels closely matching this level, compared to control individuals. In particular, 7% had an amplified expression (significantly higher than 1.5), 56% had an expression level that was significantly lower than 1.5, and 15% of the genes had highly variable expression profiles between subjects (Ait Yahya-Graison et al., 2007).
Similar results were obtained using the most widely used Ts65Dn mouse model of DS. Ts65Dn mice carry an additional chromosome, consisting of a duplicated part of the mouse chromosome 16 that is translocated to a small segment of the mouse chromosome 17 (Davisson et al., 1990). As a consequence, Ts65Dn mice are trisomic for about 50% of the genes on HSA21 (Reeves et al., 1995). However, it was demonstrated that many of these genes have transcript levels that significantly deviate from the theoretical 1.5 fold increase (Antonarakis et al., 2004; Kahlem et al., 2004; Lyle et al., 2004). For instance, Lyle et al. reported that not more than 37% of the genes in Ts65Dn matched the theoretical expression level of 1.5 (Lyle et al., 2004). Accordingly, certain genes are more dosage sensitive than others, thereby contributing in varying extents to the DS phenotypes (Antonarakis et al., 2004). Although various studies have tried to identify the crucial phenotype-determining genes, the underlying cause of the gene expression variation has been largely neglected.
Conceivably, epigenetic (epi = above (Greek)) mechanisms play a role in gene expression regulation and as such might play a crucial role in the development of the cognitive deficits in DS. An epigenetic trait is defined as “a stably heritable phenotype resulting from changes in a chromosome without alterations in the DNA sequence” (Berger et al., 2009). That is, epigenetic mechanisms, including DNA methylation, post-translational histone modifications and histone core variants, regulate gene expression without affecting the DNA itself. Importantly, epigenetic marks are reversible and thus offer a huge therapeutic potential to alleviate or cure certain genetic deficits.
As described in previous chapters, an increasing body of evidence illustrates the role of epigenetic mechanisms in synaptic plasticity, learning and memory and intellectual disabilities. Surprisingly, epigenetic mechanisms have been hardly investigated in DS. Most DS studies have focused on genomic aspects, neglecting the mounting evidence that demonstrates the contribution of epigenetics to impaired learning and memory. Importantly, epigenetic therapy is already in use for cancer, which may provide novel possibilities for cognition-enhancing treatment in DS as well. To our knowledge, no studies so far have investigated epigenetic therapy in mouse models of DS. Classical pharmacological treatment has not been successful yet in diminishing cognitive deficits in DS (Braudeau et al., 2011). Epigenetic therapy offers potentially important new avenues. This chapter summarizes and evaluates the limited knowledge on epigenetics in the neurobiology of DS, and discusses the huge potential of epigenetic therapy to reverse dysregulated gene expression.
9.2. Epigenetic mechanisms – an overview
To enable organized storage of all DNA in the nucleus, DNA in eukaryotic cells is packaged about 10.000 times into a more compact form: chromatin. The basic level of this chromatin is the nucleosome that consists of approximately 147 base pairs of DNA wrapped around a histone core in 1.7 turns. This core is an octamere, containing two copies of each histone type: histone 2A (H2A), H2B, H3 and H4 (Luger et al., 1997), as also shown in Figure 9.1. Considering the higher level of compaction, this string of nucleosomes (‘beads on a string’) is folded into a fiber, which in turn, is folded into even more condensed structures (Felsenfeld and Groudine, 2003). Epigenetic mechanisms
affect the nucleosomal packaging and thereby alter the accessibility of the DNA for molecular interactions that are required for transcription and replication (Cosgrove et al., 2004; Zentner and Henikoff, 2013). Two chromatin states are typically recognized: (i) accessible, relatively open euchromatin, which is generally associated with active gene expression and (ii) more inaccessible, tightly compacted heterochromatin that is generally associated with silenced gene expression (Figure 9.1). Three main epigenetic mechanisms are distinguished that alter this chromatin state: DNA methylation, post-translational histone modifications and nucleosomal core assembly.
Figure 9.1: Overview of the major epigenetic hallmarks and associated HSA21-products. DNA is compacted into chromatin, which consists of nucleosomes: 147 base pairs of DNA wrapped around an octamer histone core. Gene expression depends on the chromatin state: open, accessible chromatin (euchromatin) is associated with gene expression and closed, inaccessible chromatin (heterochromatin) is associated with gene silencing. DNMT, DNA methyltransferase; DSCR, Down Syndrome Critical Region; dsDNA, double-stranded DNA; HAT, histone acetyltransferase; HDAC, histone deacetylase; me, methylation; TET, ten-eleven translocation. Adapted from Falahi et al. (2014).
9.3. Epigenetic mechanisms affect learning and memory
In the human genome, cytosines preceding a guanine (CpG) are frequently methylated into 5-methylcytosine using the methyl group of S-adenosylmethionine (SAM) as donor (Weng et al., 2013). DNA methylation, generally associated with the formation of heterochromatin and repressed gene expression, is involved in the process of memory formation: increased DNA methylation of memory suppressor genes and diminished DNA
DS
CR
HAT
TET
‘Closed’ heterochromatin (repressed gene expression) ‘Open’ euchromatin
(active gene expression)
H2A H2B H3 H4 Trisomy 21 5’ 3’ 3’ 5’ me dsDNA acetylation methylation DNA methylation (CpG) DNA hydroxymethylation HSA21-products (possibly) affecting histone modifications:
DYRK1A RUNX1 HMGN-1 BRWD1 ETS2
HSA21-products (possibly) affecting DNA methylation:
DNMT3L CHAF1B CBS
HSA21-products (possibly) affecting histone variants:
H2AFZP H2BFS
T A C G T A T G C A
methylation of memory promoting genes (Day and Sweatt, 2010; Weng et al., 2013). The presence of high levels of DNA methyltransferases (DNMTs), as well as methyl-CpG-binding proteins in neurons suggests a role for DNA methylation in neuronal functioning (Sanchez-Mut et al., 2012). Indeed, the expression patterns of DNMTs change depending on the stage of neurodevelopment. Furthermore, Rett syndrome, a progressive neurodevelopmental disorder in females that results in intellectual disability, is caused by a single mutation in the methyl-CpG-binding protein 2 (MECP2) that recognizes methylated DNA (Amir et al., 1999). Moreover, neuronal plasticity was found to depend on neuronal activity-induced hydroxymethylation of certain critical genes (Ma et al., 2009). Hydroxymethylation, an early intermediate of DNA demethylation, is mediated by the ten-eleven translocation (TET) protein family (Tahiliani et al., 2009). In fact, a recent study showed that Tet1 knock-out mice presented deficits in hippocampal neurogenesis and impaired learning and memory (Zhang et al., 2013).
Next to DNA methylation, post-translational histone modifications also relate to learning and memory. Whether a certain (combination of) histone mark(s) is associated with stimulated or repressed gene transcription depends on the type and the position of a modification and the presence of particular effector proteins (Zentner and Henikoff, 2013). Histone-modifying enzymes establish (writers) or remove (erasers) particular histone marks, e.g. acetylation (generally associated with gene expression) is increased by histone acetyltransferases (HATs) and reduced by histone deacetylases (HDACs). Acetylation has been strongly associated with promoting synaptic plasticity and memory formation, while histone deacetylation was associated with memory deficits (Gräff et al., 2013). Deregulated acetylation of H4K12 has been related to memory impairment in aged mice (16 months), which was overcome by administration of HDAC inhibitors (Peleg et al., 2010). In agreement, HATs are crucial for memory formation, as is demonstrated by the intellectual disability in Rubinstein-Taybi syndrome that is caused by a loss of function mutation in the CBP/P300 HAT (Barrett and Wood, 2008). Furthermore, histone lysine methylation is associated with learning-dependent synaptic plasticity and hippocampus-dependent long-term memory formation (Jarome and Lubin, 2013).
Due to the genetic base pair mutations in epigenetic factors, Rett syndrome and Rubinstein-Taybi syndrome are classified as chromatin diseases (Berdasco and Esteller, 2013). Such mutations have not been documented in DS, but DS may be regarded, in part, as a chromatin disease as well. A growing body of evidence has, indeed, demonstrated that the triplication of HSA21, via the subsequent overexpression of various genes, directly dysregulates cellular epigenetic mechanisms in DS (Table 9.1). In turn, these disrupted epigenetic processes are associated with altered gene expression profiles and thus provide obvious candidates that might contribute to cognitive deficits in DS.
Despite the demonstrated involvement of epigenetics in learning and memory processes, only a few epigenetic studies have been conducted in DS. To the extent that it is known, the subsequent sections discuss the contribution of each of the three epigenetic mechanisms to cognitive deficits in DS, particularly focusing on epigenetic alterations due to any overexpressed HSA21 gene product (Table 9.1).
Ta bl e 9 .1 : A be rr an t e pi ge ne tic m ec ha ni sm s d ue to o ve re xp re ss ed H SA 21 -li nk ed p ro te in s i n D S. HS A2 1 pr od uc t Cl ass o f gen e exp res si on ǂ Pr im ar y f un ct io n Do w ns tr ea m ep igen et ic ef fec to r Ep igen et ic c on seq uen ce Ref er en ces DNMT 3L unk no w n DN A m et hylt ra ns fe ra se DNMT 3A , D NM T3 B DN A m et hy la tio n a nd his ton e d ea cet yla tion De pl us e t a l., 2 00 2; O oi e t a l., 2 00 7 CBS cl ass I hom ocys te in e co nv er sion SA M d ep let ion DN A m et hy la tio n a nd his ton e m et hyla tion In fa nt in o e t a l., 2 01 1; W al la ce a nd F an , 2 01 0 DY RK 1A cl ass I ki na se SI RT 1 ( HD AC) his ton e d ea cet yla tion Gu o e t a l., 2 01 0 CR EB a nd CB P/ P3 00 (HA T) his ton e a ce tyla tion Ba rr et t a nd W ood , 2 00 8; W eeb er a nd S w ea tt , 20 02 SW I/S N F c om pl ex hi st one m odi fic at io ns Lep ag nol -B es te l e t a l., 2 00 9 BR W D1 cl ass I V tr an scr ip tion al re gu la tor Hua ng e t a l., 2 00 3 RU N X1 cl ass I II tr an scr ip tion fa ct or Ba ks hi e t a l., 2 01 0 ET S2 cl ass I II CB P/ P3 00 (H AT ) his ton e a ce tyla tion Sun e t a l., 2 00 6 H2 AFZ P unk no w n hi st one v ar ia nt unk no w n unk no w n N CB I G en e, 2 01 6a ; S an ch ez -M ut e t a l., 2 01 2 H2 BF S cl ass I II Ga rdi ne r a nd Da vi sso n, 2 00 0; U ni Pr ot KB , 2 01 6 CHA F1 B cl ass I II con st itu tiv e ch rom at in pr ot ein m ult ip rot ein com plex w ith M BD1 a nd H P1 m et hyla tion -m ed ia te d tr an scr ip tion al re pr es sion Ree se e t a l., 2 00 3 HMGN1 cl ass I con st itu tiv e ch rom at in pr ot ein CB P/ P3 00 ( HA T) his ton e a ce tyla tion U ed a e t a l., 2 00 6 M ECP 2 act iv at ed or rep re ss ed gen e t ra ns cr ip tion Ab uh at zir a e t a l., 2 01 1 ǂ G ene e xpr essi on c la ssi fic at io n ba se d on Ai t Y ah ya -G ra ison et a l. ( 20 07 ). D ue to t he t rip lica tion of H SA 21 in D S, a n in cr ea sed e xp res sion lev el of 1 .5 fold is ex pe ct ed in DS com pa re d t o n on -D S con tr ols . F ou r cla ss es of g en es w er e r ep or ted w ith ex pr es sion le ve ls t ha t w er e a rou nd 1 .5 (cla ss I) , s ig nif ica nt ly h ig her t ha n 1 .5 (c la ss I I), si gni fic ant ly low er th an 1 .5 (cla ss II I) or h ig hly v ar ia ble ( cla ss IV ).
9.4. Altered DNA methylation is associated with DS
DNA methylation is conducted by four enzymes with different functions: DNMT1 maintains the methylation marks after DNA replication, DNMT3A and 3B are mainly responsible for de novo DNA methylation (Bestor, 2000; Margot et al., 2003), and DNMT3L has no methyltransferase activity, but mediates transcriptional gene repression (Deplus et al., 2002) and stimulates the methylation activity of DNMT3A and 3B by direct binding (Ooi et al., 2007; Suetake et al., 2004). DNMT3L is especially interesting in the context of DS, since it is encoded on HSA21 (Gardiner and Davisson, 2000). Previously, it was shown that female Dnmt3l knock-out mice presented specific hypomethylation of maternally imprinted genes, suggesting that DNMT3L together with DNMT3A/3B mediates de novo DNA methylation of these maternally imprinted genes (Arima et al., 2006; Bourc’his et al., 2001; Hata et al., 2002). Opposed to such a knock-out, overexpression of DNMT3L in DS likely affects DNA methylation patterns as well, potentially contributing to the cognitive deficits in DS.
Although DNA methylation patterns in DS have not been extensively investigated, studies have indicated that DNA methylation is different in DS individuals compared to the general population. To our knowledge, the first report on the presence of differential DNA methylation in DS was published in 2001, demonstrating increased genome-wide hypermethylation of lymphocyte DNA in DS children with a full trisomy 21, compared to their euploid siblings (Pogribna et al., 2001). In agreement, Chango et al. identified six DNA fragments that were hypermethylated in eight DS subjects compared to eight healthy controls. However, the applied technology did not allow for determination of the specific DNA sequence (Chango et al., 2006).
Then, in 2010, Kerkel et al. performed a high throughput screen for differentially methylated genes in DS using DNA that was extracted from total peripheral white blood cells and isolated T-lymphocytes (Kerkel et al., 2010). Compared to non-DS controls, a range of stable, gene-specific alterations in CpG methylation patterns was observed, which was independent of the differential cell counts. Strikingly, these genes were found on autosomes other than HSA21, indicating the influence of an additional copy of HSA21 on the epigenetic marks on other chromosomes. Many of the differentially methylated genes are involved in the development and functioning of white blood cells, which is supportive of the fact that DS is characterized by immune system deficiencies, amongst others, resulting in the high frequency of infections (Ram and Chinen, 2011).
Recently, Bacalini et al. confirmed most of these differentially methylated regions (DMRs) in another DS cohort. They studied DNA methylation profiles of peripheral white blood cells obtained from 29 DS subjects, as well as from their mothers and non-DS siblings to reduce the effect of genetic and environmental confounding factors. Correction for differential cell counts between the three study groups was implemented in the statistical analysis. Seven of the DMRs described by Kerkel et al. were also included: six of these probes were differentially methylated in the new DS cohort as well. Kerkel et al. reported DMRs on autosomes other than HSA21. Bacalini et al. also found a genome-wide distribution of DMRs, but the DMRs were especially enriched on HSA21. The reported DMRs primarily related to four functions: embryonic development, neuronal development, haematopoiesis (including the runt-related transcription factor 1 gene
(RUNX1), discussed below) and chromatin modulation (including TET1 and KDM2B, encoding a histone lysine demethylase) (Bacalini et al., 2015). Once more, this illustrates the role of DNA methylation in (neuro)development, possibly contributing to the intellectual disability in DS.
A first association between differential DNA methylation and a measure of cognitive functioning in DS was established by Jones and co-workers in 2013. They examined DNA obtained from cheek swabs of ten adult DS individuals and ten age-matched, healthy controls. 3300 CpGs were reported with DNA methylation levels that differed more than 10% between both groups. In accordance with Kerkel et al. but in contrast to Bacalini et al. no enrichment on HSA21 was observed. Cognitive function was briefly assessed using the Dalton Brief Praxis test and subsequently correlated with the DNA methylation results. Five differentially methylated probes correlated with cognitive functioning, indicating the relation between cognitive impairment due to trisomy 21 and altered DNA methylation. Two of those probes were observed in the TSC2 gene, which has been associated with the tau neuropathology of AD – the second major cognitive deficit in DS (Jones et al., 2013).
Interestingly, reduced levels of the methyl donor SAM have been described in DS. In contrast to the aforementioned reports on hypermethylation in DS, this suggests a reduced cellular methylation capacity (Infantino et al., 2011; Pogribna et al., 2001). The decreased SAM levels are attributed to overexpression of the HSA21-encoded cystathionine β-synthase (CBS) in DS. CBS is a central enzyme in the one-carbon metabolism, catalysing the conversion of homocysteine into cystathionine. This is the first step in the trans-sulfuration pathway, which results in the synthesis of the antioxidant glutathione (Figure 9.2). As a consequence, less homocysteine is available for the methionine cycle in which homocysteine is converted to methionine, the precursor of SAM (Pogribna et al., 2001). Therefore, reduced SAM levels and the subsequently reduced methyl transfer to DNA is a likely mechanism underlying aberrant DNA methylation patterns in DS.
In addition to nuclear DNA methylation, SAM is also required for methylation of cytosines in mitochondrial DNA (mtDNA) by the mitochondrial DNMT1, the only catalytically active DNMT found in mitochondria so far (Shock et al., 2011; van der Wijst and Rots, 2015). Importantly, despite a more than 50% increased expression of the SAM carrier, which transports SAM into the mitochondrion, significantly decreased mitochondrial SAM levels were found in lymphoblastoid cells of DS individuals compared to controls (Infantino et al., 2011). Whereas previous studies reported hypermethylation of nuclear DNA in DS (Bacalini et al., 2015; Chango et al., 2006; Infantino et al., 2011; Pogribna et al., 2001), the opposite was demonstrated for mtDNA (Figure 9.2). Compared to age-matched controls, mtDNA is hypomethylated in DS, suggesting an impaired mitochondrial methylation capacity in DS that, in turn, could lead to mitochondrial dysfunction (Infantino et al., 2011). In fact, mitochondrial dysfunction has been convincingly demonstrated in DS, including reduced expression of genes encoding mitochondrial enzymes (Conti et al., 2007; Lee et al., 2003) and impaired ATP synthesis (Valenti et al., 2010). The latter is likely to affect many epigenetic processes that require
ATP as a substrate, including the ATP-dependent production of SAM for DNA methylation (Wallace and Fan, 2010).
Figure 9.2: Schematic illustration of the CBS-induced depletion of SAM and its effects on nuclear and mitochondrial DNA methylation. Simplified: CBS is encoded on HSA21 and therefore overexpressed in DS, leading to the increased conversion of homocysteine into cystathionine. Accordingly, less homocysteine is available for conversion into the SAM-precursor methionine, yielding decreased SAM levels and a subsequently reduced DNA methylation capacity. Although hypomethylated mtDNA was indeed observed (Infantino et al., 2011), various studies reported hypermethylated nuclear DNA in DS (Chango et al., 2006; Infantino et al., 2011; Pogribna et al., 2001). Interrupted lines indicate that intermediary components are omitted. CBS, cystathionine β-synthase; mtDNA, mitochondrial DNA; SAM, S-adenosylmethionine.
The case of altered DNA methylation in DS is reinforced by two studies that demonstrated global hypermethylation in placental villi samples derived from DS fetuses compared to normal villi, illustrating that epigenetic changes are already present in early development (Eckmann-Scholz et al., 2012; Jin et al., 2013). Indeed, DNA methylation provides a new diagnostic method for the detection of DS. In contrast to the commonly used invasive (and risky) sampling procedures to obtain fetal genetic material, a novel non-invasive prenatal testing (NIPT) method was developed a few years ago using cell-free fetal DNA in the maternal peripheral blood. This method analyses differences in DNA methylation of HSA21 regions between the mother and her child and is based on the occurrence of fetal-specific DMRs on HSA21. Ratios of such DMRs for fetus over mother will be different for a trisomic child, which has an additional copy of the differentially methylated HSA21 locus, compared to a non-DS fetus. Comparing these methylation ratios for a combination of DMRs between normal and DS cases enabled correct non-invasive prenatal diagnosis of DS (Papageorgiou et al., 2011). More recently, two of these DMRs were validated as potential fetal-specific epigenetic markers (Lim et al., 2014).
To which extent DNA methylation patterns are regionally altered in the DS brain, particularly brain areas involved in learning and memory, remains to be elucidated. In that context, the hydroxylation of 5-methylcytosine into 5-hydroxymethylcytosine (5hmC), an early intermediate of DNA demethylation, might be very relevant: TET proteins mediate this hydroxylation and were shown to be differently methylated in DS and downregulated in the DS placenta, which possibly contributes to the aforementioned hypermethylated regions (Bacalini et al., 2015; Guo et al., 2011a, 2011b; Jin et al., 2013). Notably, 5hmC has
SAM Cystathionine Glutathione CBS Homocysteine Methionine DNMTs SAM carrier
SAM DNMT1 mtDNA methylation nuclear DNA methylation
Tr ans sul fur at io n pa th w ay M et hi oni ne cy cl e ↑ ↓ ↑ ↓ ↓ ↓ ↓ ↑ ↓ ↑
the highest prevalence in mature neurons compared to other mammalian cells and global 5hmC expression increases with aging, particularly at genes involved in neurological disorders, such as AD (Weng et al., 2013). The relationship between 5hmC and learning and memory is still in its infancy. However, its high expression in neurons and increased levels during aging and AD, seem to indicate a relevant role for 5hmC and TET.
In conclusion, increasing evidence indicates aberrant nuclear and mitochondrial DNA methylation in DS. Although DNA methylation is generally regarded as a stable epigenetic modification, recent findings have elucidated that intermediate DNA modifications are present, especially in neurons (Weng et al., 2013). The reversibility of DNA methylation in neurons might offer a great potential for treatment of neurological disorders, such as the cognitive deficits in DS.
9.5. Altered histone modifications are associated with DS
Post-translational histone modifications form docking platforms for chromatin-associated effector proteins that alter the chromatin structure, thus affecting gene expression. Such histone modifications are not static, but rather the equilibrium of continuous addition and removal of these chemical groups by epigenetic writers and erasers (Ueda et al., 2006). The chromatin structure is also affected by constitutive chromatin proteins and by incorporation of different histone core variants.
Post-translational histone tail modifications
Specific amino acids of histones are subjected to covalent post-translational modifications, such as acetylation, methylation and phosphorylation. Although these modifications have been reported to affect synaptic plasticity, learning and memory, only one study so far analysed a post-translational histone modification in DS, namely trimethylation of H3 lysine 4 (H3K4me3) (Letourneau et al., 2014). Letourneau et al. studied differential gene expression in fibroblasts from monozygotic twins discordant for trisomy 21. Such a pair of twins enables comparison of gene expression levels between trisomic and disomic cells without the noise of genomic variability. It was found that differential gene expression was distributed in defined domains along the chromosomes, so-called gene expression dysregulation domains (GEDDs). The authors suggested that the observed gene expression differences could be related to an altered chromatin state in the trisomic cells. Hence, DNA methylation and H3K4me3 were compared between the trisomic and disomic fibroblasts. Whereas differences in DNA methylation were not correlated to the GEDDs, the differences in H3K4me3 profiles between the twins markedly correlated with the GEDDs for nearly all chromosomes. Therefore, the altered H3K4me3-associated chromatin state relates to the altered gene expression in trisomic compared to disomic fibroblasts (Letourneau et al., 2014). Further direct proof for altered histone marks in DS is lacking. However, an increasing body of evidence suggests that post-translational histone modifications contribute to the neurological deficits observed in DS and other intellectual disabilities. So far, studies have found five HSA21 genes to influence particular histone modifications, namely DYRK1A, ETS2, HMGN1, BRWD1 and RUNX1, thus suggesting abnormal modifications in DS.
regulated kinase (DYRK) family, has been implicated in the learning deficits in DS (Smith et al., 1997). This highly conserved subfamily of protein kinases catalyzes auto-phosphorylation on tyrosine residues and auto-phosphorylation of serine/threonine residues on exogenous substrates (Becker and Joost, 1999). Studies in Drosophila and mice have revealed that the DYRK1A protein is necessary for normal brain development in a dose-sensitive way (Fotaki et al., 2002; Tejedor et al., 1995). Increased DYRK1A expression in primary murine cortical neurons reduced dendritic growth and complexity (Lepagnol-Bestel et al., 2009). Furthermore, transgenic mice that overexpressed DYRK1A showed significant impairment in cognitive flexibility and spatial learning. This indicates a causative role of DYRK1A overexpression in the intellectual disability associated with DS (Altafaj et al., 2001).
Interestingly, DYRK1A has a bipartite effect on histone modifications. First of all, DYRK1A directly phosphorylates the threonine residue 522 of the SIRT1 histone deacetylase, thereby promoting deacetylation and possibly deteriorating cognitive capacities (Guo et al., 2010). In addition, DYRK1A phosphorylates the cyclic AMP response element-binding protein (CREB) at the serine residue 133, inducing the recruitment of the CREB binding protein (CBP/P300). CBP/P300 is a histone acetyltransferase that promotes CREB-mediated expression of genes (Weeber and Sweatt, 2002; Yang et al., 2001). A loss of function mutation in CBP/P300 results in the aforementioned Rubinstein-Taybi syndrome that includes intellectual disability (Barrett and Wood, 2008; Bartholdi et al., 2007). Besides DYRK1A, two other HSA21 proteins influence the activity of CBP/P300: the erythroblastosis virus E26 oncogene homolog 2 (ETS2) and the nucleosome-binding high-mobility group N1 (HMGN1) (Sun et al., 2006; Ueda et al., 2006). Accordingly, it is conceivable that the HAT/HDAC balance is dysregulated in DS, causing aberrant histone acetylation patterns that affect learning and memory processes.
Apart from its direct phosphorylation effects, DYRK1A also alters gene expression via the neuron-restrictive silencer factor (NRSF, also known as REST). NRSF regulates the expression of a range of neuronal genes involved in the function of, amongst others, ion channels, neurotransmitter receptors and synapses (Canzonetta et al., 2008; Schoenherr and Anderson, 1995; Sun et al., 2005). NRSF represses transcription of these neuronal genes in non-neuronal cells by binding to the neuron-restrictive silencer element (NRSE). Besides non-neuronal cells, NRSF is also present in undifferentiated neuronal progenitors. However, it ceases to be expressed in differentiated neurons, thereby enabling gene expression. Therefore, NRSF has been termed as ‘a master negative regulator of neurogenesis’ (Schoenherr and Anderson, 1995).
In DS, NRSF levels seem to be perturbed. For instance, decreased NRSF expression was observed in the aforementioned placental villi samples from DS fetuses compared to non-DS (Jin et al., 2013). Moreover, various NRSF-regulated genes were repressed in neurospheres derived from fetal DS brain cells, whilst non-NRSF-regulated genes with similar functions were unaffected (Bahn et al., 2002). Furthermore, Canzonetta et al. reported a 30 to 60% reduced NRSF expression in the transchromosomic TgDyrk1A mouse model of DS, resulting in increased transcript levels of downstream targets (Canzonetta et al., 2008). This inverse correlation, however, was lost in another transgenic mouse model of DS that overexpressed DYRK1A (Lepagnol-Bestel et al., 2009). In
agreement, silencing the third copy of DYRK1A by RNA interference rescued NRSF levels, confirming the role of DYRK1A in NRSF-mediated gene regulation (Canzonetta et al., 2008). Importantly, DYRK1A regulates NRSF by binding to the SWI/SNF chromatin remodeling complex (Lepagnol-Bestel et al., 2009). This complex uses ATP to mobilize nucleosomes and rearrange the chromatin structure and induces the expression of multiple other genes involved in histone modifications, e.g. the histone methyl transferase L3MBTL2, the histone demethylase JARID1D and the HDAC interactor NCOR (Lepagnol-Bestel et al., 2009; Lu and Roberts, 2013; Sanchez-Mut et al., 2012). Therefore, the overexpression of DYRK1A in DS is likely to affect a range of epigenetic mechanisms, which strongly indicates that epigenetic marks are presumably altered in DS compared to the non-DS population.
The contribution of DYRK1A to learning and memory deficits in DS is further supported by findings from Altafaj et al. (2013). They administered short hairpin RNA against DYRK1A to Ts65Dn mice, resulting in normalized DYRK1A protein levels, improved synaptic plasticity and partial amelioration of the hippocampal-dependent search strategy in the Morris water maze. Another study showed that the DYRK1A inhibitor epigallocatechin-gallate (EGCG), a green tea flavonol, rescued visuospatial memory (Morris water maze) and object recognition memory (novel object recognition test) in both Ts65Dn and TgDyrk1A mice. A pilot study with EGCG-treatment (3 months) in young adults with DS, however, did not convincingly improve cognitive functioning: marginal positive effects were reported on visual memory recognition (p=0.04), working memory (p=0.08), and social functioning (p=0.05), as compared to placebo-treated subjects (De la Torre et al., 2014).
In addition to DYRK1A, two other HSA21-encoded proteins interact with the SWI/SNF complex, thereby altering histone modifications and likely gene expression: BRWD1 and RUNX1. The bromodomain and WD repeat-containing 1 (BRWD1) modulates the chromatin by binding through its two bromodomains and by associating with the SWI/SNF complex (Huang et al., 2003). Moreover, the RUNX1 forms multiprotein complexes at target gene promoters to which the SWI/SNF subunits BRG1 and INI1 bind. RUNX1 is associated with histone modifications that are typical of euchromatin, such as dimethylated H3K4 and acetylated H4 (Bakshi et al., 2010). As described in section 9.4, the RUNX1 gene was found to be hypermethylated in DS compared to controls (Bacalini et al., 2015; Eckmann-Scholz et al., 2012), suggesting altered RUNX1 gene expression. Consequently, altered RUNX1 protein levels likely affect epigenetic marks as well.
Unfortunately, the role of the SWI/SNF complex has not been investigated in DS yet. A growing body of evidence indicates the involvement of this chromatin-remodelling complex in neurodevelopment and hence might be critical in the cognitive deficits in DS. For instance, the expression of the SWI/SNF subunit BRG1 is enriched in the brain and the spinal cord of mice embryos (Randazzo et al., 1994) and the dorsal neural tube of chick embryos (Schofield et al., 1999). In zebrafish, Eroglu et al. revealed that BRG1-deficiency leads to impaired neurogenesis and neural crest cell differentiation (Eroglu et al., 2006). Furthermore, the contribution of the SWI/SNF complex to cognitive deficits is reinforced by the finding that the alpha-thalassemia X-linked intellectual disability (ATRX) syndrome is caused by mutations in the gene that encodes the SWI/SNF protein ATRX (Gibbons et
al., 1997; Villard et al., 1996). Therefore, aberrant functioning of the SWI/SNF complex due to overexpressed HSA21 products might lead to intellectual disability in a similar way.
Finally, post-translational histone modifications might be altered due to mitochondrial dysfunction in DS. Mitochondria are the major cellular source of high energy intermediates like acetyl-coenzyme A, nicotinamide adenine dinucleotide (NAD+), SAM and ATP, which are respectively involved in acetylation, deacetylation, methylation and phosphorylation of histones (Wallace and Fan, 2010). Aberrant mitochondrial production of these high energy intermediates in DS, for example due to the overexpression of CBS, is likely to cause alterations in post-translational histone marks in DS. Interestingly, a recent study revealed that incubation of DS lymphoblasts and fibroblasts with EGCG counteracted mitochondrial dysfunction. In particular, EGCG stimulated mitochondrial biogenesis and rescued ATP synthase catalytic activity and oxidative phosphorylation (Valenti et al., 2013), probably restoring the levels of one or more high energy intermediates. In addition to its inhibitory effect on DYRK1A, EGCG might thus improve learning and memory by rescuing mitochondrial functioning in DS.
Histone core variants and constitutive chromatin proteins
In addition to histone tail modifications, histone core variants can also influence gene expression. Incorporation of different histone variants into the nucleosomal core affects the chromatin structure (Luger et al., 2012). Most variants have been discovered for H2A and H3 and were reported to have a diverse role in gene expression regulation. Incorporation of macroH2A, for instance, is associated with gene repression (Creppe et al., 2012; Luger et al., 2012). HSA21 encodes the H2A histone family member Z pseudogene 1 (H2AFZP) and the H2B histone family member S pseudogene (H2BFS) (Gardiner and Davisson, 2000; NCBI Gene, 2016a, 2016b; Sanchez-Mut et al., 2012). DNA hyper-methylation of the H2BFS gene has been reported in three villi samples of DS fetuses compared to non-DS controls (Eckmann-Scholz et al., 2012). Whereas H2BFS has been described as a component of the nucleosomal core, it is currently unknown whether the H2AFZP encodes a protein (Sanchez-Mut et al., 2012; UniProtKB, 2016). The effects of both on gene regulation needs to be established as well.
Besides these histone variants, two HSA21-encoded constitutive chromatin proteins contribute to nucleosome assembly: chromatin assembly factor 1B (CHAF1B) and HMGN1. The CHAF1B protein is involved in nucleosome assembly on to newly replicated DNA by recruiting H3 and H4 (Kaufman et al., 1995; Verreault et al., 1996). CHAF1B forms a multiprotein complex with the methyl-CpG binding protein 1 (MBD1) and hetero-chromatin protein 1 (HP1), which, again, demonstrates the involvement of epigenetics in DS (Reese et al., 2003). HMGN1 affects post-translational histone modifications, in particular it inhibits phosphorylation of H3S10 and H3S28 and enhances H3K14 acetylation via CBP/P300 (Abuhatzira et al., 2011; Ueda et al., 2006), and has been described to regulate MECP2 expression. MECP2 is highly expressed in the brain and can activate or repress gene transcription (Brink et al., 2013). Altered MECP2 activity may result in intellectual disability (Abuhatzira et al., 2011; Samaco and Neul, 2011). Abuhatzira et al. (100) reported transcript levels of HMGN1 to be increased with 50% and transcript levels of MECP2 to be decreased with 30% in brain tissue of DS patients compared to non-DS
age-matched controls. In mice, it was found that altered HMGN1 protein levels resulted in histone modifications in the MECP2 promoter and a modified chromatin structure. Therefore, overexpressed HMGN1 might disturb normal learning and memory processes through reduced MECP2 protein expression and subsequently altered epigenetic marks.
9.6. Epigenetics in DS: a link to Alzheimer’s disease?
In addition to their intellectual disability, the cognitive capacities of people with DS are likely to deteriorate later in life due to a higher risk for early-onset AD. The triplication of the APP gene has been regarded as the main cause of this strongly increased risk. Post-mortem analysis revealed that virtually all DS individuals have an extensive AD-like neuropathology from 40 years of age (Mann, 1988). Despite the presence of extra-neuronal Aβ plaques and intraextra-neuronal neurofibrillary tangles of hyperphosphorylated tau protein, 30-50% of the DS individuals do not show signs of dementia, i.e. cognitive decline, impaired activities of daily living, and behavioural and psychological alterations (Dekker et al., 2015). In those who develop dementia, the age at which the first clinical symptoms appear varies greatly (Zigman and Lott, 2007). Accordingly, the central question emerges: what causes the one to get clinically demented and the other to live free of AD symptoms until death, while the AD-like neuropathology is present in all of them?
Since the discovery of trisomy 21, more than 50 years of genetic research has not yielded convincing answers yet. We, therefore, argue that solely focusing on the consequences of the overexpressed APP is too limited to explain this considerable inter-individual variability of AD in DS. Clearly, other factors determine why one DS inter-individual becomes demented and the other not. Which factors are essential contributors to AD in DS is currently far from understood. However, it is conceivable that different expression levels of these factors determine the presence or absence of clinical dementia symptoms. Again, epigenetic mechanisms affect gene expression, and are thus potential therapeutic targets to interfere with AD in DS. Despite the scarce epigenetic studies on AD in DS, an increasing body of evidence demonstrates altered nuclear epigenetic modifications in AD in the general population (an up-to-date review is provided in Bennett et al., 2015). Very recently, mtDNA methylation was implicated in AD as well (Blanch et al., 2016).
Importantly, various overexpressed HSA21 genes have been attributed a role in the progression of AD, possibly via their downstream epigenetic effects. Therefore, various epigenetic marks are likely altered in demented DS individuals as well. For instance, besides phosphorylating APP at threonine residue 668 (Ryoo et al., 2008), DYRK1A is also able to phosphorylate multiple sites of tau proteins, contributing to AD pathology as such. Indeed, these tau sites were found to be hyperphosphorylated in adult DS brains (Liu et al., 2008). Whether DYRK1A-mediated histone (de)acetylation and the SWI/SNF complex contributes to this is currently unknown. Furthermore, various studies show altered expression of microRNAs in AD (Tan et al., 2013), e.g. microRNA-125b is increased in the AD brain. In addition to the function of microRNAs in mRNA cleavage and degradation and translational repression (Morris, 2009; van den Berg et al., 2008), endogenously expressed microRNAs have been reported to mediate the condensation of heterochromatin (Farazi et al., 2008; Mercer and Mattick, 2013; Morris, 2009). Interestingly, microRNA-125b2 is located on HSA21 and thus overexpressed in DS (Lukiw,
2007). Therefore, disturbed microRNA-125b2 levels in DS might relate to the increased risk for AD in DS.
This high risk for AD is part of the so-called accelerated aging in DS (Zigman, 2013). Recently, Horvath’s group (Horvath, 2013) used a new molecular marker of aging, referred to as the ‘epigenetic clock’, to study human aging. The ‘epigenetic clock’ was defined by measuring DNA methylation levels in different human tissues in various age ranges. Consequently, a specific set of 353 CpGs was selected to constitute the ‘epigenetic clock’. The authors suggested that DNA methylation levels of these selected CpGs can predict, or estimate, an individual’s chronological age. This estimated age is referred to as the DNA methylation age. In non-DS individuals, the DNA methylation age is strongly correlated with the chronological age in multiple cell types and tissues, including neurons and glial cells (Horvath, 2013).
To assess whether the ‘epigenetic clock’ confirms the accelerated aging due to trisomy 21, DNA methylation datasets were analysed, including blood samples and various brain areas (Horvath et al., 2015). Again, the DNA methylation age and chronological age were strongly correlated in non-DS control individuals, establishing a reference regression line. For DS samples, the DNA methylation age tended to be higher than this regression line, suggesting an accelerated aging effect. In the analysis of brain tissue samples, the authors included AD patients without DS (mean age = 60.1 (58-64) years) in addition to DS individuals (mean age = 49.6 (42-57) years, no information provided on clinical dementia symptoms) and controls (mean age = 50.5 (32-64) years). Considering the overall brain samples, DS individuals showed a significantly higher accelerated aging effect than AD patients, despite their lower age. Similar results were obtained in focused analyses of frontal lobe and cerebellar samples (Horvath et al., 2015). As described before, extensive AD-like neuropathology is present in DS from the age of 40 (Mann, 1988), decades earlier than in the general population. Regarding the age range of the DS group, all DS individuals most likely presented this neuropathology. Conceivably, the significantly higher age acceleration in DS brain tissue samples reflects the early deposition of neuropathology in the DS brain.
In short, whereas AD in DS has hardly received attention in epigenetic studies, altered epigenetic signatures have been reported for accelerated aging in DS, and for AD patients in the general population. Consequently, the essential next step should be to epigenetically study AD in DS. Proper documentation on clinical dementia symptoms greatly matters. Since all DS individuals aged 40 years and older present the AD-like neuropathology, we need to distinguish between clinically demented and clinically non-demented DS individuals in order to study possible mechanisms causing the one to get demented at a relatively young age, while another grows old without any symptoms.
9.7. Epigenetic therapy may alleviate cognitive deficits in DS
Currently, no treatment is available to prevent or alleviate the intellectual disability or AD dementia in DS, though education and a stimulating environment may (slightly) improve cognitive capacities. Various studies have tried to develop pharmacological treatments to improve cognition in DS. In DS mouse models, such as the widely used Ts65Dn mice, promising cognitive-enhancing results have been described (Wiseman et al., 2009).
However, none of the investigated pharmacological treatments for DS patients have reached the market. Moreover, almost all clinical trials with drugs for AD in the general population have failed so far. The US Food and Drug Administration (FDA) has approved symptomatic AD treatment with cholinesterase inhibitors, such as donepezil. However, these drugs only provide short-term relief by improving cognition and to a certain, rather limited extent behaviour, but do not interfere with the underlying neurodegeneration (Mangialasche et al., 2010). Interestingly, a 10-week administration of donepezil (2.5-10 mg/day) to DS children 10-17 years of age failed to show cognitive improvement (Kishnani et al., 2010). To our knowledge, other AD drugs, such as those targeting the accumulation of Aβ, did not receive a market approval or are still in the process of clinical trials (Mangialasche et al., 2010; US National Institutes of Health, 2016).
Hence, investigating new targets and therapies is important. Here, we have pointed out the major role that epigenetics plays in synaptic plasticity and learning and memory. Various aberrant epigenetic modifications contribute to intellectual disabilities, and hence, might be involved in the cognitive deficits in DS. In contrast to the triplication of HSA21, epigenetic marks are reversible, and thus ideal targets to alleviate certain features in DS. Therefore, drugs that inhibit epigenetic enzymes, so-called epidrugs, offer a promising new way of treatment as an alternative for, or to act synergistically with, classical pharmacology.
The field of epigenetics is booming, especially in cancer research, and provides new approaches to address a wide variety of diseases. Currently, the FDA has approved five epidrugs against cancer (Table 9.2): two DNMT inhibitors (5-azacytidine (Vidaza) and decitabine (Dacogen)) and three HDAC inhibitors (vorinostat (Zolinza), romidepsin (Istodax), and belinostat (Beleodaq)) (US Food Drug Administration, 2016; Valdespino and Valdespino, 2015). In addition, it was demonstrated that valproic acid, which is already used against epilepsy and bipolar disorders for years, has anti-cancer properties as a HDAC inhibitor, and thus can be considered an epidrug too (Papi et al., 2010). Previously, studies showed that HDAC inhibitors may rescue memory impairments, thereby arousing many researchers’ interest in histone acetylation and finding specific HDAC inhibitors to combat learning and memory deficits (for a comprehensive review see Gräff et al., 2013). A wide array of clinical trials of epidrugs, however, is ongoing for neurological disorders, or to investigate new indications for approved epidrugs.
To treat the cognitive deficits in DS, the aforementioned DYRK1A inhibitor EGCG has received substantial attention in clinical trials (e.g. de la Torre et al., 2016). Interestingly, cancer studies currently investigate EGCG as an epidrug, since it can inhibit DNMT activity. Indeed, re-expression of genes that were silenced through promoter methylation have been observed (Fang et al., 2003; Nandakumar et al., 2011). Furthermore, Ramakrishna et al. (2012) reported significantly lower expression levels of the presynaptic alpha-synuclein protein in brain tissue of DS individuals and Ts65Dn mice, and suggested that alpha-synuclein plays a key role in deficient synaptic activity. Subsequent EGCG treatment of Ts65Dn mice resulted – counterintuitively – in increased DNA methylation of the synuclein promotor, and increased expression of the alpha-synuclein protein in Ts65Dn mice (Ramakrishna et al., 2014). Despite the inconsistent effect of EGCG on DNA methylation, it has become evident that EGCG does not only affect
DYRK1A, but may also alter the methylation profile of DS individuals.
In spite of the promising results of the approved treatments (Table 9.2), epidrugs have genome-wide and non-chromatin effects, thereby altering a range of biological processes. Accordingly, specific targeting of genes or proteins in particular tissues is the major challenge. In combatting the cognitive deficits in DS, specific targeting of (a part of) the third copy of HSA21 in the brain is required without affecting peripheral epigenetic modifications. Recently, the complete third copy was silenced in vitro using the large non-coding RNA molecule X inactive specific transcript (XIST), which endogenously silences the second X-chromosome in females. Jiang et al. introduced an inducible XIST transgene into the DYRK1A locus in induced DS pluripotent stem cells. As a consequence, stable heterochromatin marks were observed (repressive histone marks and DNA methylation), leading to chromosome-wide transcriptional silencing. As confirmation it was shown that transcription of e.g. DYRK1A and APP was repressed (Jiang et al., 2013). Future studies should demonstrate if this method would be successful in in vivo models as well.
Table 9.2: Currently approved epidrugs Epidrug (active
ingredient) Trade name Class Indication approval * FDA approval ** EMA
5-Azacytidine Vidaza DNMT
inhibitor Myelodysplastic syndromes
19-05-2004 17-12-2008 Decitabine Dacogen 02-05-2006 20-09-2012 Vorinostat Zolinza HDAC inhibitor Cutaneous T-cell lymphoma 06-10-2006 Withdrawn Romidepsin Istodax Cutaneous T-cell lymphoma Peripheral T-cell lymphoma 05-11-2009 12-02-2013 Refused on
Belinostat Beleodaq Peripheral T-cell lymphoma 03-07-2014 - Valproic acid various Bipolar disorder Epilepsy approval dates Various approval dates Various * FDA, US Food and Drug Administration (US Food Drug Administration, 2016)
** EMA, European Medicines Agency (European Medicines Agency, 2016)
Such innovative approaches open new avenues to pinpoint pathways or genes underlying the DS phenotype. In this respect, the rapidly developing technology of epigenetic editing offers a targeted approach to modulate the expression of (combinations of) individual genes, such as DYRK1A and APP, thereby validating their role in DS. Epigenetic editing comprises the targeting of particular epigenetic enzymes (writers or erasers) to specific genes with the use of lab-engineered DNA binding domains that target the endogenous gene of interest (de Groote et al., 2012). Such engineered domains, including designer zinc finger proteins, transcription activation-like effectors (TALEs) and the CRISPR/dCas9 platform (Jurkowski et al., 2015), are subsequently fused to an epigenetic enzyme with desired properties, e.g. a certain DNMT (Rivenbark et al., 2012; Siddique et al., 2013), DNA demethylase (Chen et al., 2014; Maeder et al., 2013) or histone modifiers (Falahi et al., 2013; Heller et al., 2014; Konermann et al., 2013). As a consequence, epigenetic modifications at the target gene are actively overwritten, causing long-term modulation of gene expression, either stimulating or repressing its expression (de Groote et al., 2012;
Jurkowski et al., 2015). Interestingly, epigenetic editing has demonstrated its promise to specifically modulate the expression of one or more neuronal genes in vivo (Bustos et al., 2014; Heller et al., 2014; Konermann et al., 2013), opening new avenues to ameliorate the cognitive deficits in DS.
The question then is, which targets? In this review we have considered epigenetic alterations due to overexpression of certain HSA21 genes (summarized in Table 9.1). Obviously, these genes would serve as a first group of targets. However, it is conceivable that non-HSA21 products influence the epigenetic mechanisms in DS as well. Unfortunately, not more than a mere handful of studies have conducted epigenetic profiling in DS. Therefore, future studies should more comprehensively investigate epigenetic marks in DS, preferentially including subgroups with and without dementia and groups of different ages.
9.8. Conclusion
DS is the most common genetic intellectual disability. Despite promising results in DS mouse models and ongoing clinical trials, no (preventive) treatment for the two major cognitive hallmarks – intellectual disability and early-onset AD – is currently available. Findings in other intellectual disabilities, including Rett syndrome and Rubinstein-Taybi syndrome, have elucidated the role of epigenetics in synaptic plasticity and learning and memory. However, epigenetics in DS has been largely neglected so far.
DS is characterized by extensive inter-individual variability. The triplication of HSA21 would theoretically result in a 1.5 fold increased expression level, but transcript levels of various HSA21 genes deviate from this, thus differently contributing to the DS phenotypes. Conceivably, epigenetic mechanisms play a crucial role herein. Although the role of epigenetics in DS is currently far from understood, an increasing body of evidence indicates the involvement of DNA methylation, post-translational histone modifications and histone core variants in DS. As summarized in Table 9.1, various overexpressed HSA21 gene products are epigenetic modulators, thereby dysregulating epigenetic mechanisms in DS. In turn, these disturbed mechanisms might contribute to the observed learning and memory deficits.
Importantly, epigenetic marks are reversible, and thus may offer great therapeutic potential to prevent or improve the cognitive symptoms in DS. Most promising is the technique of epigenetic editing in which specific epigenetic enzymes are recruited to specific genes by means of a laboratory-engineered DNA binding domain. Thereupon, epigenetic modifications are actively overwritten, causing a potentially persisting modulation of gene expression. Partial repression of overexpressed HSA21 genes can yield physiological expression levels that might alleviate the cognitive deficits in DS.
In short, it has become clear that one cannot ignore the involvement of epigenetics in intellectual disabilities, including DS. Although the number of studies on epigenetics in DS is relatively limited, these studies indicate disturbed epigenetic mechanisms due to overexpression of multiple HSA21 genes. Regarding the aforementioned cognition-enhancing therapies, future studies should identify the aberrant epigenetic marks in DS compared to the general population and between DS
individuals with and without dementia, in order to subsequently overwrite these marks by using epigenetic editing.
References
Abuhatzira, L., Shamir, A., Schones, D.E., Schaffer, A.A., Bustin, M., 2011. The chromatin-binding protein HMGN1 regulates the expression of methyl CpG-binding protein 2 (MECP2) and affects the behavior of mice. J. Biol. Chem. 286, 42051–42062. Ait Yahya-Graison, E., Aubert, J., Dauphinot, L., Rivals, I., Prieur, M., Golfier, G., Rossier, J., Personnaz, L., Creau, N., Blehaut, H.,
Robin, S., Delabar, J.-M., Potier, M.-C., 2007. Classification of human chromosome 21 gene-expression variations in Down syndrome: impact on disease phenotypes. Am. J. Hum. Genet. 81, 475–491.
Altafaj, X., Dierssen, M., Baamonde, C., Marti, E., Visa, J., Guimera, J., Oset, M., Gonzalez, J.R., Florez, J., Fillat, C., Estivill, X., 2001. Neurodevelopmental delay, motor abnormalities and cognitive deficits in transgenic mice overexpressing Dyrk1A (minibrain), a murine model of Down’s syndrome. Hum. Mol. Genet. 10, 1915–1923.
Altafaj, X., Martin, E.D., Ortiz-Abalia, J., Valderrama, A., Lao-Peregrin, C., Dierssen, M., Fillat, C., 2013. Normalization of Dyrk1A expression by AAV2/1-shDyrk1A attenuates hippocampal-dependent defects in the Ts65Dn mouse model of Down syndrome. Neurobiol. Dis. 52, 117–127.
Alzheimer’s Assocation, 2015. 2015 Alzheimer’s disease facts and figures. Alzheimer’s Dement. 11, 332–384.
Amir, R.E., Van den Veyver, I.B., Wan, M., Tran, C.Q., Francke, U., 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.
Antonarakis, S.E., Lyle, R., Dermitzakis, E.T., Reymond, A., Deutsch, S., 2004. Chromosome 21 and down syndrome: from genomics to pathophysiology. Nat. Rev. 5, 725–738.
Arima, T., Hata, K., Tanaka, S., Kusumi, M., Li, E., Kato, K., Shiota, K., Sasaki, H., Wake, N., 2006. Loss of the maternal imprint in Dnmt3Lmat-/- mice leads to a differentiation defect in the extraembryonic tissue. Dev. Biol. 297, 361–373.
Bacalini, M.G., Gentilini, D., Boattini, A., Giampieri, E., Pirazzini, C., Giuliani, C., Fontanesi, E., Scurti, M., Remondini, D., Capri, M., Cocchi, G., Ghezzo, A., Del Rio, A., Luiselli, D., Vitale, G., Mari, D., Castellani, G., Fraga, M., Di Blasio, A.M., Salvioli, S., Franceschi, C., Garagnani, P., 2015. Identification of a DNA methylation signature in blood cells from persons with Down Syndrome. Aging (Albany. NY). 7, 82–96.
Bahn, S., Mimmack, M., Ryan, M., Caldwell, M.A., Jauniaux, E., Starkey, M., Svendsen, C.N., Emson, P., 2002. Neuronal target genes of the neuron-restrictive silencer factor in neurospheres derived from fetuses with Down’s syndrome: a gene expression study. Lancet 359, 310–315.
Bakshi, R., Hassan, M.Q., Pratap, J., Lian, J.B., Montecino, M.A., van Wijnen, A.J., Stein, J.L., Imbalzano, A.N., Stein, G.S., 2010. The human SWI/SNF complex associates with RUNX1 to control transcription of hematopoietic target genes. J. Cell. Physiol. 225, 569–576.
Barrett, R.M., Wood, M.A., 2008. Beyond transcription factors: the role of chromatin modifying enzymes in regulating transcription required for memory. Learn. Mem. 15, 460–467.
Bartholdi, D., Roelfsema, J.H., Papadia, F., Breuning, M.H., Niedrist, D., Hennekam, R.C., Schinzel, A., Peters, D.J., 2007. Genetic heterogeneity in Rubinstein-Taybi syndrome: delineation of the phenotype of the first patients carrying mutations in EP300. J. Med. Genet. 44, 327–333.
Becker, W., Joost, H.G., 1999. Structural and functional characteristics of Dyrk, a novel subfamily of protein kinases with dual specificity. Prog. Nucleic Acid Res. Mol. Biol. 62, 1–17.
Bennett, D.A., Yu, L., Yang, J., Srivastava, G.P., Aubin, C., De Jager, P.L., 2015. Epigenomics of Alzheimer’s disease. Transl. Res. 165, 200–220.
Berdasco, M., Esteller, M., 2013. Genetic syndromes caused by mutations in epigenetic genes. Hum. Genet. 132, 359–383. Berger, S.L., Kouzarides, T., Shiekhattar, R., Shilatifard, A., 2009. An operational definition of epigenetics. Genes Dev. 23, 781–
783.
Bestor, T.H., 2000. The DNA methyltransferases of mammals. Hum. Mol. Genet. 9, 2395–2402.
Bittles, A.H., Bower, C., Hussain, R., Glasson, E.J., 2007. The four ages of Down syndrome. Eur. J. Public Health 17, 221–225. Blanch, M., Mosquera, J.L., Ansoleaga, B., Ferrer, I., Barrachina, M., 2016. Altered Mitochondrial DNA Methylation Pattern in
Alzheimer Disease-Related Pathology and in Parkinson Disease. Am. J. Pathol. 186, 385–97.
Bourc’his, D., Xu, G.L., Lin, C.S., Bollman, B., Bestor, T.H., 2001. Dnmt3L and the establishment of maternal genomic imprints. Science 294, 2536–2539.
Braudeau, J., Dauphinot, L., Duchon, A., Loistron, A., Dodd, R.H., Herault, Y., Delatour, B., Potier, M.-C., 2011. Chronic Treatment with a Promnesiant GABA-A alpha5-Selective Inverse Agonist Increases Immediate Early Genes Expression during Memory Processing in Mice and Rectifies Their Expression Levels in a Down Syndrome Mouse Model. Adv. Pharmacol. Sci. 2011, 153218.
Brink, M.C., Piebes, D.G., de Groote, M.L., Luijsterburg, M.S., Casas-Delucchi, C.S., van Driel, R., Rots, M.G., Cardoso, M.C., Verschure, P.J., 2013. A role for MeCP2 in switching gene activity via chromatin unfolding and HP1gamma displacement. PLoS One 8, e69347.
Bustos, F.J., Varela-Nallar, L., Aguilar, R., Henriquez, B., Falahi, F., Rots, M.G., Montecino, M., van Zundert, B., 2014. Epigenetic editing of the PSD95 gene promoter impacts neuronal architecture, in: Neuroscience 2014 - Society for Neuroscience. Washington DC.
Canzonetta, C., Mulligan, C., Deutsch, S., Ruf, S., O’Doherty, A., Lyle, R., Borel, C., Lin-Marq, N., Delom, F., Groet, J., Schnappauf, F., De Vita, S., Averill, S., Priestley, J. V, Martin, J.E., Shipley, J., Denyer, G., Epstein, C.J., Fillat, C., Estivill, X., Tybulewicz, V.L.J., Fisher, E.M., Antonarakis, S.E., Nizetic, D., 2008. DYRK1A-dosage imbalance perturbs NRSF/REST levels, deregulating pluripotency and embryonic stem cell fate in Down syndrome. Am. J. Hum. Genet. 83, 388–400.
Chango, A., Abdennebi-Najar, L., Tessier, F., Ferre, S., Do, S., Gueant, J.L., Nicolas, J.P., Willequet, F., 2006. Quantitative methylation-sensitive arbitrarily primed PCR method to determine differential genomic DNA methylation in Down Syndrome. Biochem. Biophys. Res. Commun. 349, 492–496.
Chen, H., Kazemier, H.G., de Groote, M.L., Ruiters, M.H., Xu, G.L., Rots, M.G., 2014. Induced DNA demethylation by targeting Ten-Eleven Translocation 2 to the human ICAM-1 promoter. Nucleic Acids Res. 42, 1563–1574.
Conti, A., Fabbrini, F., D’Agostino, P., Negri, R., Greco, D., Genesio, R., D’Armiento, M., Olla, C., Paladini, D., Zannini, M., Nitsch, L., 2007. Altered expression of mitochondrial and extracellular matrix genes in the heart of human fetuses with
chromosome 21 trisomy. BMC Genomics 8, 268.
Cosgrove, M.S., Boeke, J.D., Wolberger, C., 2004. Regulated nucleosome mobility and the histone code. Nat. Struct. Mol. Biol. 11, 1037–1043.
Creppe, C., Posavec, M., Douet, J., Buschbeck, M., 2012. MacroH2A in stem cells: a story beyond gene repression. Epigenomics 4, 221–227.
Davisson, M.T., Schmidt, C., Akeson, E.C., 1990. Segmental trisomy of murine chromosome 16: a new model system for studying Down syndrome. Prog. Clin. Biol. Res. 360, 263–280.
Day, J.J., Sweatt, J.D., 2010. DNA methylation and memory formation. Nat. Neurosci. 13, 1319–1323.
de Groote, M.L., Verschure, P.J., Rots, M.G., 2012. Epigenetic Editing: targeted rewriting of epigenetic marks to modulate expression of selected target genes. Nucleic Acids Res. 40, 10596–613.
de la Torre, R., de Sola, S., Hernandez, G., Farré, M., Pujol, J., Rodriguez, J., Espadaler, J.M., Langohr, K., Cuenca-Royo, A., Principe, A., Xicota, L., Janel, N., Catuara-Solarz, S., Sanchez-Benavides, G., Bléhaut, H., Dueñas-Espín, I., del Hoyo, L., Benejam, B., Blanco-Hinojo, L., Videla, S., Fitó, M., Delabar, J.-M., Dierssen, M., 2016. Safety and efficacy of cognitive training plus epigallocatechin-3-gallate in young adults with Down’s syndrome (TESDAD): a double-blind, randomised, placebo-controlled, phase 2 trial. Lancet Neurol. 15, 801–810.
De la Torre, R., De Sola, S., Pons, M., Duchon, A., de Lagran, M.M., Farré, M., Fitó, M., Benejam, B., Langohr, K., Rodriguez, J., Pujadas, M., Bizot, J.C., Cuenca, A., Janel, N., Catuara, S., Covas, M.I., Blehaut, H., Herault, Y., Delabar, J.-M., Dierssen, M., 2014. Epigallocatechin-3-gallate, a DYRK1A inhibitor, rescues cognitive deficits in Down syndrome mouse models and in humans. Mol. Nutr. Food Res. 58, 278–88.
Dekker, A.D., Strydom, A., Coppus, A.M.W., Nizetic, D., Vermeiren, Y., Naudé, P.J.W., Van Dam, D., Potier, M.-C., Fortea, J., De Deyn, P.P., 2015. Behavioural and psychological symptoms of dementia in Down syndrome: Early indicators of clinical Alzheimer’s disease? Cortex 73, 36–61.
Deplus, R., Brenner, C., Burgers, W.A., Putmans, P., Kouzarides, T., de Launoit, Y., Fuks, F., 2002. Dnmt3L is a transcriptional repressor that recruits histone deacetylase. Nucleic Acids Res. 30, 3831–3838.
Down, J.L.H., 1866. Observations on an Ethnic Classification of Idiots. London Hosp. Reports 3, 259–262.
Eckmann-Scholz, C., Bens, S., Kolarova, J., Schneppenheim, S., Caliebe, A., Heidemann, S., von Kaisenberg, C., Kautza, M., Jonat, W., Siebert, R., Ammerpohl, O., 2012. DNA-methylation profiling of fetal tissues reveals marked epigenetic differences between chorionic and amniotic samples. PLoS One 7, e39014.
Eroglu, B., Wang, G., Tu, N., Sun, X., Mivechi, N.F., 2006. Critical role of Brg1 member of the SWI/SNF chromatin remodeling complex during neurogenesis and neural crest induction in zebrafish. Dev. Dyn. 235, 2722–2735.
European Medicines Agency, 2016. European public assessment reports (EPAR) for human medicines [website]. URL http:// ema.europa.eu/ema/index.jsp?curl=pages/medicines/landing/epar_search.jsp&mid=WC0b01ac058001d124 (accessed 05.01.16).
Falahi, F., Huisman, C., Kazemier, H.G., van der Vlies, P., Kok, K., Hospers, G.A., Rots, M.G., 2013. Towards sustained silencing of HER2/neu in cancer by epigenetic editing. Mol. Cancer Res. 11, 1029–1039.
Falahi, F., van Kruchten, M., Martinet, N., Hospers, G.A.P., Rots, M.G., 2014. Current and upcoming approaches to exploit the reversibility of epigenetic mutations in breast cancer. Breast Cancer Res. 16, 412.
Fang, M.Z., Wang, Y., Ai, N., Hou, Z., Sun, Y., Lu, H., Welsh, W., Yang, C.S., 2003. Tea polyphenol (-)-epigallocatechin-3-gallate inhibits DNA methyltransferase and reactivates methylation-silenced genes in cancer cell lines. Cancer Res. 63, 7563– 7570.
Farazi, T.A., Juranek, S.A., Tuschl, T., 2008. The growing catalog of small RNAs and their association with distinct Argonaute/Piwi family members. Development 135, 1201–1214.
Felsenfeld, G., Groudine, M., 2003. Controlling the double helix. Nature 421, 448–453.
Fotaki, V., Dierssen, M., Alcantara, S., Martinez, S., Marti, E., Casas, C., Visa, J., Soriano, E., Estivill, X., Arbones, M.L., 2002. Dyrk1A haploinsufficiency affects viability and causes developmental delay and abnormal brain morphology in mice. Mol. Cell. Biol. 22, 6636–6647.
Gardiner, K.J., Davisson, M.T., 2000. The sequence of human chromosome 21 and implications for research into Down syndrome. Genome Biol. 1, 1–9.
Gibbons, R.J., Bachoo, S., Picketts, D.J., Aftimos, S., Asenbauer, B., Bergoffen, J., Berry, S.A., Dahl, N., Fryer, A., Keppler, K., Kurosawa, K., Levin, M.L., Masuno, M., Neri, G., Pierpont, M.E., Slaney, S.F., Higgs, D.R., 1997. Mutations in transcriptional regulator ATRX establish the functional significance of a PHD-like domain. Nat. Genet. 17, 146–148. Gräff, J., Tsai, L.-H.H., Graff, J., 2013. Histone acetylation: molecular mnemonics on the chromatin. Nat. Rev. Neurosci. 14, 97–
111.
Guo, J.U., Su, Y., Zhong, C., Ming, G.L., Song, H., 2011a. Emerging roles of TET proteins and 5-hydroxymethylcytosines in active DNA demethylation and beyond. Cell Cycle 10, 2662–2668.
Guo, J.U., Su, Y., Zhong, C., Ming, G.L., Song, H., 2011b. Hydroxylation of 5-methylcytosine by TET1 promotes active DNA demethylation in the adult brain. Cell 145, 423–434.
Guo, X., Williams, J.G., Schug, T.T., Li, X., 2010. DYRK1A and DYRK3 promote cell survival through phosphorylation and activation of SIRT1. J. Biol. Chem. 285, 13223–13232.
