Research Article Transcriptionally Repressive Chromatin...

Hindawi Publishing CorporationJournal of Nucleic AcidsVolume 2013 Article ID 567435 16 pageshttpdxdoiorg1011552013567435

Research ArticleTranscriptionally Repressive Chromatin Remodelling andCpG Methylation in the Presence of Expanded CTG-Repeats atthe DM1 Locus

Judith Rixt Brouwer12 Aline Huguet12 Annie Nicole12

Arnold Munnich12 and Geneviegraveve Gourdon12

1 Inserm U781 Hopital Broussais Batiment Leriche porte 9 96 rue Didot 75014 Paris France2Universite Paris Descartes-Sorbonne Paris Cite Institut Imagine 156 rue de Vaugirard 75015 Paris France

Correspondence should be addressed to Genevieve Gourdon genevievegourdoninsermfr

Received 28 August 2013 Accepted 22 October 2013

Academic Editor Emery H Bresnick

Copyright copy 2013 Judith Rixt Brouwer et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

An expanded CTG-repeat in the 31015840 UTR of the DMPK gene is responsible for myotonic dystrophy type I (DM1) Somatic andintergenerational instability cause the disease to become more severe during life and in subsequent generations Evidence isaccumulating that trinucleotide repeat instability and disease progression involve aberrant chromatin dynamics We explored thechromatin environment in relation to expanded CTG-repeat tracts in hearts from transgenic mice carrying the DM1 locus withdifferent repeat lengths Using bisulfite sequencing we detected abundant CpG methylation in the regions flanking the expandedCTG-repeat CpG methylation was postulated to affect CTCF binding but we found that CTCF binding is not affected by CTG-repeat length in our transgenicmiceWe detected significantly decreasedDMPK sense and SIX5 transcript expression levels inmicewith expanded CTG-repeats Expression of the DM1 antisense transcript was barely affected by CTG-repeat expansion In line withaltered gene expression ChIP studies revealed a locally less active chromatin conformation around the expanded CTG-repeatnamely decreased enrichment of active histone mark H3K914Ac and increased H3K9Me3 enrichment (repressive chromatinmark) We also observed binding of PCNA around the repeats a candidate that could launch chromatin remodelling cascadesat expanded repeats ultimately affecting gene transcription and repeat instability

1 Introduction

Over twenty unstable and expanded microsatellite repeatshave been identified as the cause of human neurologicaldisorders These repeats mostly consisting of trinucleotidesor tetranucleotides are considered dynamic mutations theypossess the unusual characteristic that repeat tract length isvariable Most microsatellite repeats show a normal rangeof relatively short and stable repeats and disease-causinglonger tracts that are often unstable Although differencesexist between diseases some molecular mechanisms overlapA longer repeat is typically associated with more clinicalproblems on top of earlier onset of symptoms Since muta-tion rate increases with repeat length successive generationsare faced with larger risks of developing more severe diseasea phenomenon called anticipation [1 2]

Myotonic dystrophy type I (DM1) is caused by anexpanded CTG-repeat in the 31015840 UTR of the DMPK gene thatis quite unstable when transmitted to the next generation [3ndash5] Myotonic dystrophy type I is a multisystem disorder withpatients showing not onlymuscle problems but also cataractcardiac anomalies testicular atrophy gastrointestinal andendocrine abnormalities as well as problems originating inthe central nervous system Ongoing somatic expansion inDM1 patients is thought to contribute to disease progression[6]

In addition to DM1 many other trinucleotide repeat(TNR) diseases are highly debilitating Efforts are thereforeaimed at understanding not only pathogenesis but alsomechanisms of TNR instability Thus far replication (bidi-rectional) transcription andDNA repair processes have beendescribed to play a role in TNR instability mechanisms [7ndash9]

2 Journal of Nucleic Acids

One of themajor pathogenicmodels proposed to underlieDM1 is a toxic effect of the presence of expanded CUG-containing transcripts Mutant DMPK mRNAs are retainedin the nucleus accumulate in foci [10] and form com-plexes with regulatory proteins thereby preventing theseproteins from exerting their normal function [11] AberrantmiRNA metabolism has also been described in patientswith expanded CTG-repeats [12 13] Recent evidence thatbidirectional transcription [14] and nonconventional RNAtranslation [15] are taking place at several TNR loci is compli-cating the traditional picture of RNA toxicity These findingspoint at a scenario where not just one single expanded RNAtranscript is responsible for disease development [11]

Moreover chromatin dynamics are increasingly recog-nised to influence both TNR instability and gene expressionat TNR loci and thereby probably disease outcome TNRscan affect nucleosome positioning [16 17] and CTG-repeatsspecifically have been identified as preferential location fornucleosome formation [18] CTG- and CAG-repeats havebeen described to form a functional component of insulatorelements thereby influencing gene expression levels At theDM1 locus the CTG-repeat forms an insulator together withthe two CTCF-binding sites (CTCFbs) that flank the repeat[19] Long CTG-tracts were shown to induce condensationof DNA at the DM1 locus which could hinder access ofgene regulators to this region [20] Transcription of theSIX5 gene that neighbours DMPK was decreased in patientcells expressing expanded CTG-repeats [21] These findingstogether led to the proposal that the expanded CTG-repeatinduces a transcriptionally repressive region [21] IndeedlongCTG-repetitionswere shown to induce heterochromatinformation which can then spread into neighbouring regions[22]

Aberrant chromatin remodelling has also been observedat other TNR disease loci For instance an expanded CGG-repeat is associated with CpG hypermethylation heterochro-matinisation and silencing of the FMR1 gene which is thecause of Fragile X syndrome [23] DNA methylation andhistonemodifications representative of silent chromatin havebeen observed around the expanded GAA-repeat in intron 1of the Frataxin gene that causes Friedreichrsquos ataxia [9]

Although the chromatin context is now consideredimportant for DM1 [14] and other TNR loci [9 23ndash25] fewcomprehensive studies addressing multiple factors involvedin or associated with chromatin remodelling simultaneouslyhave been performed Recently studies in mouse models forHuntingtonrsquos disease suggested that the chromatin contextof the transgene integration site determines CAG-repeatinstability and transcription levels [26]Thus these processesseem tightly linked underscoring the need to understandchromatin dynamics at TNR loci

We therefore set out to study the consequences ofCTG-repeat expansion for CpGmethylation CTCF-bindingchromatin conformation and gene expression at the DM1locus making use of the transgenic mouse model previouslygenerated in our laboratory [27] These mice carry a largehuman genomic transgene that encompasses the DMPKgene as well as the neighbouring genes DMWD and SIX5The transgene includes either a normal CTG-repeat of 20

trinucleotides or disease-associated expanded repeats withthe latter showing CTG-repeat instability patterns similar toDM1 patients (strongly biased towards expansions length-and age-dependent somatic instability albeit showing smallerrepeat length changes per instability event in mice) [28ndash30] The transgene also encompasses important regulatorysequences such as the two CTCF-binding sites (CTCFbs) thatflank the CTG-repeat in humans and the enhancer of thedownstream SIX5 gene

The phenotype of homozygous mice carrying up to 1600CTGs has recently been characterised [31] Since these micedisplay multiple characteristics seen in DM1 [27 28 3233] they are considered a valuable model to investigatemechanisms implicated in CTG-repeat instability and DM1pathogenesis [34 35] Benefitting from this DM1 mousemodel we aimed to study the epigenetic consequences of theexpanded CTG-repeat at theDM1 locusWe present evidencethat expanded CTG-repeats induce CpG methylation andlocal heterochromatinisation and concurrent decreased tran-scription around the repeat without affecting significantlyCTCF binding at the DM1 locus We also found binding ofPCNA around the CTG-repeat and propose that it mightlie at the basis of CTG-repeat expansion-induced repressivechanges in chromatin dynamics

2 Materials and Methods

21 Transgenic Mice Mice used in this study harbour atransgene consisting of 45 kb of human genomicDNA clonedfrom a DM1 patient and have been described previously(crossbred to gt90 C57BL6 background) [27 28 31] Micewere genotyped by PCR amplification of tail DNA usingoligonucleotide primers DMHR8 (51015840-TGACGTGGATGG-GCAAACTG-31015840) DMHR9 (51015840-AGCTTTGCACTTTGC-GAACC-31015840) and Dmm9 (51015840-GCTTGTAACTGATGGCTG-GG-31015840) which amplify the endogenous murine Dmpk(DMHR8 and Dmm9) and the human transgene DMPK(DmHR8 and DMHR9) [36] CTG-repeat length was deter-mined by PCR amplification of DNA extracted from tailat weaning with oligonucleotide primer 101 (51015840-CTTCCC-AGGCCTGCAGTTTGCCCATC-31015840) and primer 102 (51015840-GAACGGGGCTCGAAGGGTCTTGTAGC-31015840) as describedbefore [37] followed by electrophoresis of PCR products ona large 08 (wv) agarose gel [36] For the current paperheterozygous mice with the following CTG-repeat lengthswere used DM20 mice of the DM20-949 line that carrythe normal unexpanded human allele DM300 mice of theDM300-328 line that currently have an average of 610 CTGs(range 545ndash700) and DMSXL mice of the DM300-328line that have undergone large expansions and now carryalleles with over 1000 CTGs [28] In this particular studymice with sim1300ndash1600 CTG-repeats (mean 1435 CTGs)were used Adult mice of 3ndash5 months of age were used forall experiments described in this study (mean age did notdiffer among the different genotype groups as tested withKruskal-Wallis DM20 mean age 44 months DM300 meanage 43 months and DMSXL 41 months 119875 = 0634) Wechose heart as a representative tissue for disease which

Journal of Nucleic Acids 3

shows the highest DMPK expression levels in our mice [31]Hearts were dissected snap-frozen in liquid nitrogen andstored at minus80∘C until use

Animals were housed and cared for according to guide-lines by the French Council on Animal Care ldquoGuide for theCare and Uses of Laboratory Animalsrdquo EEC86609 CouncilDirectivemdashDecree 2001-131

22 Bisulfite Sequencing Methylation status of the sequencesflanking the CTG-repeat was studied by bisulfite conversionof DNA isolated from adult hearts (extracted with QiagenDNeasy Blood and Tissue Kit according to manufacturerrsquosinstructions)

500 ng of DNA was bisulfite-converted with QiagenrsquosEpitect Bisulfite kit Bisulfite converts an unmethylatedcytosine (C) into a thymine (T) while leaving methylated Csunchanged Subsequent PCR amplification and sequencingof the PCR product and comparison with the target (genomicDNA) sequence then allow distinction between Cs that wereor were not methylated at the time of bisulfite conversionPrimers were chosen so as not to contain any CpGssuch that DNA templates can be amplified irrespective oftheir methylation status Methylation interference studiesidentified guanine nucleotides whose methylation preventsbinding by CTCF predominantly on the noncoding strand[19] As most DNase I-hypersensitive sites induced by CTCFbinding are on this strand we conclude that the status of thisstrand is most relevant for our studiesTherefore we directedour CpG methylation studies at the non-CTG strandApproximate amplicon locations can be seen in Figure 2(a)Seminested PCR was performed with the following primersfor CTCFbs1 F (forward) 51015840-TAGTAGTAGTAGTATTTT-31015840 R1 (reverse) 51015840-TAGTAGTAGTAGTATTTT-31015840 and R2(for seminested PCR in combinationwith primer F) 51015840-CTT-TCCCTACTCCTATT-31015840 CTCFbs2 was amplified with F151015840-GTTTTGGGTAGATGGAGGGTT-31015840 R 51015840-AATCAC-AAACCATTTCTTTCT-31015840 and F2 51015840-GGTTTTAGGTGG-GGATAGATA-31015840 Three parallel seminested PCR reactionswere performed with 4 120583L of PCR product (total reactionvolume 25 120583L) of the first amplification as input with anannealing time 2∘C higher than the one used in the first PCRand 3 more cycles (30 and 33 cycles for subsequent PCRrounds) to obtain sufficient PCR product PCR productswere loaded onto an 15 (wv) agarose gel Productswere subsequently cut out and snap-frozen in liquid N

2

in columns of the Millipore DNA gel extraction kit Thesnap-frozen agarose band was spun down and DNA inthe flow-through was precipitated using classical NaCl andethanol precipitation These purified PCR products wereused to subclone into pMosBlue vector using the pMosBlueblunt-ended PCR cloning kit according to manufacturerrsquosinstructions (GE Healthcare) Colony PCR was performedwith primers T7 and U19 and correctly sized clones weresent for sequencing at the Sequencing Platform of CochinHospital in Paris France Sequences and CpG methylationof individual CpGs were subsequently analysed using BiQAnalyser software [38] Of each mouse at least 10 cloneswere sequenced An average percentage of methylation perCpG was calculated per mouse of which an overall weighted

average percentage methylation was calculated per CpG pergenotype

23 Chromatin Immunoprecipitation (ChIP) andQuantitative PCR (qPCR)

231 Chromatin Preparation Approximately 30mg of frozentissue was cut into small pieces 500120583L of cold PBS with1 formaldehyde was added Protein-DNA interactions werecross-linked for 10 minutes at room temperature (RT) whilerotating Fixation was quenched by adding glycine to afinal concentration of 0125M and rotating for 5 minutesat RT Samples were spun at sim470 g (2500 rpm in table topcentrifuge) for 10 minutes at 4∘C with slow decelerationThepellet waswashedwith cold PBS for 10minutes at 4∘C settingthe centrifuge to slowly decelerate Pellet was resuspendedvigorously vortexed in 150 120583L SDS lysis buffer (1 SDS10mM EDTA 50mM Tris pH 8 with PhosStop phos-phatase inhibitors (Roche) and complete protease inhibitors(Roche)) and left on ice for 15 minutes with repeatedvortexing The tissue samples were then sonicated (BransonSonifier cell disruptor B15) in lysis buffer until macroscop-ically homogenised while being kept cold Samples werespun down at 11000 rpm for 10 minutes at 4∘C (normaldeceleration from here onwards) 100120583L of supernatant wastransferred to a separate tube and kept on ice 100 120583L ofSDS lysis buffer was added to the pellet in which the pelletwas resuspended and vortexed These samples were snap-frozen in liquid nitrogen twice to aid further lysis Sampleswere centrifuged at 11000 rpm for 10 minutes at 4∘C Thesupernatant was taken and added to the first supernatantfraction The supernatant fraction was again sonicated toshear the chromatin obtaining fragments of 200ndash1000 bpThe sheared chromatin was diluted 5 times with ChIPdilution buffer (001 SDS 11 Triton X-100 12mM EDTA167mM Tris-HCl pH 81 and 167mM NaCl) supplementedwith phosphatase and protease inhibitors

232 Chromatin Immunoprecipitation (ChIP) 40 120583L ofDynal proteinADynabeads (Life technologies) per immuno-precipitation (IP) reaction was incubated with antibody atRT for one hour while rotating The following antibodieswere used for ChIP rabbit anti-CTCF (Abcam ab70303 2 120583gper IP) rabbit anti-H3K914Ac (Millipore 06-599 5 120583g)mouse anti-H3K27Me (Abcam ab6002 2 120583g) rabbit anti-H3K9Me3 (Millipore 07-442 3 120583g) and mouse anti-PCNA(Santa Cruz sc56 5120583g) 200120583L of 5x diluted chromatinwas taken per IP and subjected to a further 2x dilutionwith ChIP dilution buffer supplemented with phosphataseand protease inhibitors and added to the antibody-coveredDynabeads (the abChIP reaction) For each sample a controlIP was taken along using rabbit or mouse IgG (SantaCruz) as appropriate An aliquot of the same cross-linkedand sheared chromatin was kept aside and purified inparallel to be used as input chromatin control for PCRAntibody-covered beads were incubated with the chromatinovernight at 4∘C while rotating The following day thebeads were washed using a Dynal Dynamag-Spin Magnet


5998400DMPK promoter

Amplicons for qPCR after ChIPAmplicons for bisulfite sequencingCTG-repeat

CTCF binding sitesTSS transcription start site

Exons 1ndash15

TSS sense transcript

Q-bs1 Q-bs2CTCFbs1 37CpGs

3998400 untranslated region

CTCFbs2 30CpGsQ-Enh

sim4000gt50ndash2005ndash37

Congenital formAdult onset formUnaffected

TSS antisensetranscript

End of DMPK gene

TSS SIX5DNAse hyper-sensitive site

3998400

sim2000 bpReplication origin

(CTG)n

Figure 1 DM1 locus schematic drawing of the DM1 locus indicating relevant sites and regions including the expanded CTG-repeat withflanking CTCF-binding sites (CTCFbs) the DNAse hypersensitive site enhancer of SIX5 and transcription start sites (TSS) of the geneslocated at the DM1 locus Approximate locations of amplicons used for bisulfite sequencing and qPCR after ChIP are indicated The 45 kbfragment of genomic DNA that was used to generate the DM300 transgenic mouse line used in this study contains all of the features indicatedin this scheme

(Life technologies) for 4 minutes at 4∘C while rotatingtwice with ChIP dilution buffer 1x with low salt immunecomplex washing buffer (01 SDS 1 Triton X-100 2mMEDTA 20mM Tris-HCl pH 81 and 150mM NaCl) 2x withhigh salt immune complex washing buffer (01 SDS 1Triton X-100 2mM EDTA 20mM Tris-HCl pH 81 and500mM NaCl) 1 time with LiCl immune complex buffer(025M LiCl 1 Igepal-CA630 1 deoxycholic acid (sodiumsalt) 1mM EDTA 10mM Tris pH 81 Millipore 20-156)and twice in TE buffer (10mM Tris-HCl 1mM EDTA pH80) Beads were then taken up in 150 120583L complete elutionbuffer (20mM Tris-HCl pH 76 5mM EDTA 50mM NaCl1 SDS and 130 120583gmL proteinase K (Life technologies))and incubated at 67∘C for at least 4 hours while rotatingto elute the protein-DNA complexes off the beads reversecross-links and digest proteins simultaneously Eluates werethen collected and beads were rinsed with 75120583L elutionbuffer (20mM Tris-HCl pH 76 5mM EDTA and 50mMNaCl) Eluates were combined and purified with QiaquickDNA purification columns (Qiagen 28106) according toprotocol with the exception of the addition of 1 volume ofisopropanol in addition to 5 volumes of buffer PB in thefirst step to optimise isolation of small fragments DNA waseluted in two steps using twice 25120583L elution buffer providedwith the kit

233 Quantitative PCR (qPCR) Relative abundance attarget loci in the ChIPed DNA was analysed using quan-titativez PCR using Power SYBR green master mix (AppliedBiosystems) in an AB7300 real-time PCR system (AppliedBiosystems) For all antibodies used amplicons were meas-ured at CTCF binding site 1 (Q-bs1) CTCF binding site 2(Q-bs2) and the enhancer region (Q-Enh) as indicated inFigure 1 Primer sequences were as follows Q-bs1-Forward(F) 51015840-CTGCCAGTTCACAACCGCTC-31015840 Q-bs1-Reverse(R) 51015840-CGAGCCCCGTTCGCCG-31015840 Q-bs2-F 51015840-CGT-CCGTGTTCCATCCTC-31015840 Q-bs2-R 51015840-CGTCCGTGT-TCCATCCTC-31015840 Q-Enh-F 51015840-GGAGGCGTGTGGAGG-CGG-31015840 and Q-Enh-R 51015840-TCCCCCAACCCTGATTCG-31015840

The following amplicons were used as positive or nega-tive PCR controls for ChIP reactions Myc-F 51015840-CTTGTT-CTATTGCCTTTCCGTTTC-31015840 and Myc-R 51015840-AACCCA-TCCCTACTTTCTGACAGTC-31015840 (positive control forCTCF-ChIP) Gapdh-F 51015840-ATAAGCAGGGCGGGAGGC-31015840 and Gapdh-R 51015840-CGTCTCTGGAACAGGGAGGAG-31015840(positive control for active histone mark negative controlfor repressive histone modifications and CTCF) Amylase-F51015840-CTCCTTGTACGGGTTGGT-31015840 and Amylase-R 51015840-AAT-GATGTGCACAGCTGAA-31015840 (negative control for activehistone modification positive control for PCNA) andHOXd9-F 51015840-TGCTCCGGGGCTTTGGATAA-31015840 andHOXd9-R 51015840-CTCTCTGGGTCCTGCGATCT-31015840 (positivecontrol for repressive histone modifications) A standardcurve of serial dilutions of genomic DNA was always takenalong A dissociation curve was run in every experiment toassess quality of the reaction and ensure absence of primer-dimer or other nonspecific PCR products Making use ofthe formula derived of the standard curve quantities werecalculated from the obtained quantification cycle (119862

119902) values

for each sample Each samplewas performed in triplicate andquantities were averaged Quantities obtained in IP reactions(antibody-ChIP ldquoabChIPrdquo or IgG-mock IP ldquoIgG-IPrdquo) werenormalised by division by the quantity obtained in inputchromatin (non-IPed) samples (enrichment) Enrichmentswere normalised against the abChIP enrichment value ofthe positive control amplicon for each respective CTG-repeat length category to correct for possible differences inchromatin density All pairs of normalised IgG- and abChIP-enrichment values for a given genotype and amplicon weresubjected to a Mann-Whitney 119880 test to see if enrichment inabChIP was statistically significantly larger than in IgG-IPaccording to expectations (expected for positive controlamplicons but not for negative control amplicons) Theinput-corrected enrichment values of all specific abChIPsamples were further subjected to statistical analysis usingIBM SPSS Statistics Standard Edition version 20 Softwareto analyse possible effects of increasing CTG-repeat lengthGraphs and statistical tests were obtained with GraphPadPrism version 50c All effects are reported at a 005 level ofsignificance


5998400 Exon 15

CTCFbs1 361bp 37CpGs

CTCFbs1 (CTG)n

CTG-repeatAmplicons for bisulfite sequencing

CTCF binding sites

CTCFbs23998400


(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

DM20

DM300

DMSXL

0 7 5 0 0 8 2 3 10 0 0 0 0 0 0 7 0 0 0 5 0 5 5 5 8 3 0 8 0 0 5 5 2 0 6 9 18

56 29 37 49 61 37 22 25 29 25 43 33 45 27 17 20 35 31 36 45 41 28 20 33 73 4 65 6 17 7 41 28 26 35 26 47 46

89 70 79 83 95 70 52 61 66 57 69 65 80 49 48 56 66 50 83 79 84 78 67 86 82 26 70 28 42 23 53 45 67 76 48 80 90

CTCFbs2CpG nr

CpG nr

DM20

DM300

DMSXL

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

0 2 5 2 2 7 0 2 5 4 2 0 0 0 4 3 2 0 4 2 2 2 2 0 4 6 2 0 2 0

38 29 21 12 20 57 21 16 16 21 23 16 0 38 3 32 20 3 7 5 29 19 10 11 0 10 3 0 6 2

78 17 17 22 17 40 18 19 26 30 36 26 28 26 19 13 4 26 17 2 43 30 24 26 7 6 2 4 4 2

CpG methylation ()Nevergt0ndash249925ndash4999

(CTG)n

(CTG)n

50ndash749975ndash100

(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

CpG nrDM20

Nonmethylated CpGMethylated CpG

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

DM300

DMSXL

CTCFbs1

CTCFbs1(CTG)n

(CTG)n

(CTG)n

(c)

Figure 2 CpG methylation increases with CTG-repeat length (a) Schematic drawing of the position of the amplicons obtained withseminested PCR used for bisulfite sequencing relative to both CTCF-binding sites the CTG-repeat and exon 15 in the 31015840 region of theDMPK transgene (b) CpGmethylation in hearts of adult mice increases with expanding CTG-repeat length For each repeat length category4 mice were used and per mouse at least 10 clones were sequenced after bisulfite conversion CpGs are numbered from 51015840 to 31015840 Shaded CpGnumbers lie within the CTCF-binding site recognition sequence Underlined CpG numbers are CpG dinucleotides that contain G residuesessential for recognition by CTCF [19] Numbers indicate the weighted average percentage of methylation at a particular CpG seen across allclones in all 4mice Colour coding further indicates the approximate degree ofmethylation at a givenCpG (c) Example ofmethylation profilesobtained in different clones representing different cardiac cells of 1 mouse per repeat length category Individual clones show substantiallydistinct methylation patterns


24 Expression Studies

241 RNA Isolation Snap-frozen adult hearts taken fromDM20 DM300 and DMSXLmice were homogenized in Tri-zol (Life Technologies) using a tissue lyser (Retsch MM4002x 25 minutes using 2 stainless steel beads (Qiagen)) Afterchloroform extraction the aqueous phase was mixed with70 ethanol and transferred to a Spin Cartridge of the Pure-Link RNA Mini Kit (Ambion by Life Technologies) TotalRNAwas extracted according to manufacturerrsquos instructionsA PureLink DNAse step was inserted in the protocol asrecommended after binding of the RNA to the column RNAconcentrations were determined by absorbance at 260 nmusing a NanoDrop 1000 spectrophotometer (ThermoFisherScientific) and quality and absence of genomic DNA wereverified on agarose gel

Different RNA extraction methods have been comparedof which the one described here was found to give thehighest yield Efficacy of the various methods to recoverexpanded RNA was assessed by comparing the total RNArecovery and DMPK expression levels between hemizygousand homozygous mice [31]

242 Reverse Transcriptase (RT) PCR cDNA was synthe-sized from 04ndash06 120583g RNA using Superscript II ReverseTranscriptase (Life Technologies) when using random hex-amer primers (for SIX5 mRNA quantification) or usingSuperscript III Reverse Transcriptase (Life Technologies)when using strand-specific primers (for DMPK sense andantisensemRNAquantification) according tomanufacturerrsquosinstructions (also see [31]) cDNA was treated with RNAseA for 20 minutes at 37∘C Please refer to Table 1 for primersequences used for strand-specific RT

243 Quantitative RT-PCR (qRT-PCR) Relative abundanceof transcripts was analysed using Power SYBR green mastermix (Applied Biosystems) in an AB7300 real-time PCRsystem (Applied Biosystems) Annealing temperatures andsample dilutions were optimized for each amplicon DMPKSIX5 and antisense transcript levels were calculated rela-tive to 18s and endogenous murine Dmpk mRNA levelsOligonucleotide primer sequences are described in Table 1For qRT-PCR we used standard curves of serial dilutionsof a plasmid carrying the amplicon Reverse transcriptaseefficiency for each gene and each primer set was verified usingincreasing amounts of RNA as input A dissociation curvewas run in every experiment to assess quality of the reactionand ensure absence of primer-dimer or other nonspecificPRCR products Reverse transcriptase was performed induplo followed by separate qPCR analyses on each cDNAsample All qPCR reactions were performed in triplicate andexperiments (from RT reaction to qPCR analysis) were donetwice RNA from hearts of 8 DM20 5 DM300 and 6 DMSXLmice was used for these expression studies Averages oftriplicate quantities obtained for eachmousewere normalisedagainst a control sample that was taken along in every qPCRexperiment The average expression level of the two parallelqRT-PCR experiments was subjected to statistical analyses

Jonckheere Terpstra test for trend (IBM SPSS StatisticsStandard Edition version 20 Software) was performed toinvestigate whether expression levels change with increasingCTG-repeat length Differences between repeat length cate-gories were further investigated by means of non-parametricMann-Whitney pairwise comparisons Graphs and statisticaltests were obtained with GraphPad Prism version 50c Alleffects are reported at a 005 level of significance

3 Results

31 Elevated CpG Methylation with Increasing CTG-RepeatLength We studied CpG methylation around the CTG-repeat since DNA methylation can affect binding of tran-scription factors or attract chromatin-remodelling enzymesPrevious methylation analyses [39] had shown substantialCpG methylation in various tissues of DM300 mice inboth upstream and downstream regions flanking the CTG-repeat while DM20 tissues were almost completely devoidof CpG methylation CpG methylation analysis in DM1patient tissues showed a clearly polarised pattern withonly methylated Cs at and around CTCFbs1 and not atCTCFbs2 Substantial variability between individual patientscan be observed [39] We extended these observations byperforming a detailed analysis of the CpG methylationpattern by bisulfite sequencing in individual mice carryingnormal alleles with 20 CTGs (DM20) or mice expressingexpanded alleles (DM300 545ndash700 repeats or DMSXL 1300ndash1600 CTGs) We studied the region flanking the CTG-repeatas shown in Figure 1 We chose to study heart which showsthe highest DMPK expression levels in our mice [31] Heartabnormalities including arrhythmias and conduction defectsare a central feature to the disease Bisulfite sequencingwas directed at the non-CTG strand because a methylationinterference assay indicated that CTCF showed strongercontacts on this strand [19] Approximate amplicon locationsused in this study are indicated in Figure 2(a)

Figure 2(b) illustrates the results of 10 clones (represent-ing 10 different cells) of four mice for each repeat lengthcategory It shows that CpGmethylation is very low in DM20mice more prominent in DM300 mice and quite abundantin DMSXL mice Further expansion of the CTG-repeatfrom around 600 to sim1450 CTGs is associated with morepronounced CpG methylation This trend is seen aroundboth CTCFbs1 and CTCFbs2 although less pronouncedat CTCFbs2 Overall the region around CTCFbs1 is moremethylated than the region around CTCFbs2 Note thatCTCFbs1 itself is relatively spared from CpG methylation ascompared to its flanking sequences Figure 2(c) illustratingthe methylation pattern of each clone shows that the CpGmethylation pattern is variable in individual cells within atissue Other mice that we analysed showed a similar patternof distinct methylation patterns in individual cells (data notshown)

32 CTCF Still Binds CTCF-Binding Sites When CTG-RepeatIs Expanded Based on in vitro observations it had beenpostulated that binding of CTCF is lost when CpGs in


Table 1 List of primers used for expression analysis

DMPK senseprimer for RT CGACTGGAGCACGAGGACACTGACTTGCTCAGCAGTGTCAGCAGGTCCCCGCC

qPCR primers Forward CGACTGGAGCACGAGGACACTGAReverse GGAGAGGGACGTGTTG

DMPK antisenseprimer for RT CGACTGGAGCACGAGGACACTGAGACCATTTCTTTCTTTCGGCCAGGCTGAGGC

qPCR primers Forward GGAGCACGAGGACACTGAReverse TGCGAACCAACGATAG

SIX5primer for RT Random hexamers

qPCR primers Forward TGGTGGTGCTGGGGGTTGTATCReverse GGGGCAGGGTGTTCCGCTTAC

Dmpk (specific priming)primer for RT CGACTGGAGCACGAGGACACTGACTCAGCAGCGTTAGCA

qPCR primers Forward GGAAGAAAGGGATGTATTAReverse CTCAGCAGCGTTAGCA

Dmpk (random priming)primer for RT Random hexamers


18S (specific priming)primer for RT CGGGTTGGTTTTGATCTG

qPCR primers Forward CAGTGAAACTGCGAATGGReverse CGGGTTGGTTTTGATCTG

18S (random priming)primer for RT Random hexamers


Gapdh (specific priming)primer for RT TGTAGACCATGTAGTTGAGGTCA

qPCR primers Forward AGGTCGGTGAACGGATTTGReverse TGTAGACCATGTAGTTGAGGTCA

Gapdh (random priming)primer for RT Random hexamers


the CTCF recognition sequence are methylated or mutated[19] Since we did not observe an all-or-nothing CpG methy-lation pattern we investigated CTCF binding to the twoCTCFbs flanking the CTG-repeat We performed ChIP onchromatin preparations of adult heart comparing the threerepeat length categories followed by qPCR to analyse quan-tities of immunoprecipitated DNA Positions of ampliconsat the DM1 locus are drawn in Figure 1 For ChIP analysesof histone modifications and CTCF binding we performedcontrol experiments with both a positive and a negativecontrol amplicon Overall these control experiments led tosatisfactory results Details of statistical analyses can be foundin Supplementary Tables 1 and 2 (see SupplementaryMaterialavailable online at httpdxdoiorg1011552013567435)

We observed enrichment in CTCF-immunoprecipitatedsamples of all CTG-repeat length categories at CTCFbs1but no significant difference between mice carrying normalor expanded repeat (Figure 3) Although the enrichment

appeared slightly lower inDMSXL as compared toDM20 andDM300 this trend did not reach statistical significance

At CTCFbs2 binding of CTCF seemed lower than bind-ing at CTCFbs1 and enrichment was significant only inDMSXL mice probably due to experimental variability Nosignificant trend across the categories was observed for CTCFenrichment at CTCFbs2

At the enhancer region no statistically significant enrich-ment was seen showing as expected no CTCF binding inthis region (Figure 3)

33 CTG-Repeat Expansion Is Associated with Local Chro-matin Remodelling around the CTCFbs Methylated CpGsmay attract chromatin-remodelling enzymes thus we anal-ysed chromatin remodelling in the presence of an expandedCTG-repeat in our mice We performed ChIP with anti-bodies directed against histone modifications that representactively transcribed (H3K914Ac) or repressed (H3K27Me3


25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1

Test for trend P = 0210

lowast lowast lowastlowast

IgG CTCF CTCF CTCFIgG IgGDM20 DM300 DMSXL

(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs2Test for trend P = 0872

NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)

Enhancer regionTest for trend P = 0936

NSS NSS NSS


(c)

Figure 3 CTCF binds to CTCFbs1 also in the presence of expanded CTG-repeats These graphs show enrichment ((Qt(IP)Qt(input)normalised against the abChIP enrichment value of the positive control amplicon of each respective repeat length category) for CTCF inabChIP versus IgG-IP at three regions at the DM1 locus CTCFbs1 shows statistically significant enrichment for CTCF in abChIP versus IgG-IP in all repeat length categories No such CTCF-binding is seen at CTCFbs2 or at the enhancer region Height of the bars indicates medianenrichment error bars indicate the interquartile range (25th to 75th percentile of observations) Mann-Whitney tests were performed to testfor a statistically significant difference between the abChIP and IgG reactions Results are summarised here with lowast being 119875 lt 005 lowastlowast being119875 lt 001 and lowastlowastlowast being 119875 lt 0001 Details of the statistical analysis can be found in Supplementary Table 1 Enrichment values obtainedfor abChIP reactions were subjected to the Jonckheere Terpstra test for trend which tests whether a trend exists across the categories withincreasing CTG-repeat length 119875 values are indicated in the graphs and details of this statistical analysis can be found in SupplementaryTable 2

and H3K9Me3) chromatin Enrichment for these histonemodifications around CTCFbs1 CTCFbs2 and the enhancerregion was studied by qPCR on chromatin immunoprecip-itated DNA Approximate amplicon locations at the DM1locus are indicated in Figure 1

331 H3K914Ac Histone Modification Representative ofTranscriptionally Active Chromatin We observed statisti-cally significant enrichment with an antibody directedagainst acetylated H3K914 (H3K914Ac) around CTCFbs1CTCFbs2 and to a lower extend at the enhancer region(Figure 4(a)) At both CTCFbs1 and bs2 Jonckheere Terpstratest for trend revealed a statistically significant decrease of

H3K914Ac enrichment across CTG-repeat length categories(CTCFbs1 119911-score minus2931 119875 = 0003 CTCFbs2 119911-scoreminus2996 119875 = 0003 where a negative 119911-score indicates adescending trend thus a lower median enrichment was seenwith increasing CTG-repeat length) Hence chromatin ofmice with longer repeats was less enriched for the activehistonemodification than chromatin of controlmice Enrich-ment at the enhancer regionwas low and did not show a trendacross the different categories of mice

332 H3K27Me3 and H3K9Me3 Histone ModificationsIndicative of Transcriptionally Repressed Chromatin As afirst investigation of possible heterochromatinisation we


lowastlowastlowastlowastlowastlowastlowastlowastlowast

H3KAcH3KAcH3KAcIgG IgG IgGDM20 DM300 DMSXL

3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)

lowastlowastlowast lowastlowastlowast lowastlowast

H3K27MeH3K27MeH3K27MeIgG IgG IgGDM20 DM300 DMSXL

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00

IgG H3K9Me IgG H3K9Me IgG H3K9MeDM20 DM300 DMSXL

Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS

Test for trend P = 0002CTCFbs2

Enhancer region



lowastlowastlowast

(c)

Figure 4 Expanded CTG-repeats are associated with local chromatin remodelling towards a less active chromatin conformation aroundthe CTG-repeat Less occupancy of active histone marks H3K914Ac (H3KAc in graph) (a) and more repressive histone marks (H3K9Me3)(c) around CTCFbs1 and bs2 is observed upon CTG-repeat expansion while enrichment of repressive mark H3K27Me3 (b) is unaffectedThese graphs show enrichment (Qt(IP)Qt(input) normalised against the abCHIP enrichment value of the positive control amplicon of eachrespective repeat length category) for different histonemodifications in abChIP versus IgG-IP at three regions at the DM1 locus Height of thebars indicates themedian enrichment and error bars indicate the interquartile range (25th to 75th percentile of observations)Mann-Whitneytests were performed to test for a statistically significant difference between the abChIP and IgG reactions Results are summarised here withlowast being 119875 lt 005 lowastlowast being 119875 lt 001 and lowastlowastlowast being 119875 lt 0001 Details of the statistical analysis can be found in Supplementary Table 1 Theenrichment values obtained for the specific abChIP reactions were subjected to the Jonckheere Terpstra test for trend which tests whether atrend exists across the categories with increasing CTG-repeat length 119875 values are indicated in the graphs and details of this statistical analysiscan be found in Supplementary Table 2

performed ChIP with an antibody directed against therepressive histone mark trimethylated H3K27 (H3K27Me3)We saw low but statistically significant enrichment at allamplicons and in all CTG-repeat length categories as shownin Figure 4(b) Although enrichment appeared slightlyhigher in the DMSXL mice for the CTCFbs1 and CTCFbs2amplicons no statistically significant trend with increasingrepeat length was seen

We next investigated another histone methylation markrepresentative of transcriptionally repressed chromatintrimethylated H3K9 (H3K9Me3 Figure 4(c)) DM20 andDM300 did not show statistically significant enrichment inthe specific antibody ChIP reaction (abChIP) versus IgG-IPat CTCFbs1 and bs2 whereas DMSXL did Thus H3K9Me3only binds at and around the CTCFbs in DMSXL samplesThis was confirmed by a statistically significant trend across

categories (CTCFbs1 119911-score 3084 119875 = 0002 CTCFbs2119911-score 2599 119875 = 0009) No statistically significantenrichment nor a statistically significant trend for H3K9Me3enrichment was seen at the enhancer region (Figure 4(c))

34 Lower DMPK Sense and SIX5 Transcript Levels in Micewith Expanded CTG-Repeats While Antisense TranscriptLevels Appear Unaffected by CTG-Repeat Length We inves-tigated possible changes in expression levels at the DM1locus since chromatin remodelling is generally accompaniedwith changes in gene expression Sense DMPK transcriptlevels showed a sharp decrease between DM20 and micewith expanded repeats (Figure 5) An overall statisticallysignificant trend was observed across the repeat lengthcategories for both reference genes (see Supplementary Table


8

6

4

2

0

DM20 DM300 DMSXL

DMPKDmpkTest for trend P = 002

RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL

DMPK antisenseDmpkTest for trend P = 1000

RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00

SIX5DmpkTest for trend P lt 0001

(c)

Figure 5Decreased sense transcript levels at theDM1 locus in the presence of expandedCTG-repeats UponCTG-repeat expansion decreaseof DMPK sense and SIX5 transcript levels is observed while DMPK antisense levels are unaffected by CTG-repeat length These graphsshow relative abundance (in arbitrary units) of transcripts of interest corrected by two different reference genes (18s (a) and endogenousmurineDmpk (b)) Height of bars indicates the median enrichment and error bars indicate the interquartile range (25th to 75th percentile ofobservations) Relative abundance values were subjected to the Jonckheere Terpstra test for trend which tests whether a trend exists acrossthe categories with increasing CTG-repeat length 119875 values are indicated in the graphs and details of this statistical analysis can be found inSupplementary Table 3

3 for results of Jonckheere Terpstra test for trend normalisedagainst 18s 119911-score minus3160 119875 = 0002 Dmpk 119911-scoreminus2332 119875 = 002) Mann-Whitney pairwise comparisonsdid not show a statistically significant decrease in DMPKmRNA expression between DM300 and DMSXL (mediansand interquartile ranges are shown in Supplementary Table3 data of Mann-Whitney analysis not shown) The DMPKantisense transcript did not show a similar trend of changingexpression across repeat length categories when antisensemRNA levels were normalised against Dmpk (Figure 5)(Dmpk Jonckheere Terpstra test for trend 119911-score 0 119875 = 1Supplementary Table 3) However a statistically significanttrend was observed when antisense transcript levels werenormalised against 18s (119911-score minus2107 119875 = 0035) Post hocpairwise Mann-Whitney comparisons however did not showstatistically significant differences between any of the repeatlength categories when correcting the 119875 values for multiplecomparisons (data not shown)

SIX5 expression levels are affected by increasing repeatlength as they show a significant decrease across the repeatlength categories The decrease is lower than that seen forDMPK sense transcripts but consistent with both referencegenes (Jonckheere Terpstra test for trend versus 18s 119911-scoreminus2332 119875 = 002 versus Dmpk 119911-score minus3912 119875 lt 0001Supplementary Table 3) No difference was observed betweenDM300 and DMSXL mice

35 PCNA Binding Near the Expanded CTG-Repeat PCNA(proliferating cell nuclear antigen) is involved in manycellular processes [40] In addition to its role in replicationPCNA recruits a variety of epigenetic regulators [41] Loops ofslipped-strand structures formed by expanded CTG-repeatscould serve as loading sites for PCNA and binding of PCNAto the expanded CTG-repeat could be the beginning of thecascade of chromatin remodelling event [8 42]


15

10

05

00

lowastlowast lowastlowast lowastlowastlowast

IgG PCNA IgG PCNA IgG PCNADM20 DM300 DMSXL

Amylase positive controlTest for trend P = 1

Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS

PCNA IgG PCNA IgG PCNADM20 DM300 DMSXL

Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00

lowastlowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)lowast


(d)

Figure 6 PCNA seems to bind to expanded CTG-repeats CTCFbs1 and bs2 but not the enhancer region show statistically significantenrichment of PCNA in abChIP versus IgG-IP in the expanded CTG-repeat length categories but not at DM20 suggesting that PCNAbinds the expanded CTG-repeat These graphs show enrichment (Qt(IP)Qt(input) normalised against the abCHIP enrichment value of thepositive control amplicon of each respective repeat length category) for PCNA in abChIP versus IgG-IP at a positive control amplicon andthree regions at the DM1 locus Height of the bars indicates the median enrichment and error bars indicate the interquartile range (25th to75th percentile of observations) Mann-Whitney tests were performed to test for a statistically significant difference between the abChIP andIgG reactions Results are summarised here with lowast being119875 lt 005 lowastlowast being119875 lt 001 and lowastlowastlowast being119875 lt 0001 Details of the statistical analysiscan be found in Supplementary Table 1 The enrichment values obtained for the specific abChIP reactions were subjected to the JonckheereTerpstra test for trend which tests whether a trend exists across the categories with increasing CTG-repeat length 119875 values are indicated inthe graphs and details of this statistical analysis can be found in Supplementary Table 2

We therefore investigated whether PCNA binds near theCTG-repeats in our mice by ChIP followed by qPCR analy-sis Enrichments measured with the enhancer amplicon weremodest compared to the positive Amylase control and weresignificant only for DM300mice (119875 = 001 IgG versus PCNAantibody Figure 6) At both CTCFbs1 and 2 we did not seestatistically significant enrichment at DM20 but enrichmentwas detected for DM300 and DMSXL suggesting bindingof PCNA to expanded CTG-repeats Jonckheere Terpstratest for trend did not reveal a statistically significant trendindicating that similar PCNA-binding was detected despitea longer repeat At the enhancer region on the contrary wedid not observe statistically significant enrichment except for

modest enrichment in DM300 Thus these preliminary dataseem to suggest that PCNA specifically binds close to theamplicons at CTCFbs1 and 2 but not the enhancer region

4 Discussion

We here show that expanded CTG-repeats induce a locallyrepressed chromatin state and accompanying reduced sensegene transcription at the DM1 locus in adult transgenicmouse hearts

Mice with expanded repeats showed substantial methyla-tion at and around the CTCFbs as opposed to DM20 whichshowed very little CpG methylation This CpG methylation


is not seen at all CpGs nor in all individual cells but overallDNA methylation levels are higher with increasing repeatlength DM20 and DM300SXL are independent transgeniclines and we cannot exclude an influence of the transgeneintegration site sequences However the transgene is large(45 kb) and contains the major regulatory sequences betweenDMPK and SIX5 [14 19] All different lines obtained withdifferent repeat lengths showed the same tissue-specific pat-tern of DMPK expression which is also similar to the mouseDmpk gene and to the DMPK gene in human tissues [31 32]In addition expression levels correlate to copy number of theintegrated transgene indicating that the surrounding mousesequences have no or minimal impact on the transgene

We observed that the percentage of cells carrying amethyl group at a given CpG was substantial in DM300and even higher in DMSXL both at CTCFbs1 and bs2CpG methylation was more abundant at CTCFbs1 than atCTCFbs2 confirming a polarised localisation of methylationat the DM1 locus as described before [39] This is in linewith evidence pointing at a more important regulatory rolefor CTCFbs1 [19 43] Our observation that the CTCFbs1recognition sequence itself was relatively spared from CpGmethylation as opposed to surrounding sequences is alsoworth noting in this respect

CpG methylation around the DM1 CTG-repeat hadpreviously been assessed by measuring the height of chro-matogram peaks obtained for cytosine after sequencing ofbisulfite-converted DNA [39] That study showed differencesin methylation patterns between DM300 mice and humans[43] Most importantly adult human DM1 samples nevershowed methylation at CTCFbs2 whereas DM300 mice didindicating that the mouse model does not fully mimic thehuman situation underlining limitations of animal modelsHowever one human foetus did show CpG methylation atand around CTCFbs2 indicating individual variation amonghuman patients [39] We demonstrate here that mice witharound 1450 CTGs clearly show more methylation than micewith around 600CTGwith a strong bias 51015840 of theCTG-repeatMice with longer expanded repeat tracts may therefore bettermimic the human situation

The variable DNA methylation pattern found in ourmice around theCTG-repeat resembles that observed aroundthe SCA7-CAG-repeat in a transgenic mouse model forSpinocerebellar Ataxia 7 (SCA7) [24] A strong correlationbetween the severity of disease symptoms and level of DNAmethylation has been described at the CGG-repeat andpromoter region of the FMR1 gene [44 45] In addition itwas recently proposed that also variable methylation patternsin the promoter of the ATXN2 gene explain considerablevariation in anticipation in the absence of intergenerationalCAG-repeat instability Different degrees of methylation ofthe ATXN2 promoter could be related to age of onsetin patients with SCA2 SCA3 suggesting that gene dosagethrough this epigenetic mechanism is important for diseaseoutcome [46] Thus these observations underline that CpGmethylation is no all-or-nothing phenomenon at TNR lociand underscore the importance of careful examination ofmethylation status of individual CpGs Relevant mechanistic

information might be missed when a more general approachis followed

In vitro studies have shown disrupted binding of CTCF tothe CTCFbs uponmutation ormethylation of the recognitionsequence [19] Since our data show that CpG methylation isno all-or-nothing phenomenon at the studied locus it wasunclear what to expect concerning CTCF binding For thefirst time inmammalian tissueswe show that CTCF still bindsto CTCFbs1 despite the presence of an expanded CTG-repeatof up to sim1600 units (Figure 3) We did not detect significantbinding of CTCF to CTCFbs2 which is consistent with invitro binding assays that showed stronger binding of CTCFto site 1 [19] It was surprising to observe clear CTCF bindingat CTCFbs1 in vivo despite abundant CpG methylation ofthe region However it is interesting to note that the CTCFrecognition sequence is relatively spared from methylationwhen compared to the adjacent region It is possible thatthe repeat in our mice is not large enough to induce fullmethylation of the binding site Alternatively at the H19locus binding of CTCF has been demonstrated to preventCpG methylation [47 48] Further research may shed lighton the order of events

Methylated CpGs are known to attract chromatin-remodelling enzymes [41 49 50] In vitro the nucleosomeassembly of DNA containing repeating CTG triplets showedthat the efficiency of nucleosome formation increased withexpanded triplet blocks [16ndash18] suggesting that such blocksmay profoundly alter local chromatin structure and represstranscription through the creation of stable nucleosomesWe therefore explored possible chromatin rearrangement inmice with expanded CTG-repeats as opposed to DM20 Wefound chromatin remodelling indicative of a transcriptionallyrepressed state close to the expanded CTG-repeat in DM300and DMSXL mice (Figure 4) The enhancer region generallyshowed a different enrichment pattern for histone modifica-tions than around the CTCFbs suggesting that a local regionof heterochromatin is formed close to the expanded CTG-repeat within a euchromatin region This has previouslybeen demonstrated in patient cells [14] heterochromatinspreading was seen upon expansion of the CTG-repeat asactive HK4Me3 was replaced by the repressive H3K9Me3mark When the expanded CTG-repeat induced heterochro-matinisation adjacent genes were silenced by propagationof heterochromatin along the chromosome [22] We did seeincreased enrichment of H3K9Me3 around very long CTG-repeats but heterochromatinisation did not propagate to theenhancer region Prominent decreasing trends of H3K914Acenrichment at both CTCFbs were observed across increasingCTG-repeat length categories A similar graded loss ofacetylated H3 and H4 with increasing CGG-repeat lengthhas been observed in FXS patient cells [23] Considering thatqPCR reactions for CTCFbs1 and bs2 were performed onthe same ChIPed DNA and that qPCR efficiencies were verysimilar itmight be concluded that CTCFbs2 ismore enrichedfor H3K914Ac As yet we cannot know whether this has afunctional implication orwhether it is linked to the seeminglymore important regulatory role of CTCFbs1

We chose to study one active histone modification onlyas genome-wide histone modification maps show that the


distribution of most histone marks recognised to be activeis highly similar [51] Fewer histone modifications associatedwith repressive chromatin have been described and little isknown about their global linkage [51] Our data show thatH3K9Me3 and H3K27Me3 did not respond in the same wayto the expanded CTG-repeat In a study that investigated theepigenetic status of the euchromatic region of the human Ychromosome H3K9Me3 and H3k9Ac enrichment correlatedwith the expression status whereas H3K27Me3 enrichmentdid not This suggests a mechanism where H3K9Me3 andH3K9Ac dominate over H3K27Me modifications to deter-mine expression status of the chromatin [52] A similarsituation seems to be the case in our adult mouse hearts

Consistent with chromatin changes representative oftranscriptional repression we saw lower sense DMPK andSIX5 expression in mice that carry an expanded CTG-repeat(Figure 5) We did not detect a further decrease of DMPKand SIX5 transcription levels when comparing DM300 andDMSXL at 5 months of age despite the 30 decreasepreviously observed at 2months of age in the same transgenicmice [31] This could be due to the decreased transgeneexpression we observed with age (data not shown)Thereforefactors other than chromatin may also contribute to thechange in DM1 expression

The DMPK antisense transcript emanates from the SIX5adjacent regulatory region In DMSXLmice it is expressed inmany tissues with expression being the highest in heart asis the case for the DMPK sense transcript although it doesnot follow nor mirror the same expression profile [31] Sensemessenger levels are higher than antisense Since the anti-sense transcript and SIX5 have overlapping promoter regionsit could be postulated that they are subject to similar reg-ulatory factors However in the current study SIX5 mRNAlevels decreased whereas antisense transcript levels remainedunaffected in the presence of an expanded CTG-repeatInterestingly this finding demonstrates that regulation ofDMPK antisense is independent although some regulatorysequences might be shared with DMPK and SIX5 Futureresearchwill likely shedmore light on the role of bidirectionaltranscription in the DMPK gene and at other TNR loci

Previous studies have demonstrated that PCNA can beloaded onto dsDNA-ssDNA junctions in DNA-loops orloops of slipped-strand structures formed by expanded CTG-repeats [8 42]We therefore recognised in PCNA a candidatemolecule that might bind to the expanded CTG-repeat andthen cause a cascade of chromatin-modifying eventsWe hereshow evidence that PCNA indeed binds to or close to theexpanded CTG-repeat (Figure 6) PCNA appears to bind toa similar extent to CTCFbs1 and bs2 amplicons which isaccording to expectations since both amplicons lie very closeto the CTG-repeat It is possible that more PCNA was boundto the longer CTG-repeat inDMSXLmice but that we cannotdetect this due to the size of our sheared fragments

By recruiting epigenetic regulators PCNA might be theinstigator of multiple downstream chromatin modifications[41] PCNA is known to interact with DNA methyltrans-ferases DNMT1 -3A and -3B as well as with histone methyl-transferases These interactions cause H3K9 and H3K27to become trimethylated yielding a repressed chromatin

environment Via another route DNMT1 interacts withhistone deacetylases (HDACs) which also contributes toa repressed chromatin context [41] The observed enrich-ment pattern of histone modifications and hypermethylationaround the expanded CTG-repeat in adult hearts of mice fitwith this model Importantly we detected enrichment forPCNA in mice with expanded CTG-repeats to a similarextent at both CTCFbs but not at the enhancer regionsuggesting that PCNA binds to or very near the expandedCTG-repeat specifically

DNA methylation and histone modifications appear toreciprocally influence each other [41] Thus multiple par-allel pathways seem responsible for the establishment of arepressed chromatin status

The involvement of PCNA needs to be confirmed and theprecise order of events remains to be elucidated

We here presented evidence that expanded CTG-repeatsinduce CpG methylation and local heterochromatinisationclose to the repeat This is accompanied by decreasedlevels of sense DMPK and SIX5 transcription CTCFbinding at the DM1 locus was not affected by the expansionof the CTG-repeat We found that PCNA binds in thevicinity of expanded CTG repeats and might be recruitedto the expanded CTG-repeat We propose that it couldsubsequently attract chromatin-remodelling enzymes thatyield the repressive changes in chromatin dynamics A betterunderstanding of the precise cascade of processes induced byexpanded TNRs and importantly the starting point of thesechanges will provide us with therapeutic targets to alleviatedisease progression and limit further TNR expansion

Conflict of Interests

The authors have no conflict of interests to report

Funding

This work was supported by Agence Nationale de RechercheFrance (to Genevieve Gourdon) Association Francaisecontre les Myopathies (to Genevieve Gourdon) InstituteNational de la Sante et Recherche Medicale France (toGenevieve Gourdon) Universite Paris Descartes France(to Genevieve Gourdon) Association Francaise contre lesMyopathies (to Judith Rixt Brouwer) and Marie Curie Intra-European Fellowship of 7th Framework Programme of theEuropean Commission (Grant no 252610 to Judith RixtBrouwer)

Acknowledgment

The authors are grateful to personnel from CERFE (CentredrsquoExploration et de Recherche Fonctionelle ExperimentaleGenopole Evry France) for attentively caring for the mice

References

[1] J R Brouwer R Willemsen and B A Oostra ldquoMicrosatelliterepeat instability and neurological diseaserdquo BioEssays vol 31no 1 pp 71ndash83 2009


[2] R I Richards and G R Sutherland ldquoHeritable unstable DNAsequencesrdquo Nature Genetics vol 1 no 1 pp 7ndash9 1992

[3] C Aslanidis G Jansen C Amemiya et al ldquoCloning of theessential myotonic dystrophy region and mapping of the puta-tive defectrdquo Nature vol 355 no 6360 pp 548ndash551 1992

[4] J D Brook M E McCurrach H G Harley et al ldquoMolecularbasis of myotonic dystrophy expansion of a trinucleotide(CTG) repeat at the 31015840 end of a transcript encoding a proteinkinase family memberrdquo Cell vol 68 no 4 pp 799ndash808 1992

[5] Y-H Fu A Pizzuti R G Fenwick Jr et al ldquoAn unstabletriplet repeat in a gene related tomyotonicmuscular dystrophyrdquoScience vol 255 no 5049 pp 1256ndash1258 1992

[6] T Ashizawa and P S Sarkar ldquoMyotonic dystrophy types 1 and2rdquo Handbook of Clinical Neurology vol 101 pp 193ndash237 2011

[7] A L Castel J D Cleary and C E Pearson ldquoRepeat instabilityas the basis for human diseases and as a potential target fortherapyrdquo Nature Reviews Molecular Cell Biology vol 11 no 3pp 165ndash170 2010

[8] M Nakamori C E Pearson and C A Thornton ldquoBidirec-tional transcription stimulates expansion and contraction ofexpanded (CTG)lowast(CAG) repeatsrdquo Human Molecular Geneticsvol 20 no 3 pp 580ndash588 2011

[9] V Dion and J H Wilson ldquoInstability and chromatin structureof expanded trinucleotide repeatsrdquo Trends in Genetics vol 25no 7 pp 288ndash297 2009

[10] K L Taneja M McCurrach M Schalling D Housman andR H Singer ldquoFoci of trinucleotide repeat transcripts in nucleiof myotonic dystrophy cells and tissuesrdquo Journal of Cell Biologyvol 128 no 6 pp 995ndash1002 1995

[11] G Sicot G Gourdon and M Gomes-Pereira ldquoMyotonicdystrophy when simple repeats reveal complex pathogenicentities new findings and future challengesrdquoHumanMolecularGenetics vol 20 no 2 pp R116ndashR123 2011

[12] R Perbellini S Greco G Sarra-Ferraris et al ldquoDysregulationand cellular mislocalization of specific miRNAs in myotonicdystrophy type 1rdquo Neuromuscular Disorders vol 21 no 2 pp81ndash88 2011

[13] F Rau F Freyermuth C Fugier et al ldquoMisregulation ofmiR-1 processing is associated with heart defects in myotonicdystrophyrdquoNature Structural and Molecular Biology vol 18 no7 pp 840ndash845 2011

[14] D H Cho C P Thienes S E Mahoney E Analau G NFilippova and S J Tapscott ldquoAntisense transcription andheterochromatin at the DM1 CTG repeats are constrained byCTCFrdquoMolecular Cell vol 20 no 3 pp 483ndash489 2005

[15] T Zu B Gibbens N S Doty et al ldquoNon-ATG-initiatedtranslation directed by microsatellite expansionsrdquo Proceedingsof the National Academy of Sciences of the United States ofAmerica vol 108 no 1 pp 260ndash265 2011

[16] Y-HWang S Amirhaeri S Kang RDWells and JDGriffithldquoPreferential nucleosome assembly at DNA triplet repeats fromthe myotonic dystrophy generdquo Science vol 265 no 5172 pp669ndash671 1994

[17] C B Volle and S Delaney ldquoCAGCTG repeats alter theaffinity for the histone core and the positioning of DNA in thenucleosomerdquo Biochemistry vol 51 pp 9814ndash9825 2012

[18] Y-H Wang and J Griffith ldquoExpanded CTG triplet blocks fromthemyotonic dystrophy gene create the strongest knownnaturalnucleosome positioning elementsrdquo Genomics vol 25 no 2 pp570ndash573 1995

[19] G N Filippova C PThienes B H Penn et al ldquoCTCF-bindingsites flank CTGCAG repeats and form a methylation-sensitiveinsulator at the DM1 locusrdquo Nature Genetics vol 28 no 4 pp335ndash343 2001

[20] A D Otten and S J Tapscott ldquoTriplet repeat expansion inmyotonic dystrophy alters the adjacent chromatin structurerdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 92 no 12 pp 5465ndash5469 1995

[21] T R Klesert A D Otten T D Bird and S J TapscottldquoTrinucleotide repeat expansion at the myotonic dystrophylocus reduces expression of DMAHPrdquo Nature Genetics vol 16no 4 pp 402ndash406 1997

[22] A Savellev C Everett T Sharpe Z Webster and R Festen-stein ldquoDNA triplet repeats mediate heterochromatin-protein-1-sensitive variegated gene silencingrdquo Nature vol 422 no 6934pp 909ndash913 2003

[23] B Coffee F Zhang S Ceman S TWarren andD Reines ldquoHis-tone modifications depict an aberrantly heterochromatinizedFMR1 gene in fragile X syndromerdquoAmerican Journal of HumanGenetics vol 71 no 4 pp 923ndash932 2002

[24] R T Libby K A Hagerman V V Pineda et al ldquoCTCFcis-regulates trinucleotide repeat instability in an epigeneticmanner a novel basis for mutational hot spot determinationrdquoPLoS Genetics vol 4 no 11 Article ID e1000257 2008

[25] B L Sopher P D Ladd V V Pineda et al ldquoCTCF regulatesataxin-7 expression through promotion of a convergently tran-scribed antisense noncoding RNArdquo Neuron vol 70 no 6 pp1071ndash1084 2011

[26] A V Goula A Stys J P Chan Y Trottier R Festensteinand K Merienne ldquoTranscription elongation and tissue-specificsomatic CAG instabilityrdquo PLOS Genetics vol 8 Article IDe1003051 2012

[27] H Seznec A-S Lia-Baldini C Duros et al ldquoTransgenic micecarrying large human genomic sequences with expanded CTGrepeat mimic closely the DMCTG repeat intergenerational andsomatic instabilityrdquoHumanMolecular Genetics vol 9 no 8 pp1185ndash1194 2000

[28] M Gomes-Pereira L Foiry A Nicole et al ldquoCTG trinucleotiderepeat ldquobig jumpsrdquo large expansions small micerdquo PLoS Genet-ics vol 3 no 4 article e52 2007

[29] G Gourdon F Radvanyi A-S Lia et al ldquoModerate intergener-ational and somatic instability of a 55-CTG repeat in transgenicmicerdquo Nature Genetics vol 15 no 2 pp 190ndash192 1997

[30] A-S Lia H Seznec H Hofmann-Radvanyi et al ldquoSomaticinstability of the CTG repeat in mice transgenic for themyotonic dystrophy region is age dependent but not correlatedto the relative intertissue transcription levels and proliferativecapacitiesrdquo Human Molecular Genetics vol 7 no 8 pp 1285ndash1291 1998

[31] A Huguet F Medja and A Nicole ldquoMolecular physiologicaland motor performance defects in DMSXL mice carryinggt1000 CTG repeats from the human DM1 locusrdquo PLOS Genet-ics vol 8 Article ID e1003043 2012

[32] H Seznec O Agbulut N Sergeant et al ldquoMice transgenic forthe human myotonic dystrophy region with expanded CTGrepeats display muscular and brain abnormalitiesrdquo HumanMolecular Genetics vol 10 no 23 pp 2717ndash2726 2001

[33] C Guiraud-Dogan A Huguet M Gomes-Pereira et al ldquoDM1CTG expansions affect insulin receptor isoforms expression invarious tissues of transgenic micerdquo Biochimica et BiophysicaActa vol 1772 no 11-12 pp 1183ndash1191 2007


[34] M Gomes-Pereira L Foiry and G Gourdon ldquoTransgenicmouse models of unstable trinucleotide repeats toward anunderstanding of disease-associated repeat size mutationrdquo inGenetic Instabilities and Neurological Diseases R Wells andT Ashizawa Eds pp 563ndash583 Academic Press 2nd edition2006

[35] M Gomes-Pereira T A Cooper and G Gourdon ldquoMyotonicdystrophy mouse models towards rational therapy develop-mentrdquo Trends in Molecular Medicine vol 17 no 9 pp 506ndash5172011

[36] J R Brouwer L Foiry and G Gourdon ldquoCell recovery fromDM1 transgenic mouse tissue to study (CTG) n instability andDM1 pathogenesisrdquo in Methods in Molecular Biology vol 1010pp 253ndash264 2013

[37] S Tome I Holt W Edelmann et al ldquoMSH2 ATPase domainmutation affects CTG∙CAG repeat instability in transgenicmicerdquo PLoS Genetics vol 5 no 5 Article ID e1000482 2009

[38] C Bock S Reither T Mikeska M Paulsen J Walter and TLengauer ldquoBiQ Analyzer visualization and quality control forDNAmethylation data from bisulfite sequencingrdquo Bioinformat-ics vol 21 no 21 pp 4067ndash4068 2005

[39] A Lopez Castel M Nakamori S Tome et al ldquoExpanded CTGrepeat demarcates a boundary for abnormal CpG methyla-tion in myotonic dystrophy patient tissuesrdquo Human MolecularGenetics vol 20 no 1 pp 1ndash15 2011

[40] G-LMoldovan B Pfander and S Jentsch ldquoPCNA themaestroof the replication forkrdquo Cell vol 129 no 4 pp 665ndash679 2007

[41] B Jin Y Li and K D Robertson ldquoDNA methylation superioror subordinate in the epigenetic hierarchyrdquo Genes and Cancervol 2 no 6 pp 607ndash617 2011

[42] A Pluciennik L Dzantiev R R Iyer N Constantin F AKadyrov and PModrich ldquoPCNA function in the activation andstrand direction of MutL120572 endonuclease in mismatch repairrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 107 no 37 pp 16066ndash16071 2010

[43] J D Cleary S Tome A L Castel et al ldquoTissue-and age-specific DNA replication patterns at the CTGCAG-expandedhumanmyotonic dystrophy type 1 locusrdquoNature Structural andMolecular Biology vol 17 no 9 pp 1079ndash1087 2010

[44] A McConkie-Rosell A M Lachiewicz G A Spiridigliozzi etal ldquoEvidence thatmethylation of the FMR-1 locus is responsiblefor variable phenotypic expression of the fragile X syndromerdquoAmerican Journal ofHumanGenetics vol 53 no 4 pp 800ndash8091993

[45] M Pieretti F Zhang Y-H Fu et al ldquoAbsence of expression ofthe FMR-1 gene in fragile X syndromerdquo Cell vol 66 no 4 pp817ndash822 1991

[46] J M Laffita-Mesa P O Bauer V Kourı et al ldquoEpigeneticsDNA methylation in the core ataxin-2 gene promoter novelphysiological and pathological implicationsrdquo Human Geneticsvol 131 pp 625ndash638 2011

[47] V Pant S Kurukuti E Pugacheva et al ldquoMutation of a SingleCTCF Target Site within the H19 Imprinting Control RegionLeads to Loss of Igf2 Imprinting and Complex Patterns of deNovo Methylation upon Maternal Inheritancerdquo Molecular andCellular Biology vol 24 no 8 pp 3497ndash3504 2004

[48] C J Schoenherr J M Levorse and S M Tilghman ldquoCTCFmaintains differentialmethylation at the Igf2H19 locusrdquoNatureGenetics vol 33 no 1 pp 66ndash69 2003

[49] B Coffee F Zhang S T Warren and D Reines ldquoAcetylatedhistones are associated with FMR1 in normal but not fragile X-syndrome cellsrdquoNature Genetics vol 22 no 1 pp 98ndash101 1999

[50] X Nan H-H Ng C A Johnson et al ldquoTranscriptionalrepression by themethyl-CpG-binding proteinMeCP2 involvesa histone deacetylase complexrdquo Nature vol 393 no 6683 pp386ndash389 1998

[51] F van Leeuwen and B van Steensel ldquoHistone modificationsfrom genome-wide maps to functional insightsrdquo Genome Biol-ogy vol 6 no 6 article 113 2005

[52] N P Singh S R Madabhushi S Srivastava et al ldquoEpigeneticprofile of the euchromatic region of human y chromosomerdquoNucleic Acids Research vol 39 no 9 pp 3594ndash3606 2011

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


One of themajor pathogenicmodels proposed to underlieDM1 is a toxic effect of the presence of expanded CUG-containing transcripts Mutant DMPK mRNAs are retainedin the nucleus accumulate in foci [10] and form com-plexes with regulatory proteins thereby preventing theseproteins from exerting their normal function [11] AberrantmiRNA metabolism has also been described in patientswith expanded CTG-repeats [12 13] Recent evidence thatbidirectional transcription [14] and nonconventional RNAtranslation [15] are taking place at several TNR loci is compli-cating the traditional picture of RNA toxicity These findingspoint at a scenario where not just one single expanded RNAtranscript is responsible for disease development [11]

Moreover chromatin dynamics are increasingly recog-nised to influence both TNR instability and gene expressionat TNR loci and thereby probably disease outcome TNRscan affect nucleosome positioning [16 17] and CTG-repeatsspecifically have been identified as preferential location fornucleosome formation [18] CTG- and CAG-repeats havebeen described to form a functional component of insulatorelements thereby influencing gene expression levels At theDM1 locus the CTG-repeat forms an insulator together withthe two CTCF-binding sites (CTCFbs) that flank the repeat[19] Long CTG-tracts were shown to induce condensationof DNA at the DM1 locus which could hinder access ofgene regulators to this region [20] Transcription of theSIX5 gene that neighbours DMPK was decreased in patientcells expressing expanded CTG-repeats [21] These findingstogether led to the proposal that the expanded CTG-repeatinduces a transcriptionally repressive region [21] IndeedlongCTG-repetitionswere shown to induce heterochromatinformation which can then spread into neighbouring regions[22]

Aberrant chromatin remodelling has also been observedat other TNR disease loci For instance an expanded CGG-repeat is associated with CpG hypermethylation heterochro-matinisation and silencing of the FMR1 gene which is thecause of Fragile X syndrome [23] DNA methylation andhistonemodifications representative of silent chromatin havebeen observed around the expanded GAA-repeat in intron 1of the Frataxin gene that causes Friedreichrsquos ataxia [9]

Although the chromatin context is now consideredimportant for DM1 [14] and other TNR loci [9 23ndash25] fewcomprehensive studies addressing multiple factors involvedin or associated with chromatin remodelling simultaneouslyhave been performed Recently studies in mouse models forHuntingtonrsquos disease suggested that the chromatin contextof the transgene integration site determines CAG-repeatinstability and transcription levels [26]Thus these processesseem tightly linked underscoring the need to understandchromatin dynamics at TNR loci

We therefore set out to study the consequences ofCTG-repeat expansion for CpGmethylation CTCF-bindingchromatin conformation and gene expression at the DM1locus making use of the transgenic mouse model previouslygenerated in our laboratory [27] These mice carry a largehuman genomic transgene that encompasses the DMPKgene as well as the neighbouring genes DMWD and SIX5The transgene includes either a normal CTG-repeat of 20

trinucleotides or disease-associated expanded repeats withthe latter showing CTG-repeat instability patterns similar toDM1 patients (strongly biased towards expansions length-and age-dependent somatic instability albeit showing smallerrepeat length changes per instability event in mice) [28ndash30] The transgene also encompasses important regulatorysequences such as the two CTCF-binding sites (CTCFbs) thatflank the CTG-repeat in humans and the enhancer of thedownstream SIX5 gene

The phenotype of homozygous mice carrying up to 1600CTGs has recently been characterised [31] Since these micedisplay multiple characteristics seen in DM1 [27 28 3233] they are considered a valuable model to investigatemechanisms implicated in CTG-repeat instability and DM1pathogenesis [34 35] Benefitting from this DM1 mousemodel we aimed to study the epigenetic consequences of theexpanded CTG-repeat at theDM1 locusWe present evidencethat expanded CTG-repeats induce CpG methylation andlocal heterochromatinisation and concurrent decreased tran-scription around the repeat without affecting significantlyCTCF binding at the DM1 locus We also found binding ofPCNA around the CTG-repeat and propose that it mightlie at the basis of CTG-repeat expansion-induced repressivechanges in chromatin dynamics

2 Materials and Methods

21 Transgenic Mice Mice used in this study harbour atransgene consisting of 45 kb of human genomicDNA clonedfrom a DM1 patient and have been described previously(crossbred to gt90 C57BL6 background) [27 28 31] Micewere genotyped by PCR amplification of tail DNA usingoligonucleotide primers DMHR8 (51015840-TGACGTGGATGG-GCAAACTG-31015840) DMHR9 (51015840-AGCTTTGCACTTTGC-GAACC-31015840) and Dmm9 (51015840-GCTTGTAACTGATGGCTG-GG-31015840) which amplify the endogenous murine Dmpk(DMHR8 and Dmm9) and the human transgene DMPK(DmHR8 and DMHR9) [36] CTG-repeat length was deter-mined by PCR amplification of DNA extracted from tailat weaning with oligonucleotide primer 101 (51015840-CTTCCC-AGGCCTGCAGTTTGCCCATC-31015840) and primer 102 (51015840-GAACGGGGCTCGAAGGGTCTTGTAGC-31015840) as describedbefore [37] followed by electrophoresis of PCR products ona large 08 (wv) agarose gel [36] For the current paperheterozygous mice with the following CTG-repeat lengthswere used DM20 mice of the DM20-949 line that carrythe normal unexpanded human allele DM300 mice of theDM300-328 line that currently have an average of 610 CTGs(range 545ndash700) and DMSXL mice of the DM300-328line that have undergone large expansions and now carryalleles with over 1000 CTGs [28] In this particular studymice with sim1300ndash1600 CTG-repeats (mean 1435 CTGs)were used Adult mice of 3ndash5 months of age were used forall experiments described in this study (mean age did notdiffer among the different genotype groups as tested withKruskal-Wallis DM20 mean age 44 months DM300 meanage 43 months and DMSXL 41 months 119875 = 0634) Wechose heart as a representative tissue for disease which






2










Exons 1ndash15




CTCFbs2 30CpGsQ-Enh




End of DMPK gene


3998400


(CTG)n





119902) values



5998400 Exon 15


CTCFbs1 (CTG)n


CTCF binding sites

CTCFbs23998400


(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

DM20

DM300

DMSXL

0 7 5 0 0 8 2 3 10 0 0 0 0 0 0 7 0 0 0 5 0 5 5 5 8 3 0 8 0 0 5 5 2 0 6 9 18

56 29 37 49 61 37 22 25 29 25 43 33 45 27 17 20 35 31 36 45 41 28 20 33 73 4 65 6 17 7 41 28 26 35 26 47 46

89 70 79 83 95 70 52 61 66 57 69 65 80 49 48 56 66 50 83 79 84 78 67 86 82 26 70 28 42 23 53 45 67 76 48 80 90

CTCFbs2CpG nr

CpG nr

DM20

DM300

DMSXL

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

0 2 5 2 2 7 0 2 5 4 2 0 0 0 4 3 2 0 4 2 2 2 2 0 4 6 2 0 2 0

38 29 21 12 20 57 21 16 16 21 23 16 0 38 3 32 20 3 7 5 29 19 10 11 0 10 3 0 6 2

78 17 17 22 17 40 18 19 26 30 36 26 28 26 19 13 4 26 17 2 43 30 24 26 7 6 2 4 4 2


(CTG)n

(CTG)n


(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

CpG nrDM20


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

DM300

DMSXL

CTCFbs1

CTCFbs1(CTG)n

(CTG)n

(CTG)n

(c)









3 Results































25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology






2










Exons 1ndash15




CTCFbs2 30CpGsQ-Enh




End of DMPK gene


3998400


(CTG)n





119902) values



5998400 Exon 15


CTCFbs1 (CTG)n


CTCF binding sites

CTCFbs23998400


(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

DM20

DM300

DMSXL

0 7 5 0 0 8 2 3 10 0 0 0 0 0 0 7 0 0 0 5 0 5 5 5 8 3 0 8 0 0 5 5 2 0 6 9 18

56 29 37 49 61 37 22 25 29 25 43 33 45 27 17 20 35 31 36 45 41 28 20 33 73 4 65 6 17 7 41 28 26 35 26 47 46

89 70 79 83 95 70 52 61 66 57 69 65 80 49 48 56 66 50 83 79 84 78 67 86 82 26 70 28 42 23 53 45 67 76 48 80 90

CTCFbs2CpG nr

CpG nr

DM20

DM300

DMSXL

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

0 2 5 2 2 7 0 2 5 4 2 0 0 0 4 3 2 0 4 2 2 2 2 0 4 6 2 0 2 0

38 29 21 12 20 57 21 16 16 21 23 16 0 38 3 32 20 3 7 5 29 19 10 11 0 10 3 0 6 2

78 17 17 22 17 40 18 19 26 30 36 26 28 26 19 13 4 26 17 2 43 30 24 26 7 6 2 4 4 2


(CTG)n

(CTG)n


(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

CpG nrDM20


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

DM300

DMSXL

CTCFbs1

CTCFbs1(CTG)n

(CTG)n

(CTG)n

(c)









3 Results































25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology





Exons 1ndash15




CTCFbs2 30CpGsQ-Enh




End of DMPK gene


3998400


(CTG)n





119902) values



5998400 Exon 15


CTCFbs1 (CTG)n


CTCF binding sites

CTCFbs23998400


(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

DM20

DM300

DMSXL

0 7 5 0 0 8 2 3 10 0 0 0 0 0 0 7 0 0 0 5 0 5 5 5 8 3 0 8 0 0 5 5 2 0 6 9 18

56 29 37 49 61 37 22 25 29 25 43 33 45 27 17 20 35 31 36 45 41 28 20 33 73 4 65 6 17 7 41 28 26 35 26 47 46

89 70 79 83 95 70 52 61 66 57 69 65 80 49 48 56 66 50 83 79 84 78 67 86 82 26 70 28 42 23 53 45 67 76 48 80 90

CTCFbs2CpG nr

CpG nr

DM20

DM300

DMSXL

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

0 2 5 2 2 7 0 2 5 4 2 0 0 0 4 3 2 0 4 2 2 2 2 0 4 6 2 0 2 0

38 29 21 12 20 57 21 16 16 21 23 16 0 38 3 32 20 3 7 5 29 19 10 11 0 10 3 0 6 2

78 17 17 22 17 40 18 19 26 30 36 26 28 26 19 13 4 26 17 2 43 30 24 26 7 6 2 4 4 2


(CTG)n

(CTG)n


(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

CpG nrDM20


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

DM300

DMSXL

CTCFbs1

CTCFbs1(CTG)n

(CTG)n

(CTG)n

(c)









3 Results































25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


5998400 Exon 15


CTCFbs1 (CTG)n


CTCF binding sites

CTCFbs23998400


(a)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

DM20

DM300

DMSXL

0 7 5 0 0 8 2 3 10 0 0 0 0 0 0 7 0 0 0 5 0 5 5 5 8 3 0 8 0 0 5 5 2 0 6 9 18

56 29 37 49 61 37 22 25 29 25 43 33 45 27 17 20 35 31 36 45 41 28 20 33 73 4 65 6 17 7 41 28 26 35 26 47 46

89 70 79 83 95 70 52 61 66 57 69 65 80 49 48 56 66 50 83 79 84 78 67 86 82 26 70 28 42 23 53 45 67 76 48 80 90

CTCFbs2CpG nr

CpG nr

DM20

DM300

DMSXL

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

0 2 5 2 2 7 0 2 5 4 2 0 0 0 4 3 2 0 4 2 2 2 2 0 4 6 2 0 2 0

38 29 21 12 20 57 21 16 16 21 23 16 0 38 3 32 20 3 7 5 29 19 10 11 0 10 3 0 6 2

78 17 17 22 17 40 18 19 26 30 36 26 28 26 19 13 4 26 17 2 43 30 24 26 7 6 2 4 4 2


(CTG)n

(CTG)n


(b)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CTCFbs1

CpG nrDM20


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37CpG nr

DM300

DMSXL

CTCFbs1

CTCFbs1(CTG)n

(CTG)n

(CTG)n

(c)









3 Results































25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology








3 Results































25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




























25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)CTCFbs1




(a)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)


NSSNSS lowast

(b)

25

20

10

15

05

00

Qt(C

hIP)

Qt(i

nput

)


NSS NSS NSS


(c)









3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)




3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs2



3

2

1

0

Qt(C

hIP)

Qt(i

nput

)


CTCFbs1

(a)



Qt(C

hIP)

Qt(i

nput

)


CTCFbs2

10

08

06

04

02

00



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

CTCFbs1



Qt(C

hIP)

Qt(i

nput

)


08

06

04

02

00

(b)

Figure 4 Continued


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

25

20

10

15

05

00


Qt(C

hIP)

Qt(i

nput

)

CTCFbs1

NSS NSSNSS NSS

NSS NSS NSS


Enhancer region



lowastlowastlowast

(c)







8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


8

6

4

2

0

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

(a)

DM20 DM300 DMSXL


RNA

Dmpk

(au

)

25

20

15

10

05

00

(b)

RNA

Dmpk

(au

)

DM20 DM300 DMSXL

25

20

15

10

05

00


(c)






15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


15

10

05

00




Qt(C

hIP)

Qt(i

nput

)

(a)

15

10

05

00

IgG

NSS NSS


Qt(C

hIP)

Qt(i

nput

)


lowast

(b)

15

10

05

00

lowastlowastNSS


Qt(C

hIP)

Qt(i

nput

)

lowast


(c)

15

10

05

00



Qt(C

hIP)

Qt(i

nput

)lowast


(d)




4 Discussion
























Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology






















Funding


Acknowledgment


References






























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology





























































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Research Article Transcriptionally Repressive Chromatin...

Documents

Transcript of Research Article Transcriptionally Repressive Chromatin...