UBC Faculty Research and Publications

Distinct DNA methylation patterns of cognitive impairment and trisomy 21 in down syndrome Jones, Meaghan J; Farré, Pau; McEwen, Lisa M; MacIsaac, Julia L; Watt, Kim; Neumann, Sarah M; Emberly, Eldon; Cynader, Max S; Virji-Babul, Naznin; Kobor, Michael S Dec 27, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12920_2013_Article_443.pdf [ 372.67kB ]
JSON: 52383-1.0223310.json
JSON-LD: 52383-1.0223310-ld.json
RDF/XML (Pretty): 52383-1.0223310-rdf.xml
RDF/JSON: 52383-1.0223310-rdf.json
Turtle: 52383-1.0223310-turtle.txt
N-Triples: 52383-1.0223310-rdf-ntriples.txt
Original Record: 52383-1.0223310-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessDistinct DNA methylation patterns of cognitiveimpairment and trisomy 21 in down syndromeMeaghan J Jones1, Pau Farré2, Lisa M McEwen1, Julia L MacIsaac1, Kim Watt3, Sarah M Neumann1, Eldon Emberly2,Max S Cynader4, Naznin Virji-Babul5* and Michael S Kobor1,6*AbstractBackground: The presence of an extra whole or part of chromosome 21 in people with Down syndrome (DS) isassociated with multiple neurological changes, including pathological aging that often meets the criteria forAlzheimer’s Disease (AD). In addition, trisomies have been shown to disrupt normal epigenetic marks across thegenome, perhaps in response to changes in gene dosage. We hypothesized that trisomy 21 would result in globalepigenetic changes across all participants, and that DS patients with cognitive impairment would show anadditional epigenetic signature.Methods: We therefore examined whole-genome DNA methylation in buccal epithelial cells of 10 adults with DSand 10 controls to determine whether patterns of DNA methylation were correlated with DS and/or cognitiveimpairment. In addition we examined DNA methylation at the APP gene itself, to see whether there were changesin DNA methylation in this population. Using the Illumina Infinium 450 K Human Methylation Array, we examinedmore than 485,000 CpG sites distributed across the genome in buccal epithelial cells.Results: We found 3300 CpGs to be differentially methylated between the groups, including 495 CpGs that overlapwith clusters of differentially methylated probes. In addition, we found 5 probes that were correlated with cognitivefunction including two probes in the TSC2 gene that has previously been associated with Alzheimer’s diseasepathology. We found no enrichment on chromosome 21 in either case, and targeted analysis of the APP generevealed weak evidence for epigenetic impacts related to the AD phenotype.Conclusions: Overall, our results indicated that both Trisomy 21 and cognitive impairment were associated withdistinct patterns of DNA methylation.Keywords: Down syndrome, DNA methylation, Cognitive impairment, Aging, Illumina 450k human methylation arrayBackgroundDown syndrome (DS) occurs in approximately 1 out ofevery 600 live births in the US and is the most prevalentgenetic cause of developmental disabilities [1]. It is dueto the presence of an additional whole or partial copy ofchromosome 21 resulting in developmental changes be-ginning early in fetal life. Clinical features of DS includemental retardation, stereotypical facial features, poormuscle tone, and short stature. People with DS are atincreased risk of congenital heart disease, periodontaldisease, diabetes and leukemia, and often show acceler-ated cognitive impairment with age [2-4].Postmortem studies show that from the age of 40 up-ward, individuals with DS are at much higher risk thanthe general population of having neuropathologicalchanges that meet the clinical criteria for Alzheimer’sDisease (AD) [5-7]. These include extensive cerebral at-rophy, accumulation of β-amyloid, extracellular senileplaques and intracellular neurofibrillary tangles in thehippocampus, and frontal and temporal cortices. Inaddition, functional brain imaging studies reveals spectralslowing in the brain activity of DS subjects, particularlyin bilateral temporal regions known to be associatedwith learning and memory [8]. Memory impairments are* Correspondence: naznin.virji-babul@ubc.ca; msk@cmmt.ubc.ca5Department of Physical Therapy, University of British Columbia, Vancouver,British Columbia, Canada1Centre for Molecular Medicine and Therapeutics, Child and Family ResearchInstitute, and Department of Medical Genetics, University of British Columbia,Vancouver, British Columbia, CanadaFull list of author information is available at the end of the article© 2013 Jones et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Jones et al. BMC Medical Genomics 2013, 6:58http://www.biomedcentral.com/1755-8794/6/58hypothesized to be associated with the amyloid precursorprotein gene on chromosome 21 [5,6,9]. This gene isthought to have multiple possible implications to the eti-ology of DS which overlap with AD symptoms, includingtranscriptional modulation and amyloid plaque formation[8,10,11]. Since individuals with DS have three copies ofchromosome 21, it is suspected that an overexpressionof the amyloid precursor protein contributes to the in-creased risk of AD in this population [7,12,13].One mechanism by which cells may respond tochanges in gene dosage is altered DNA methylation.DNA methylation is one of a group of epigenetic modifi-cations to the genome which affect the ability of specificgenes to be expressed but do not modify the sequence ofthe genome itself. One of the best-characterized effectsof DNA methylation is its contribution to the inactiva-tion of one entire X chromosome in females, which re-stores dosage equality with XY males. Methyl groups areadded to CpG dinucleotides, and these modifications inturn recruit chromatin remodeling complexes whichalter the structure of the surrounding chromatin, eitherincreasing or decreasing the availability for the gene tobe expressed. Changes in DNA methylation are associatedwith both normal aging and with Alzheimer’s Disease[14-16]. Additionally, the APP promoter is specificallyhypomethylated in brain tissues from AD patients [17].Previous studies examined leukocytes and fetal tissues ofDS participants for changes in DNA methylation usinglower-resolution genome-wide approaches, and found sig-nificant differences in a number of genes, as did a recentstudy examining DS placental tissue using reduced repre-sentation bisulfite sequencing [18-20].We hypothesize that altered DNA methylation onchromosome 21 and across the genome may be associ-ated with accelerated cognitive aging in DS. To that end,we examined a cohort of 10 adult participants with DSand 10 age- and sex-matched controls. We evaluatedcognitive function and collected cheek swabs from eachparticipant. Genome-wide DNA methylation patternswere analyzed using the Illumina 450K Human Methyla-tion Array, which interrogates over 480,000 CpG dinucleo-tides in the genome, including over 4000 on Chromosome21 itself, and were correlated with scores related to cogni-tive function.MethodsParticipantsThis study was approved by the University of BritishColumbia Clinical Research Ethics Board. Participantswith DS were recruited from the Down Syndrome Re-search Foundation (DSRF), in Burnaby, B.C. Informedconsent was provided by either a parent or guardian,and assent was obtained from the participant. Age andgender matched control participants were recruited fromthe staff and students at the Child and Family ResearchInstitute (CFRI) in Vancouver, B.C. A total of 20 adultsbetween the ages of 27–46 years of age, 10 with DS (5Male, 5 Female) and 10 controls (5 Male, 5 Female) par-ticipated in this study. All participants were financiallycompensated for parking and travel costs.Dalton brief praxis testThe Dalton Brief Praxis Test (BPT) is an abbreviated,20-item version of the Dyspraxia Scale for Adults withDown’s Syndrome, a 62-item cognitive test of praxis. Itscores the ability to perform simple, highly practiced,voluntary movements in response to a verbal commandor imitation, and therefore measures verbal comprehen-sion and motor co-ordination and is used to monitorchanges in cognitive function. The BPT was adminis-tered by a trained research assistant.DNA isolation and DNA methylation arraysBuccal (cheek) swabs were collected from each partici-pant using standard protocols, and DNA isolated usingthe Isohelix Buccal DNA isolation kit (Cell Projects,Kent, UK) as per manufacturer’s instructions. DNA wasthen purified and concentrated using the DNA Clean &Concentrator kit (Zymo Research, Irvine, CA, USA). Ap-proximately 750 ng of DNA was used for bisulfite con-version using the Zymo Research EZ DNA MethylationKit (Zymo Research, Irvine, CA, USA). Sample yield andpurity was assessed after each step using a NanodropND-1000 (Thermo Scientific, Irvine, CA). After bisulfiteconversion, 160 ng of DNA was applied to the Illumina450 K methylation array, as per manufacturer’s protocols(Illumina, San Diego, CA, USA).Data quality control and normalizationData was subjected to stringent quality control beforebeing normalized in R [21]. First, probes for which de-tection p-values were greater that 0.01, probes withmissing beta values, and probes for which less than threebeads contributed to the signal were eliminated in anysample (a total of 8620). Next, 11,648 probes on the Xor Y chromosome and 65 probes examining single nu-cleotide polymorphisms were removed from further ana-lysis. More recent annotation of the Human Methylation450 k array was used to filter 32,494 probes that areknown to be polymorphic at the CpG, or probes whichhave in silico nonspecific binding to the X or Y chromo-somes [22]. Together, these measures eliminated 52,827probes, leaving a total of 432,750 probes for further ana-lysis. Raw data has been deposited in GEO, accessionnumber GSE50586. Colour correction, background ad-justment, and quantile normalization were performedusing the lumi R package, and data was normalizedusing peak-based correction [23]. ComBat was used toJones et al. BMC Medical Genomics 2013, 6:58 Page 2 of 11http://www.biomedcentral.com/1755-8794/6/58remove any effects of batch from our data [24,25]. Cor-relations for two technical replicates were 0.9949 and0.9939 before ComBat, and 0.9973 and 0.9963 afterComBat, indicating minimal batch effect, which wasnonetheless corrected.Principal component analysisPrincipal Component Analysis (PCA) decomposes themeasured methylation patterns into a set of linearly in-dependent principal component (PC) patterns that areranked according to how much variance in the data theyexplain. The methylation pattern of each probe i acrossall samples, xi→, can be written as xi→¼ x þXjaij vj→ wherex is the mean profile calculated over all the probes inthe dataset. vj→are the eigenvectors (PCs), and aij are theprojection values of each probe i onto the eigenvector j.The top ranked PCs can often be correlated with knowntraits in the cohort such as tissue type, cellular compos-ition, or disease state. Because PCs are linearly inde-pendent, a particular PC’s contribution to each probe’smethylation pattern can be subtracted out without alter-ing the information contained in the pattern arisingfrom all the others. For this dataset, PCA was performedon the normalized data set twice; first on the 10 DS and10 control samples and a second time with ten add-itional blood samples added. These blood samples werefrom unrelated healthy individuals of the same approxi-mate ages. As described in more detail in the result sec-tion, the initial PCA without the blood samples revealedan unusual clustering of samples in the first PC, possiblyindicating blood contamination of the buccal swabs. Thesecond PCA including blood samples showed that in-deed, some of our buccal swabs had scores more similarto blood than other buccal swabs for PC1. Since the de-pendency of methylation on tissue represented by thatPC1 for this study is a confounder, its contribution toeach probe was subtracted out, yielding methylation datathat no longer has variation due to tissue differences. Adataset xi→were the contribution of PC number k is sub-tracted out is constructed as xi→¼ x þXj≠kaij vj→. Ourfinal data set then had 388,607 probes, since only probesfor which we had data for all samples – the 20 from ourDS/Control study plus the ten blood samples we used todetermine the tissue-related variation in our data. Den-drograms were generated using Euclidean distance.Differential methylation analysisAll statistical analysis on normalized and corrected datawas performed using R statistical software (version 3.0).Probes with DNA methylation levels significantly differ-ent between DS and control participants were identifiedfirst using the R limma package’s moderated t-tests withempirical Bayesian variance method and Benjamini-Hochberg correction to control the false discovery rateat 0.01 [26]. Significant probes were then filtered to in-clude only those that also had a beta value difference be-tween DS and control group means (Δbeta) of at least10%. This cutoff is used to eliminate probes that haverelatively small magnitude of change between groups re-gardless of their statistical significance. In addition, weused the “Bump Hunting” method from the R CHARMpackage to discover groups of probes that show differen-tial methylation, and used this list of differentially meth-ylated regions (DMRs) to identify genes that containmultiple differentially methylated probes [27].Correlations between BPT scores in DS participantsand DNA methylation levels both for all probes and forthe APP probes specifically were performed in R using a2-sided Spearman correlation, and p-values were cor-rected with Benjamini-Hochberg or Bonferroni correc-tion, as noted. For whole-genome correlations, probeswere considered significantly correlated with BPT if theBenjamini-Hochberg corrected p-values were below 0.01and the range of highest to lowest beta values wasgreater than 10%. T-test were performed using the baset-test function in R [21].All statistical analysis was performed on transformedM-values [28]. All values given in figures and the textare expressed as beta values.DAVID analysisSignificant probe accession names were input into theDAVID online GO clustering tool [29]. Only the probesused in this analysis from the Illumina 450 k HumanMethylation array were used as a background list. Consist-ent with published approaches, clusters with enrichmentscores greater than 1.3 were considered significant [29].ResultsParticipants and brief praxis scoresScores on the BPT ranged from 35 to 80 for the 10 par-ticipants with DS, and all 10 control participants re-ceived a score of 100 (Table 1). Mean age and SD werematched across groups (means 34.13 for DS and 34.5 forcontrol, SD 6.12 for DS and 6.78 for control), and a t-test p-value of 0.90 showed good matching of cohortages. Praxis scores were not significantly correlated withage in the DS cohort (Spearman’s correlation p = 0.55).Principal component analysis to eliminate sample tissuevariabilityHierarchical clustering of global DNA methylation placedthe 5 DS cases with the higher BPT scores closer to thecontrols than participants with lower scores though thedifference in scores between the groups was not statisti-cally significant (Welsh two sample t test p value = 0.07)Jones et al. BMC Medical Genomics 2013, 6:58 Page 3 of 11http://www.biomedcentral.com/1755-8794/6/58(Figure 1a). Age was also not significantly different be-tween the groups with high and low BPT scores (Welshtwo sample t test p value = 0.55).Principal component analysis (PCA) was used to de-termine the sources of variance across all samples andprobes. Initial PCA revealed that the 5 participants withlow BPT scores had markedly different scores for thefirst PC, which accounted for 64.8% of the variation,when compared to controls or the remaining DS partici-pants (Figure 1b). This amount of variance was unlikelyto be due simply to differences in cognitive impairment,since previous studies using PCA have generally identi-fied the first PC as being associated with either tissuedifferences or probe-to-probe variation [30,31]. Giventhat people with DS are at increased risk of periodontaldisease, we were concerned about blood contaminationin our collected buccal swabs [4]. We added 10 unre-lated blood samples to our PCA and repeated the ana-lysis, showing that the new first PC, which had a verysimilar shape to the previous first PC, separated tissuetypes (Figure 1c). In particular it segregated blood frombuccal samples, with the low BPT scoring DS partici-pants having intermediate scores. This could indicatethat the different scores in the original PCA between thelow and high BPT were due to the five low BPT buccalswabs having become contaminated with a small amountof blood. This could be due to the participants havingbitten their cheeks or tongue, or may simply be due tothinner epithelium and greater probability of periodontaldisease in DS participants [4].In order to analyze the true differences between theDS participants and controls, as well as the differencescorrelated with BPT, it was important to eliminate thispossible cell type-related confounder. To control forthis, we subtracted from our data set the effects of PC1from the PCA that included the blood samples (seeMethods). This eliminated the variance due to tissuecontamination while leaving any real DNA methylationdifferences between DS and control and any correlationswith cognitive impairment. After subtraction we re-peated the hierarchical clustering and noted that the dis-tance between samples was reduced, reflecting thereduced variation in the modified data set (Figure 1d).The cluster for the non-contaminated samples changedas well, likely due to the reduction in variance from allsamples; any variance associated with the blood/buccaldifference would be eliminated, and so the noise thathad been obscuring the true relationship between thesamples was removed. The five low BPT samples nowclustered with one of the other DS samples, which couldbe due either to a small remnant of the blood differencesin the data, or a true difference between the groups. Thesubtracted data was a significant improvement over theoriginal. After the PC1 subtraction, the new PC1 ac-counted for a total of 24.1% of the variance and wascorrelated with both DS and BPT score, and PC2, ac-counting for 11.2% of the variance, was correlated withDS (Additional file 1: Table S1).Differential DNA methylation between DS and controlsUsing a linear model fitting method, we found 9,982probes that were significantly different between DS andcontrol samples after Benjamini-Hochberg correction. Wethen calculated the absolute difference between means ofbeta values for DS and controls, and refer to it as Δbeta.Of the 9,982 significant probes, 3300 had a Δbeta of morethan 0.1, meaning that the mean methylation valuesbetween groups are different by more than 10%, which werefer to as differentially methylated probes (DMPs,Figure 2a, Additional file 2: Table S2). Of these, 2,190 weremore methylated in DS and 1,110 were more methylatedin controls. DMPs were distributed up to 1 Mb away fromthe nearest TSS, though hits with higher Δbeta valuestended to be found closer to TSS sites (Figure 2b). CpGisland distribution was significantly altered from the arraybackground (chi-square pval < 0.005, Figure 2c), withhigh-density CpG islands (HC) depleted and low-densityislands enriched, while IC and ICshore proportions weresimilar. Finally, we found no enrichment or depletion forTable 1 Demographics and brief praxis scores of allparticipantsSample ID SamplegroupSex Age at timeof samplingTotal briefpraxis scoreC1 Control F 30.00 100C2 Control M 28.00 100C3 Control F 45.00 100C4 Control M 47.00 100C5 Control M 33.00 100C6 Control F 30.00 100C7 Control F 38.00 100C8 Control M 30.00 100C9 Control M 29.00 100C10 Control F 35.00 100DS01 DS M 46.56 35DS02 DS F 38.40 80DS03 DS F 30.27 80DS04 DS F 40.68 74DS05 DS F 35.69 76DS06 DS M 29.55 70DS07 DS M 29.52 80DS08 DS F 30.38 78DS09 DS M 32.96 64DS10 DS M 27.29 67Jones et al. BMC Medical Genomics 2013, 6:58 Page 4 of 11http://www.biomedcentral.com/1755-8794/6/58sites mapping to chromosome 21 (Fisher two-sided exacttest p = 0.09).To determine whether we had multiple hits per gene, weused the Bump Hunter function in the CHARM R packageto find differentially methylated regions (DMRs) [27]. Wefound 495 of our DMPs were also found in Bump Hunterclusters (BH-DMPs) (Figure 2d, Additional file 3: Table S3).BH-DMPs tended to be found very close to the TSS(Figure 2e), and island distribution was also significantlyaltered from the array background (chi-square pval < 0.005,Figure 2f).Within the DMPs and BH-DMPs, we found a number ofgenes known to have functions related to the pathology ofDS. TFAP2B had 26 out of a total of 45 probes on the arraysignificantly different between DS and control, 13 of whichoverlapped with Bump Hunter clusters, and all of whichwere more methylated in DS by a range of 10-35%. Thesignificant probes are located between 1.5 kb upstream ofthe TSS to 8 kb downstream. This gene was found in aGWAS study to be associated with type 2 diabetes, and wasfurther characterized to be involved in glucose uptake andinsulin resistance in adipocytes [32,33]. DLX5 and its nearneighbor DLX6-AS had a total of 42 probes in the DMPs,with 28 of these in the BH-DMPs. DLX5 and DLX6 havebeen identified as being important in neural crest differen-tiation, including GABAergic neurons of the developingforebrain and craniofacial development, particularly jawdevelopment [34,35]. TNXB has 23 probes significantlydifferent between DS and control, all of which are alsofound in the BH-DMPs. This gene is an extracellular matrixprotein responsible for Ehlers-Danos syndrome, which hasbeen hypothesized to have clinical overlap with DS [36].Finally, CPT1B with 12 CpGs in the DMP and BH-DMPlists, is a carnitine palmitoyltransferase specifically expressedin mitochondria of skeletal muscle and associated withmetabolic syndrome and lipid deposition [37,38]. Reflectingthese interesting hits with clear linkages to the biology ofDS and AD, DAVID analysis of our DMP list showed 17significant clusters of GO terms (Additional file 4: Table S4).The top cluster had an enrichment score of 8.05 and wasadcb-0.2 0.0 0.2PC10.30.0-0.3PC2C10C1C2C3C4C5C67C8C9DS01DS02DS03DS04DS056DS078DS09DS10B1B2B3B4B5B6B7B8B9B10-0.4 -0.2 0.0 0.2PC1C10C1 C2C3C4C5C6C7C8C9S01DS02DS03DS04DS05DS06DS07DS08DS09DS100.2-0.2PC20.02000 1000 0C2DS03DS07DS05DS08C1DS02C3C4C5DS06DS04DS09DS01DS10C8C6C7C10C9600 400 200 0C2DS03DS07DS05DS08C1DS02C3C4C5DS06DS04DS09DS01DS10C8C6C7C10C90 20 40 60 80 100Figure 1 Clustering and Principal Component analysis revealed and removed contaminating epigenetic signatures from blood inbuccal swabs. a) Dendrogram of relatedness of overall patterns of DNA methylation in all 20 individuals studied. In all figure parts, colours ofparticipant codes indicate Praxis scores, as shown in scale in centre. b) PCA plot showing individual participants’ scores for PC1 (x-axis) and PC2(y-axis) without correction. c) PCA plot showing scores for PC1 and PC2 for individual participants as well as 10 blood samples. Five DS participantsclearly scored intermediate between the other buccal samples and the blood scores for PC1. d) Dendrogram of relatedness of overall patterns of DNAmethylation in all 20 individuals studied after removal of PC1 from c. Overall distances between samples were reduced.Jones et al. BMC Medical Genomics 2013, 6:58 Page 5 of 11http://www.biomedcentral.com/1755-8794/6/58related to cell adhesion, the second had an enrichmentscore of 3.85 and included terms related to protein phos-phorylation. The third cluster had a score of 2.46 and wascentered on neural development and differentiation. Ourlist of DMPs that overlap with Bump Hunter clusters didnot reveal any significant DAVID enrichment categories.Specific CpGs correlated with cognitive functionFor correlations between BPT scores in DS participantsand DNA methylation, we used two approaches. First, wecorrelated the methylation profile of each probe genomewide with BPT scores for DS participants only. A normalQ-Q plot of correlation coefficients revealed significantskewing from a normal distribution (Figure 3a). We useda Benjamini-Hochberg corrected cutoff p-value of 0.01 todetermine significance, and a total of 79 probes met thiscriteria. To eliminate probes with very little difference be-tween the DS participants or between DS and control par-ticipants, we filtered these for probes where the totalrange of methylation exceeded 10% within the DS partici-pants and which passed a t-test for differences betweenDS and control at a p-value of 0.05, for a final total of 4ba0% 20% 40% 60% 80% 100% total in hit list LC ICshore IC HC 0% 20% 40% 60% 80% 100% Total in list LC ICshore IC HC C2C3 C4 C5C1DS04DS01DS06DS03DS05DS07DS08DS02DS09DS10C6C7 C8 C9C10C2 C3C4 C5 C1DS04DS01DS06DS03DS05DS07DS08DS02DS09DS10C6 C7C8 C9 C10-1 Mb 1 Mb0 Mb-0.200.2Δ beta-1 Mb 1 Mb0 Mb-0.200.2Δ betacedfFigure 2 Epigenetic signature of T21. a) Heatmap of beta values of probes that were significantly different between DS and control participantswith a difference between the means of the groups >10%, 3300 probes total. Yellow indicates higher methylation, and blue indicates lower methylation.DS participants are shown in blue, controls in red. b) Scatterplot showing relationship of Δbeta (mean difference in beta value between DS and controlparticipants, y-axis) and distance to transcriptional start site (TSS, x-axis). Probes in red are significantly differently methylated between DS and control asin a. c) Breakdown of CpG island type in entire array (left column) and significantly different probes shown in a (right column). d) Heatmap of beta valuesof probes from a that also overlap with regions of differential methylation identified using the “Bump Hunter” method, 495 total. e) Scatterplot showingrelationship of Δbeta (mean difference in beta value between DS and control participants, y-axis) and distance to transcriptional start site (TSS, x-axis).Probes in red are significantly differently methylated between DS and control as in d. f) Breakdown of CpG island type in entire array (left column) andsignificantly different probes shown in d (right column).Jones et al. BMC Medical Genomics 2013, 6:58 Page 6 of 11http://www.biomedcentral.com/1755-8794/6/58probes (Additional file 5: Table S5). Secondly, we per-formed the same correlation test on the 3300 DMPs fromthe previous section. With the same cutoffs, five probeswere significantly correlated with BPT in DS participants,of which two had been identified in the first correlationanalysis. Combining these two lists then results in sevenprobes that were both correlated with BPT in DS partici-pants and significantly differently methylated between DSand control. Two of these were poorly correlated withBPT when all both DS and control samples were tested,leaving a total of five hits in four genes (Figure 3b-e,Additional file 5: Table S5). Two probes are found in aCpG island in the body of TSC2, which has been shownto be required for mTOR signaling in the brain, which hasbeen associated with cognitive impairment [39,40]. Thethird is located 344 bp upstream of the TSS of RND1, aRho GTPase on chromosome 12 that regulates axon ex-tension in dendritic neurons [41,42]. The final two are lo-cated 1.1 kb upstream of the TSS of KIAA1644, anuncharacterized protein on chromosome 22, and one is26 kb downstream of BICC1, involved in kidney de-velopment and located on chromosome 10. Since so fewprobes were found, functional enrichment analysis wasnot possible.Specific analysis of amyloid precursor protein (APP)No probes from APP were found to be significantly dif-ferent between DS and control or correlated with BPTin the whole-genome analysis. To be sure that significantcorrelations were not merely being lost in the multipletesting correction, we performed a targeted analysis ofthe 15 CpGs and one non-CpG site from the array thatlocalized to the APP gene (Figure 4a). We found thatfour CpGs were significantly differently methylated be-tween DS and control (t-test p-value <0.05), one of whichhad less than 5% methylation in all samples, and theremaining three were all located in intron 1 (Figure 4b).This finding correlated with previous studies that showedhypomethylation of the promoter in DS patients [17]. Wealso found two CpGs correlated with BPT scores in DSparticipants only at a BH corrected p-value of 0.05, lo-cated on either end of the APP gene (Figure 4c). Bothprobes show higher methylation with higher BPT scores,which is consistent with the model that higher levels ofAPP are found in patients with AD, but the magnitude ofdifference between individuals is small and thus its bio-logically significance is unclear.Discussion and conclusionsWhole genome DNA methylation analysis of trisomiesand cases of age-related cognitive impairment have re-vealed patterns of changes which appear to be associatedwith the etiology of each disease [18-20]. Here we add tothe few studies examining whole-genome epigenetic per-turbation in trisomies, and additionally show that thisT21-related perturbation can coexist with an epigeneticsignature associated with cognitive impairment. Com-paring individuals with DS who show behavioural evi-dence of cognitive impairment to cognitively-normal agematched controls, we have found 3300 probes whoseDNA methylation level differed by more than 10% be-tween DS and control, and 5 probes which were corre-lated with Brief Praxis scores, a measure of cognitiveimpairment. Interestingly, however, two of the probeswe found correlated with cognitive impairment in ourDS participants were found in the TSC2 gene, a compo-nent of the mTOR pathway that has been linked toAlzheimer’s Disease progression. Neither of these listsbaceTheoretical QuantilesSpearman rhoquantiles-4 -2 0 2 4-101cg14168713BICC1 (+26kb)40 60 80 100BPT score0.00.40.8Beta valuescg18649854KIAA1644 (-1kb)40 60 80 100BPT score0.00.40.8Beta valuescg08249385TSC2 (-3.7kb)40 60 80 100BPT score0.00.40.8Beta valuescg01718071TSC2 (-3.3kb)40 60 80 100BPT score0.00.40.8Beta valuescg22514963RND1 (-344bp)40 60 80 100BPT score0.00.40.8Beta valuesdFigure 3 DNA methylation sites correlated with cognitiveimpairment in DS participants. a) Normal Q-Q plot of Spearmanrho correlation coefficients for all probes. S-shape indicates small tailson the distribution of coefficients. Black points are non-significantprobes, red points are probes which survive Benjamini-Hochbergcorrection for correlation between DNA methylation and BPT in DSparticipants. b-e) BPT scores (x-axis) plotted against beta values(y-axis) for five probes found to be significantly correlated with BPTin DS participants and significantly different between DS andcontrols. Blue points indicate DS participants and red points indicatecontrols. b) RND1 c) BICC1 d) KIAA1644 and e) Two probes in TSC2.Jones et al. BMC Medical Genomics 2013, 6:58 Page 7 of 11http://www.biomedcentral.com/1755-8794/6/58was enriched for sites on chromosome 21, and targetedanalysis of the APP gene on chromosome 21 revealedweak if any evidence for epigenetic impacts on the genehypothesized to cause the Alzheimer’s Disease-like phe-notype often seen in people with DS. Particularly strik-ing in this study was the fact that we found thesesignatures in buccal epithelial cells after correction forcontaminating blood cells, highlighting both the pitfallsand potential rewards of using these easily-accessibletissues.Comparing our results with previous studies examin-ing DNA methylation in DS samples, we found an overlapin two of seven genes (TSC2 and DIO3) that are differ-ently methylated in fetal trisomy 21 skin and muscle, andfive genes (RYK, CASP10, MAP2K6, MSTR1, and RARA)that are differentially methylated in trisomy 18 skin [19].In another study using adult samples and parallel mea-surements of approximately 28,000 probes, 118 probeswere found to be significantly different between thegroups [18]. Interestingly, our data overlaps with 18 of thespecific probes they found to be different in lym-phocytes, and an additional 14 genes with differentprobes (Additional file 6: Table S6). For those probeswhich overlapped exactly, the direction of differencein DNA methylation between DS and control was thesame in both studies [18]. A recent study used re-duced representation bisulfite sequencing to examineDNA methylation in DS placenta. They found 629sites in 597 unique genes that were significantly dif-ferent between DS and control placenta. Of these, 93genes were also found to differ between DS and con-trol in our analysis, of which three HOXA2, CPT1B,and GRM6 were found in all three studies (Additionalfile 7: Table S7) [18,20]. These sites must then be-come differently methylated in people with DS veryearly in development, since the placental tissue usedin the latter study is of extraembryonic origin. It isreassuring that despite using different tissues andtechnologies, and considering that all three studieshad relatively small cohorts, similar genes were foundacross studies. Also similarly to the two latter studies,we observed a bias towards hypermethylation in DScompared to control, with more than 66% of our sig-nificantly different probes being hypermethylated inDS [18,20].Tissue composition differences are a bane of epigenet-ics studies, and thus it is important to continue to de-velop methods such as the one presented here to controlfor these differences. Signatures of different cell types,even within a given tissue, can easily mask true differ-ences between groups [30,43], or, as we have shown,spuriously contribute to differences which may trulyexist. Given that buccal swabs are a popular choice forpopulation epigenetic studies because of their ease of1 2 103 4 5 6 7 8 9 11 12 13 14 15APPabA7:cg1286133 9:cg185974218:cg19788250Control DS0.750.850.95Beta valueControl DS Control DSc 15:cg2615395440 60 80 100BPT score1:cg2100099940 60 80 1000.0BPT scoreBetasBetas0. 4 Targeted analysis of APP. a) Structure of the APP gene showing the locations of the 15 CpGs (numbers) and one CpH (letter A)analyzed in this study. Exons are shown as black boxes, intronic regions as lines. Grey box indicates position of the CpG island. b) Boxplots ofthree probes found to be significantly differently methylated between DS and control. CpG identification number as well as numbered positionfrom A are indicated. c) Scatterplots of beta values (y-axis) versus Brief Praxis scores (X-axis) for the two probes which are significantly correlatedwith BPT. Blue points indicate DS participants and red points indicate controls.Jones et al. BMC Medical Genomics 2013, 6:58 Page 8 of 11http://www.biomedcentral.com/1755-8794/6/58collection and storage, it will be important for the re-search community to begin looking for and correctingthese types of differences. A recent study examined buc-cal and blood whole-genome methylation, and deter-mined that buccal cells cluster together with many othertissues, while blood methylation patterns are very dis-tinct [44]. This makes correcting for potential bloodcontamination in buccal swabs especially important,since blood contamination would have a greater effecton buccal methylation profiles than other tissues. Ourapproach of using principal component analysis allowedus to robustly remove the variation caused by these tis-sue differences without removing entire probes or sam-ples from the analysis. One potential problem is that wemay be removing more variation than is required; ifsome true variation has the same projections as the tis-sue differences, we may be losing it as well. On the otherhand, the current analysis gave very robust and unam-biguous results after this correction, so any improperlylost variation would have had a minor contribution tothe results.The fact that an epigenetic signature that includesgenes with functions related to the biology of DS isstrong support for validity of our approach. The top fourCpG clusters as mentioned in the Results section wererelated to diabetes, lipid metabolism, neural crest andcraniofacial development, and connective tissue, all ofwhich are connected to clinical features of DS. OurDAVID analysis further supported this, with enrichmentclusters overlapping these functions as well as regulationof apoptosis and skeletal development. The massive en-richment for genes involving adhesion in our hit list fordifferentially methylated probes is interesting as, while asingle CpG site in the Down Syndrome Cell AdhesionMolecule (DSCAM) gene is present on our hit list, therewere 106 total adhesion-related genes in our DAVID hitlist. DSCAM has been proposed to contribute to thecongenital heart disease feature of DS, and it is possiblethat mis-regulation of a large number of adhesion-related genes through epigenetic modification may ex-plain the increased risk in the DS population [45]. It isalso notable that we found probes which differed be-tween DS and control across all chromosomes. The lackof enrichment for changes in DNA methylation ofprobes on chromosome 21 is perhaps counterintuitive,despite the fact that it is supported by previous studiesshowing similar results [18,19]. It could be anticipatedthat having three copies of a chromosome would resultin specific epigenetic modifications to attempt and con-trol for dosage across the trisomy. Unfortunately the DSparticipants in our study were not karyotyped, so thatwe do not know the extent of their trisomy 21. Giventhe differences observed between DS and control, how-ever, it is clear that the presence of at least a part of anextra chromosome 21 is capable of causing these genome-wide epigenetic changes.Given that only five probes in four genes were foundto be both correlated with BPT scores in DS participantsand different between DS and controls, it is impossibleto assess functional enrichment. Two of these fourgenes, however, have known functions in neural develop-ment or degeneration. RND1 is a Rho GTPas that func-tions to promote dendritic cell growth and axon guidance,and loss of RND1 expression suppresses axon formationin hippocampal neurons [41,42]. We found that methyla-tion of RND1 was positively correlated with BPT, meaningthat lower levels of DNA methylation were associated withmore severe cognitive impairment. If DNA methylationsuppresses RND1 expression, the opposite pattern mightbe expected in DS participants, but without knowing themechanistic relationship between RND1 expression andDNA methylation, it is difficult to interpret. The othergene, TSC2, had two CpGs in an island that overlaps withexons 27, 28, and 29 that showed decreasing DNA methy-lation with cognitive impairment. TSC2 is involved inmTOR signaling, which regulates levels of tau in mouseneuronal models, resulting in neuropathological symp-toms which are alleviated when mTOR signaling is re-duced [39,40,46,47]. Levels of tuberin protein, the TSC2gene product, are decreased in brain samples of adultmales with DS and AD-like symptoms, and a mouseknockdown of Tsc2 showed an increase in tau-positiveaxon formation in hippocampal neurons [48,49]. In cancercells, demethylation of the promoter of TSC2 was shownto result in an increase in expression [50]. On the surface,this methylation data appears conflicting, however the is-land in which our two target CpGs are found is in thebody of the gene, and island CpGs in gene bodies can beeither positively or negatively correlated with gene expres-sion [51]. We can therefore hypothesize that decreasedmethylation of TSC2 in DS participants is associated withthe decreased protein expression, which results in an in-creased probability of accumulation of tau through mTORregulation, which predisposes these patients to AD-likedisease. Thus RND1 and especially TSC2 are interestingtargets for future studies of epigenetic alterations in cogni-tive impairment.Cross-sectional studies such as these are important todiscover associations between DNA methylation andcognitive impairment. Given our relatively small samplesize, we were surprised to find such clear differences be-tween our study groups, but in the future, it will be im-portant to perform larger and more powerful longitudinalstudies on patients with DS and AD to track the dynamicsof DNA methylation changes during cognitive decline.Post-mortem brain tissue analysis would also shed furtherlight on the relationship between epigenetics and cognitivedecline in these patients. Together, these data will helpJones et al. BMC Medical Genomics 2013, 6:58 Page 9 of 11http://www.biomedcentral.com/1755-8794/6/58illuminate whether DNA methylation is changed withcognition, whether altered DNA methylation is a predict-ive factor for cognitive decline, or whether the two pro-cesses occur simultaneously.Additional filesAdditional file 1: Table S1. Table of correlation coefficients andp-values for correlations of PCs after subtraction of PC1 and BPT scores orpresence of DS.Additional file 2: Table S2. Table of CpG probes that are significantlydifferently methylated between DS and control participants.Additional file 3: Table S3. Table of CpG probes that are significantlydifferently methylated between DS and control participants and overlapwith clusters found to be significantly differently methylated by BumpHunter.Additional file 4: Table S4. DAVID results for probes that aresignificantly differently methylated between DS and control participants.Additional file 5: Table S5. Table of probes that are significantlycorrelated with BPT.Additional file 6: Table S6. Overlap of our DMPs with DMPs from theKerkel study [18].Additional file 7: Table S7. Overlap of our DMPs with DMPs from theJin study [20].AbbreviationsDS: Down syndome; AD: Alzheimer’s disease; GO: Gene ontology; BPT: Briefpraxis test; PCA: Principal component analysis; DMP: Differentially methylatedprobe; DMR: Differentially methylated region; BH-DMP: Bump-hunterdifferentially methylated probe.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsMJJ performed the data analysis and drafted the manuscript. PF performedthe PCA and PC1 subtraction. KW collected the samples and performed theBPT assessment. SN, LMM, and JLH processed the samples, performed thearrays, and did preliminary data analysis. EE conceived the PCA analysis andassisted with statistical analysis. MC participated in the study design andedited the manuscript. NVB conceived the study and design, recruited theparticipants, and contributed to analysis and interpretation of the data. MSKconceived the study and design, and helped draft the manuscript. Allauthors read and approved the final manuscript.AcknowledgementsThe authors would like to thank all the participants who took part in thisstudy, and the Down Syndrome Research Foundation for their support. Thisstudy was supported by the Brain Research Centre, University of BritishColumbia. MJJ was supported by a Mining for Miracles fellowship from theChild and Family Research Institute. MSK is a Senior Fellow of the CanadianInstitute for Advanced Research and a Scholar of the MowafaghianFoundation.Author details1Centre for Molecular Medicine and Therapeutics, Child and Family ResearchInstitute, and Department of Medical Genetics, University of British Columbia,Vancouver, British Columbia, Canada. 2Department of Physics, Simon FraserUniversity, Burnaby, British Columbia, Canada. 3Department of Psychology,Simon Fraser University, Burnaby, British Columbia, Canada. 4Brain ResearchCentre, University of British Columbia, Vancouver, British Columbia, Canada.5Department of Physical Therapy, University of British Columbia, Vancouver,British Columbia, Canada. 6Human Early Learning Partnership, School ofPopulation and Public Health, University of British Columbia, Vancouver,British Columbia, Canada.Received: 27 August 2013 Accepted: 19 December 2013Published: 27 December 2013References1. Parker SE, Mai CT, Canfield MA, Rickard R, Wang Y, Meyer RE, Anderson P,Mason CA, Collins JS, Kirby RS, Correa A, National Birth Defects PreventionNetwork: Updated National Birth Prevalence estimates for selected birthdefects in the United States, 2004–2006. Birth Defects Res Part A Clin MolTeratol 2010, 88:1008–1016.2. Van Goor JC, Massa GG, Hirasing R: Increased incidence and prevalence ofdiabetes mellitus in Down’s syndrome. Arch Dis Child 1997, 77:183.3. Loudin MG, Wang J, Leung H-CE, Gurusiddappa S, Meyer J, Condos G,Morrison D, Tsimelzon A, Devidas M, Heerema NA, Carroll AJ, Plon SE,Hunger SP, Basso G, Pession A, Bhojwani D, Carroll WL, Rabin KR: Genomicprofiling in Down syndrome acute lymphoblastic leukemia identifieshistone gene deletions associated with altered methylation profiles.Leukemia 2011, 25:1555–1563.4. Sakellari D, Arapostathis KN, Konstantinidis A: Periodontal conditions andsubgingival microflora in down syndrome patients - a case–controlstudy. J Clin Periodontol 2005, 32:684–690.5. Mann DM, Yates PO, Marcyniuk B: Alzheimer “s presenile dementia, seniledementia of Alzheimer type and Down” s syndrome in middle age forman age related continuum of pathological changes. Neuropathol ApplNeurobiol 1984, 10:185–207.6. Wisneiwski KE, Wisniewski HM, Wen GY: Occurrence of neuropathologicalchanges and dementia of Alzheimer“s disease in Down”s syndrome.Ann Neurol 1985, 17:278–282.7. Zigman WB: Atypical aging in down syndrome. Dev Disabil Res Rev 2013,18:51–67.8. Ismail S, Sun W, Nathoo FS, Babul A, Moiseev A, Beg MF, Virji-Babul N: ASkew-t space-varying regression model for the spectral analysis ofresting state brain activity. Stat Methods Med Res 2013, 22:424–438.9. Rumble B, Retallack R, Hilbich C, Simms G, Multhaup G, Martins R, Hockey A,Montgomery P, Beyreuther K, Masters CL: Amyloid A4 protein and itsprecursor in Down“s syndrome and Alzheimer”s disease. N Engl J Med1989, 320:1446–1452.10. Masters CL, Simms G, Weinman NA, Multhaup G, McDonald BL, BeyreutherK: Amyloid plaque core protein in Alzheimer disease and Downsyndrome. Multe Values Selected 1985, 82:4245–4249.11. Trazzi S, Mitrugno VM, Valli E, Fuchs C, Rizzi S, Guidi S, Perini G, Bartesaghi R,Ciani E: APP-dependent up-regulation of Ptch1 underlies proliferationimpairment of neural precursors in Down syndrome. Hum Mol Genet2011, 20:1560–1573.12. Hof PR, Bouras C, Perl DP, Sparks DL, Mehta N, Morrisson JH: Age-relateddistribution of neuropathologic changes in the cerebral cortex ofpatients with down“s syndrome. Quantitative regional analysis andcomparison with alzheimer”s disease. Arch Neurol 1995, 52:379–391.13. Hyman BT, West HL, Rebeck GW, Lai F, Mann D: Neuropathologicalchanges in Down’s syndrome hippocampal formation: effect of age andapolipoprotein E genotype. Arch Neurol 1995, 52:373.14. Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B,Bibikova M, Fan J-B, Gao Y, Deconde R, Chen M, Rajapakse I, Friend S, IdekerT, Zhang K: Genome-wide methylation profiles reveal quantitative viewsof human aging rates. Mol Cell 2013, 49:359–367.15. Liu L, Van Groen T, Kadish I, Tollefsbol TO: DNA methylation impacts onlearning and memory in aging. Neurobiol Aging 2009, 30:549–560.16. Bakulski KM, Dolinoy DC, Sartor MA, Paulson HL, Konen JR, Lieberman AP,Albin RL, Hu H, Rozek LS: Genome-wide DNA methylation differencesbetween late-onset Alzheimer’s disease and cognitively normal controlsin human frontal cortex. J Alzheimers Dis 2012, 29:571–588.17. West RL, Lee JM, Maroun LE: Hypomethylation of the amyloid precursorprotein gene in the brain of an Alzheimer’s disease patient. J MolNeurosci 1995, 6:141–146.18. Kerkel K, Schupf N, Hatta K, Pang D, Salas M, Kratz A, Minden M, Murty V,Zigman WB, Mayeux RP, Jenkins EC, Torkamani A, Schork NJ, Silverman W,Croy BA, Tycko B: Altered DNA methylation in leukocytes with trisomy 21.PLoS Genet 2010, 6:e1001212.19. Yuen RK, Neumann SM, Fok AK, Penaherrera MS, McFadden DE, RobinsonWP, Kobor MS: Extensive epigenetic reprogramming in human somatictissues between fetus and adult. Epigenetics Chromatin 2011, 4:7.Jones et al. BMC Medical Genomics 2013, 6:58 Page 10 of 11http://www.biomedcentral.com/1755-8794/6/5820. Jin S, Lee YK, Lim YC, Zheng Z, Lin XM, Ng DPY, Holbrook JD, Law HY, KwekKYC, Yeo GSH, Ding C: Global DNA hypermethylation in down syndromeplacenta. PLoS Genet 2013, 9:e1003515.21. R Development Core Team: R: a Language and Environment for StatisticalComputing. Vienna, Austria: R Foundation for Statistical Computing; 2008.22. Price ME, Cotton AM, Lam LL, Farré P, Emberly E, Brown CJ, Robinson WP,Kobor MS: Additional annotation enhances potential for biologically-relevant analysis of the Illumina Infinium HumanMethylation450BeadChip array. Epigenetics Chromatin 2013, 6:4.23. Dedeurwaerder S, Defrance M, Calonne E, Denis H, Sotiriou C, Fuks F:Evaluation of the infinium methylation 450 K technology. Epigenomics2011, 3:771–784.24. Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD: The sva package forremoving batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 2012, 28:882–883.25. Johnson WE, Li C, Rabinovic A: Adjusting batch effects in microarrayexpression data using empirical Bayes methods. Biostatistics 2006, 8:118–127.26. Smyth GK: Linear models and empirical bayes methods for assessingdifferential expression in microarray experiments. Stat Appl Genet Mol Biol2004, 3:1 Article3.27. Irizarry RA, Ladd-Acosta C, Carvalho B, Wu H, Brandenburg SA, Jeddeloh JA,Wen B, Feinberg AP: Comprehensive high-throughput arrays for relativemethylation (CHARM). Genome Res 2008, 18:780–790.28. Du P, Zhang X, Huang C-C, Jafari N, Kibbe WA, Hou L, Lin SM: Comparisonof Beta-value and M-value methods for quantifying methylation levelsby microarray analysis. BMC Bioinforma 2010, 11:587.29. Huang DW, Sherman BT, Lempicki RA: Systematic and integrative analysisof large gene lists using DAVID bioinformatics resources. Nat Protoc 2008,4:44–57.30. Lam LL, Emberly E, Fraser HB, Neumann SM, Chen E, Miller GE, Kobor MS:Factors underlying variable DNA methylation in a human communitycohort. Proc Natl Acad Sci U S A 2012, 109(Suppl 2):17253–17260.31. Ziller MJ, Gu H, Müller F, Donaghey J, Tsai LTY, Kohlbacher O, De Jager PL,Rosen ED, Bennett DA, Bernstein BE, Gnirke A, Meissner A: Charting adynamic DNA methylation landscape of the human genome. Nature2013, 500:477–481.32. Maeda S, Tsukada S, Kanazawa A, Sekine A, Tsunoda T, Koya D, Maegawa H,Kashiwagi A, Babazono T, Matsuda M, Tanaka Y, Fujioka T, Hirose H, EguchiT, Ohno Y, Groves CJ, Hattersley AT, Hitman GA, Walker M, Kaku K, IwamotoY, Kawamori R, Kikkawa R, Kamatani N, McCarthy MI, Nakamura Y: Geneticvariations in the gene encoding TFAP2B are associated with type 2diabetes mellitus. J Hum Genet 2005, 50:283–292.33. Tao Y, Maegawa H, Ugi S, Ikeda K, Nagai Y, Egawa K, Nakamura T, Tsukada S,Nishio Y, Maeda S, Kashiwagi A: The transcription factor AP-2{beta} causescell enlargement and insulin resistance in 3 T3-L1 adipocytes.Endocrinology 2006, 147:1685–1696.34. Stühmer T, Anderson SA, Ekker M, Rubenstein JLR: Ectopic expression ofthe Dlx genes induces glutamic acid decarboxylase and Dlx expression.Development 2002, 129:245–252.35. Kraus P, Lufkin T: Dlx homeobox gene control of mammalian limb andcraniofacial development. Am J Med Genet A 2006, 140A:1366–1374.36. Pasmatzi E, Vlastos D, Monastirli A, Stephanou G, Georgious S, Sakkis T, TsambaosD: Ehlers-Danlos type IV syndrome in a patient with down syndrome: simpleco-occurrence or true association? Am J Med Sci 2006, 331:48–50.37. Shi W, Hu S, Wang W, Zhou X, Qiu W: Skeletal muscle-specific CPT1deficiency elevates lipotoxic intermediates but preserves insulin sensitivity.J Diabetes Res 2013, 2013:163062.38. Auinger A, Rubin D, Sabandal M, Helwig U, Rüther A, Schreiber S, FoelschUR, Döring F, Schrezenmeir J: A common haplotype of carnitinepalmitoyltransferase 1b is associated with the metabolic syndrome.Br J Nutr 2013, 109:810–815.39. Huang J, Dibble CC, Matsuzaki M, Manning BD: The TSC1-TSC2 complex isrequired for proper activation of mTOR complex 2. Mol Cell Biol 2008,28:4104–4115.40. Caccamo A, Magrì A, Medina DX, Wisely EV, López-Aranda MF, Silva AJ,Oddo S: mTOR regulates tau phosphorylation and degradation:implications for Alzheimer’s disease and other tauopathies. Aging Cell2013, 12:370–380.41. Li Y-H, Ghavampur S, Bondallaz P, Will L, Grenningloh G, Püschel AW: Rnd1regulates axon extension by enhancing the microtubule destabilizingactivity of SCG10. J Biol Chem 2009, 284:363–371.42. Ishikawa Y, Katoh H, Negishi M: Small GTPase Rnd1 is involved inneuronal activity-dependent dendritic development in hippocampalneurons. Neurosci Lett 2006, 400:218–223.43. Liu Y, Aryee MJ, Padyukov L, Fallin MD, Hesselberg E, Runarsson A, Reinius L,Acevedo N, Taub M, Ronninger M, Shchetynsky K, Scheynius A, Kere J,Alfredsson L, Klareskog L, Ekström TJ, Feinberg AP: Epigenome-wideassociation data implicate DNA methylation as an intermediary ofgenetic risk in rheumatoid arthritis. Nat Biotechnol 2013, 31:142–147.44. Lowe R, Gemma C, Beyan H, Hawa MI, Bazeos A, Leslie RD, Montpetit A,Rakyan VK, Ramagopalan SV: Buccals are likely to be a more informativesurrogate tissue than blood for epigenome-wide association studies.Epigenetics 2013, 8:445–454.45. Barlow GM, Chen XN, Shi ZY, Lyons GE, Kurnit DM, Celle L, Spinner NB,Zackai E, Pettenati MJ, Van Riper AJ, Vekemans MJ, Mjaatvedt CH, KorenbergJR: Down syndrome congenital heart disease: a narrowed region and acandidate gene. Genet Med 2001, 3:91–101.46. Habib SL, Michel D, Masliah E, Thomas B, Ko HS, Dawson TM, Abboud H,Clark RA, Imam SZ: Role of tuberin in neuronal degeneration. NeurochemRes 2008, 33:1113–1116.47. Chong ZZ, Shang YC, Wang S, Maiese K: Shedding new light onneurodegenerative diseases through the mammalian target ofrapamycin. Prog Neurobiol 2012, 99:128–148.48. Ferrando-Miguel R, Rosner M, Freilinger A, Lubec G, Hengstschläger M:Tuberin–a new molecular target in Alzheimer's disease? Neurochem Res2005, 30:1413–1419.49. Choi YJ, Di Nardo A, Kramvis I, Meikle L, Kwiatkowski DJ, Sahin M, He X:Tuberous sclerosis complex proteins control axon formation. Genes Dev2008, 22:2485–2495.50. Chakraborty S, Mohiyuddin SMA, Gopinath KS, Kumar A: Involvement ofTSC genes and differential expression of other members of the mTORsignaling pathway in oral squamous cell carcinoma. BMC Cancer 2008,8:163.51. Varley KE, Gertz J, Bowling KM, Parker SL, Reddy TE, Pauli-Behn F, Cross MK,Williams BA, Stamatoyannopoulos JA, Crawford GE, Absher DM, Wold BJ,Myers RM: Dynamic DNA methylation across diverse human cell linesand tissues. Genome Res 2013, 23:555–567.doi:10.1186/1755-8794-6-58Cite this article as: Jones et al.: Distinct DNA methylation patterns ofcognitive impairment and trisomy 21 in down syndrome. BMC MedicalGenomics 2013 6:58.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitJones et al. BMC Medical Genomics 2013, 6:58 Page 11 of 11http://www.biomedcentral.com/1755-8794/6/58


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items