Integration of miRNA and Protein Profiling Reveals Coordinated Neuroadaptations in the Alcohol-Dependent Mouse Brain

of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Fashion & Beauty

Published:

Views: 64 | Pages: 13

Extension: PDF | Download: 0

Share
Description
The molecular mechanisms underlying alcohol dependence involve different neurochemical systems and are brain region-dependent. Chronic Intermittent Ethanol (CIE) procedure, combined with a Two-Bottle Choice voluntary drinking paradigm, represents one
Transcript
  Integration of miRNA and Protein Profiling RevealsCoordinated Neuroadaptations in the Alcohol-Dependent Mouse Brain Giorgio Gorini * , Yury O. Nunez, R. Dayne Mayfield Waggoner Center for Alcohol and Addiction Research, The University of Texas at Austin, Austin, Texas, United States of America Abstract The molecular mechanisms underlying alcohol dependence involve different neurochemical systems and are brain region-dependent. Chronic Intermittent Ethanol (CIE) procedure, combined with a Two-Bottle Choice voluntary drinking paradigm,represents one of the best available animal models for alcohol dependence and relapse drinking. MicroRNAs, masterregulators of the cellular transcriptome and proteome, can regulate their targets in a cooperative, combinatorial fashion,ensuring fine tuning and control over a large number of cellular functions. We analyzed cortex and midbrain microRNAexpression levels using an integrative approach to combine and relate data to previous protein profiling from the same CIE-subjected samples, and examined the significance of the data in terms of relative contribution to alcohol consumption anddependence. MicroRNA levels were significantly altered in CIE-exposed dependent mice compared with their non-dependent controls. More importantly, our integrative analysis identified modules of coexpressed microRNAs that werehighly correlated with CIE effects and predicted target genes encoding differentially expressed proteins. Coexpressed CIE-relevant proteins, in turn, were often negatively correlated with specific microRNA modules. Our results provide evidencethat microRNA-orchestrated translational imbalances are driving the behavioral transition from alcohol consumption todependence. This study represents the first attempt to combine  ex vivo  microRNA and protein expression on a global scalefrom the same mammalian brain samples. The integrative systems approach used here will improve our understanding of brain adaptive changes in response to drug abuse and suggests the potential therapeutic use of microRNAs as tools toprevent or compensate multiple neuroadaptations underlying addictive behavior. Citation:  Gorini G, Nunez YO, Mayfield RD (2013) Integration of miRNA and Protein Profiling Reveals Coordinated Neuroadaptations in the Alcohol-DependentMouse Brain. PLoS ONE 8(12): e82565. doi:10.1371/journal.pone.0082565 Editor:  Barbara Bardoni, CNRS UMR7275, France Received  July 17, 2013;  Accepted  October 24, 2013;  Published  December 16, 2013 Copyright:    2013 Gorini et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the srcinal author and source are credited. Funding:  This work was supported by NIAAA grants AA016648, AA0107838, AA019382, AA020683, and Integrative Neuroscience Initiative on Alcoholism(INIA).The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests:  The authors have declared that no competing interests exist.* E-mail: gorini@utexas.edu Introduction Cerebral cortex (CTX) and midbrain (MB) are two brainregions particularly susceptible to the effects of long-term alcoholabuse. CTX is prone to most severe brain damage in alcoholichuman brain [1–3], has major connections with the mesolimbicreward pathway [4,5], and is crucial for cognitive, executive andother important functions that are impaired in alcoholics [6,7].The development of alcoholism involves molecular alterationswithin the brain’s reward neuronal circuits [8–10], and MB withits subdomains is well known to comprise addiction-relatedpathways which are crucial for drug responses.Synergetic molecular interactions regulate neurobiologicalevents associated with complex traits responsible for alcoholdependence. Although lists of candidate molecules have beenreported based on their altered expression levels in alcohol-relatedstudies [11,12], their use is limited given the lack of contextualinformation and the documented low and inaccurate correspon-dence between transcript and protein levels [13–17]. Moreover,even small changes in the levels of certain microRNAs (miRNAs),which finely orchestrate gene and protein expression, couldextensively impact brain function. For these reasons, alcoholresearchers have recently started to focus on miRNAs in order todevelop a more integrated profile of the effects of alcohol abuse[18].miRNAs, short noncoding RNA molecules, have been effec-tively described as master regulators of the cellular transcriptomeand proteome [19]. They can regulate their target genes in acooperative, combinatorial fashion, where a single miRNA cantarget multiple mRNA transcripts and distinct miRNAs can targetthe same mRNA, ensuring fine tuning and control over a largenumber of cellular functions [20]. miRNAs are capable of inhibiting protein synthesis both by repressing translation and byfacilitating deadenylation and subsequent degradation of mRNAtargets [21]. In certain cases, miRNAs have even been reported toactivate translation of targeted mRNAs [22]. miRNAs have beenproposed as novel diagnostic biomarkers of human disease incirculating fluids such as plasma/serum [23], and there is recentevidence that they can act as signals for membrane receptoractivation [24].Previous microarray studies in human alcoholics and animalmodels have shown miRNA regulation. Differential miRNAexpression has been reported following chronic intermittentethanol exposure and withdrawal in primary cortical neuronalcultures [25], and persistent, coordinated changes in the expres- PLOS ONE | www.plosone.org 1 December 2013 | Volume 8 | Issue 12 | e82565  sion of miRNAs and their target mRNAs have also been shown inrat medial prefrontal cortex following an alcohol dependenceparadigm [26]. Furthermore, miRNA is up-regulated in thefrontal cortex of human alcoholics who spent most of their adultlives consuming high quantities of ethanol without developing complicated alcohol-related disorders, indicating a robust homeo-static adaptation to the effects of alcohol [27]. Interestingly, thismiRNA up-regulation might explain the down-regulation of certain genes in human alcoholic cases [28–32]. Indeed, recentevidences suggest that alcohol dependence results from changes inco-regulation that might not be detectable using single molecular-based analysis, since no one approach can fully account for therepercussion of individual changes on the complex interactionsthat regulate brain function. A number of studies have shown theinterconnectedness among different stages of information process-ing at the molecular level. Quantitative proteomics studies haveshown that miRNAs participate in fine-tuning the production of their targets, both at the mRNA and the protein level [33,34].Select clusters of gene expression profiles identified by arraystudies can be used to predict meaningful networks of interacting proteins [35–37]. Several components of protein complexes maybe regulated simultaneously by a single miRNA or by severalcoexpressed miRNAs, and miRNAs that target the same proteincomplexes are frequently coexpressed [38]. Finally, several studieshave demonstrated that the targets of single miRNAs are moreconnected in the protein-protein interaction (PPI) network thanexpected by chance [39–41].Since miRNAs may regulate their targets at the translationallevel, without affecting mRNA abundance, the use of proteomictechniques is crucial to identify miRNA targets and to quantify thecontribution of translational repression by miRNAs [42–45]. Ourgroup has recently reported the differential regulation in cerebralcortex (CTX) and midbrain (MB) proteomes from C57BL/6J micesubjected to a chronic intermittent ethanol (CIE), two bottle choice(2BC) paradigm [46], which represents one of the best currentlyavailable animal models for alcohol dependence and relapsedrinking. Here, we investigated the changes in global miRNAexpression levels from the same brain samples and integrate thetwo datasets to investigate the molecular mechanisms of miRNAdirect translational control during alcohol dependence. Novelsystems-biology approaches have been utilized to comprehensivelyexamine brain alterations in human alcoholics [47]. To improveour current molecular model of addiction, we used a systemsapproach to data analysis that combines miRNA and proteindifferential expression, miRNA and protein coexpression net-works, miRNA target predictions, PPIs, and gene annotations. Materials and Methods Animals and Ethics Statement Brain samples of male C57BL/6J mice subjected to ChronicIntermittent Ethanol paradigm combined with Two-Bottle Choiceethanol voluntary consumption were provided by Dr. Amanda J.Roberts (The Scripps Research Institute, La Jolla, CA), aspreviously described [46]. All procedures were conducted inaccordance with the guidelines established by the NationalInstitutes of Health in the Guide for the Care and Use of Laboratory Animals and were approved by The Scripps ResearchInstitute’s Animal Care and Use Committee (protocol: 11-0026).The paradigm used is summarized in Figure S1 and was based onearlier reports [48,49]. Blood alcohol concentrations (BACs) werein the range shown to be critical for escalated ethanol drinking [50]. Brains were collected 72 hours after the last drinking session.CTX and MB were dissected from 7 CIE-2BC ethanol vapor-exposed (alcohol-dependent, high drinkers) mice, 7 Air-2BC air-exposed matched controls (which have also had access to alcohol),plus 7 ethanol-Naı¨ ve mice. miRNA expression analysis Total RNA was isolated from the same 21 CTXs and 21 MBsanalyzed in our previous study [46] using mirVana PARIS kit (LifeTechnologies, Carlsbad, CA), according to the manufacturer’sinstructions. Yield and quality of the RNA was determined using a2100 Bioanalyzer (Agilent, Palo Alto, CA). Microarray hybridiza-tion was performed at the Moffitt Cancer Center microarrayfacility (Tampa, FL). Total RNA was hybridized with miRCURY6th generation LNA TM microRNA arrays (Exiqon, Vedbaek,Denmark) to assess miRNA expression. The arrays included 1,223human, 1,055 mouse, and 680 rat probes as referenced bymiRBase v.16, in addition to many proprietary probes. Sampleswere labeled with Cy3, while the red channel was used for qualitycontrol and reference purposes. Images were analyzed using Imagene 8.0 (BioDiscovery, Hawthorne, CA). miRNA microarraydata analysis was implemented in R environment using the LinearModels for Microarray Data (LIMMA) [51] and the WeightedGene Correlation Network Analysis (WGCNA) [52,53] packagesfrom Bioconductor (http://www.bioconductor.org).Data preprocessing included between-arrays quantile normal-ization [54] of single (green) channel, removal of flagged spots, andweighting. Background subtraction was not necessary. For eachmiRNA, a median was calculated from the intensity values of fourreplicates. Data were fitted into a linear model with an appropriatedesign matrix. Statistical differences between groups werecalculated using an empirical Bayesian approach. False discoveryrate (FDR) was assessed using the method of Benjamini andHochberg [55]. Microarray data have been deposited online atNCBI Gene Expression Omnibus (GEO [56], accession numberGSE48576). Protein expression analysis Two-dimensional differential in-gel electrophoresis (2D-DIGE)was previously used to measure protein expression levels from thesame 21 CTX and 21 MB samples. A comparative cross analysisof 28 gels (14 for CTX and 14 for MB samples) was performed asdescribed [46,57]. Briefly, 2,369 spots were detected and matched,and protein abundance values for individual samples as well asdifferential expression ratios between experimental groups werecalculated. The standardized log abundances from 1,255 gel spotswith at least 69/84 appearances were used for protein coexpres-sion analysis. Based on significance, fold change, correlations, andappearances 93 spots were selected and identified by MALDItandem mass spectrometry (MS) (Figure S2). Related MS datahave been deposited to the ProteomeXchange Consortium via thePRIDE partner repository (dataset PXD000349, DOI 10.6019/PXD000349). The resulting protein coexpression dataset wasintegrated with miRNA coexpression data from the same samplesin the present study. Weighted Gene Correlation Network Analysis (WGCNA) General information and purpose.  WGCNA is a bioinformatics toolwhich identifies significant over-represented patterns of directionalchanges in expression levels, consistently repeated across all thesamples studied. We have previously used WGCNA to analyzeprotein coexpression [46]. In the present study, WGCNA wasapplied to miRNA expression data from the same samples. Thecorrelation between miRNAs measures the degree of similaritybetween their expression patterns, and linkage hierarchicalclustering can be used to detect modules, which are groups of  MiRNA Translational Control in Alcohol DependencePLOS ONE | www.plosone.org 2 December 2013 | Volume 8 | Issue 12 | e82565  interconnected miRNAs showing over-represented patterns of coexpression. We used WGCNA to examine modules of interconnected miRNAs, as well as proteins, showing over-represented patterns of coexpression. Module Eigengenes (MEs)summarize and represent each module in one synthetic expressionprofile. We used MEs to treat modules as single units and relatethem to external information used as trait (CIE- and Air-2BCphenotypes) via simple measures (correlation).miRNA microarray normalized intensity data were subjected tocoexpression analysis implemented in R environment using ‘‘WGCNA’’ package from Bioconductor. Expression data for1,023 miRNAs were used.  Analytical procedure.  The general framework of WGCNA has beendescribed in detail previously [47,52]. We ran separate analyses formiRNAs in each region, as we have previously done with proteins[46]. First, a measure of similarity between the miRNA expressionprofiles was defined. This similarity measures the level of concordance between miRNA expression profiles across samples.Pearson’s correlations were calculated for all pairs of miRNAs, andthen a signed similarity (S ij  ) matrix was derived from S ij =(1  + cor(x i ,x  j  ))/2, where expression profiles x i  and x  j  consisted of theexpression of miRNAs ‘‘i’’ and ‘‘j’’ across multiple microarraysamples. In the signed networks, the similarity between miRNAsreflects the sign of the correlation of their expression profiles. Next,the similarity matrix was transformed into a weighted adjacencymatrix of connection strengths by using a power adjacencyfunction. The adjacency function (a ij  ) depends on certainparameters, which determine the sensitivity and specificity of thepairwise connection strengths. The signed similarity (S ij  ) was raisedto power  b  (soft thresholding) to represent the connection strength(a ij  ): a ij =S ij b  . This step aims to emphasize strong correlations andreduce the emphasis of weak correlations on an exponential scale[53]. We chose the appropriate ‘‘Soft Power’’ (  b  =10) so that theresulting networks exhibited approximate scale-free topology (thelowest power for which the scale-free topology fit index Rˆ2 . 0.8),according to the criterion proposed by Zhang and Horvath [52].Next, a topological overlap matrix (TOM) that measures therelative inter-connectedness of a pair of miRNAs in the network was built from the adjacency matrix, and the corresponding dissimilarity was obtained as d ij (a) ; 1  2  a ij . Average linkagehierarchical clustering of the topological overlap dissimilaritymatrix was then used to identify clusters of co-expressed miRNAs(modules) [52]. miRNA modules correspond to branches of thehierarchical clustering tree (dendrogram). The resulting miRNAdendrogram was used for module detection with the ‘‘DynamicTree Cut’’ method: a ‘‘cutting height’’ of 0.995 with the preset‘‘deep split’’ option (=2) were chosen to cut branches off the tree,and the resulting branches correspond to miRNA modules. The‘‘minimum module size’’ parameter was set to limit the size of thesmallest modules to at least 5 miRNAs, in order to avoid a largenumber of modules. The branches cut off of the miRNA treecorresponding to modules were labeled in unique colors.Unassigned miRNAs were labeled in gray. Relating modules to sample information.  In our analysis, we relatedmodules to CIE paradigm. As a trait, the ‘‘Escalation of Consumption’’ (EoC) trait was intended as increased ethanolconsumption, with ‘‘0’’ for the Naı¨ ve group, ‘‘1’’ for Air-2BC, and‘‘2’’ for CIE-2BC. We also used actual average ethanol drinking amounts for the last 5-days 2BC session, plus miRNA expressionand protein expression information from the same samples [46] astraits. miRNAs whose coexpression was highly and significantlycorrelated with the EoC trait were used as a trait for proteincoexpression modules and  vice versa  . Real Time PCR analysis Single-stranded cDNA was synthesized from total RNA using the TaqMan TM miRNA Reverse Transcription (Applied Biosys-tems, Foster City, CA). Following reverse transcription, quantita-tive RT-PCR (qRT-PCR) was performed in triplicate using TaqMan TM miRNA Assays (P/N: 4427975, Applied Biosystems)according to the manufacturer’s instructions. All 7 samples foreach experimental group were included in every reaction. Theidentification numbers for the single assays used are indicated inTable 1.qRT-PCR was carried out in a ViiA TM 7 Real-Time PCRSystem (Applied Biosystems), data collected using ViiA TM 7Software v. 1.2.2 (Applied Biosystems), and qRT-PCR resultsimported into qbasePLUS software v. 2.4 (Biogazelle, Zwijnaarde,Belgium) [58]. Data were normalized to the average of the bestendogenous control genes based on their M scores calculated bythe software (Table 1). Unpaired t-test with correction for multipletesting was used to assess statistical significance. Target correlationwas calculated using Pearson correlation. Functional annotations and bioinformatics tools Functional annotations of differentially expressed miRNAs wereobtained by using Ingenuity Pathway Analysis (IPA) (IngenuitySystems, www.ingenuity.com). IPA Target Filter module was usedto associate detected miRNAs with experimentally observed andpredicted mRNA targets encoding for the differentially expressedor coexpressed proteins identified with 2D-DIGE and massspectrometry from the same samples [46]. Target informationdata were filtered by considering the following: for differentialexpression data, miRNA p , 0.06 (CTX) or p , 0.05 (MB) andproteins p # 0.2 (CTX and MB); for coexpression data, miRNAand proteins p # 0.01 (CTX) or p , 0.05 (MB) and correlation $ 0.5(CTX and MB) (Table S1).Integrative networks were built by combining our differentialexpression and coexpression data with miRNA target predictionsobtained from IPA database and known/predicted PPIs from theSearch Tool for the Retrieval of Interacting Genes/Proteins(String) database v.9.05 (confidence score: 0.15, http://string-db.org/). Collected information was loaded on Cytoscape v.2.8.3(http://www.cytoscape.org/), and networks were generated andanalyzed with several topology-based scoring methods [59–63]. Results miRNA differential expression miRNA microarrays were used to measure expression profiles inthe CTX and MB of mice subjected to CIE-2BC or Air-2BC andalcohol-naı¨ ve mice. The 2BC paradigm induced significantchanges in miRNA levels with the most significant differencesbetween the 2BC treated mice versus the Naı¨ ve group (DatasetS1). When comparing CIE-2BC to Naı¨ ve, approximately 200miRNAs were differentially expressed in the CTX (p , 0.05, foldchange between 5 and 130%), and about 260 miRNAs weredifferentially expressed in MB (p , 0.05, fold change between 5 and120%). When comparing Air-2BC to Naı¨ ve, approximately 210miRNAs were differentially expressed in the CTX (p , 0.05, foldchange between 5 and 118%), and 300 miRNAs were differen-tially expressed in the MB (p , 0.05, fold change between 5 and89%). The relative heatmaps for the top 10 differentially expressedmiRNAs show complete group separation (Figure 1 B-C, E-F).The differences were smaller and less significant whencomparing CIE-2BC mice with their Air-2BC matched controls.Forty-one miRNAs were differentially expressed in the CTX(p , 0.05) with a fold change between 5 and 135% compared to 95 MiRNA Translational Control in Alcohol DependencePLOS ONE | www.plosone.org 3 December 2013 | Volume 8 | Issue 12 | e82565  miRNAs in the MB (p , 0.05) with a fold change between 5 and46% (Dataset S1). The relative heatmaps for the top 10differentially expressed miRNAs show incomplete group separa-tion (Figure 1 A, D). Over-represented functional categories Top IPA functional categories for differentially expressedmiRNAs between CIE-2BC and Air-2BC mice in CTX and MBinclude the following: cell cycle, endocrine system disorders, andinflammatory response (Table S2 A, D). Similar IPA analyses werecarried out for other group comparisons in CTX and MB (TableS2 B-C, E-H). A comparison between regions shows greateradaptations in endocrine, hematological, and inflammatorydisorders in the MB, with more differentially expressed miRNAsinvolved in related functional categories (Table S2 I). About , 10% of differentially expressed and coexpressed miRNAs arepredicted to target genes related to immune response, in bothCTX and MB. miRNA-protein overall changes The overall significant directional changes in miRNA andprotein expression, for each comparison and brain region, aresummarized in Table 2. When comparing CIE-2BC versus theNaı¨ ve group, a predominant miRNA up- and protein down-regulation was observed. On the contrary, when comparing CIE-2BC with their Air-2BC matched controls, an overall miRNAdown- and protein up-regulation was prevalent. WGCNA analysis WGCNA analysis was performed on normalized expressiondata from 1,023 miRNAs. Average linkage hierarchical clustering identified 39 distinct modules of coexpressed miRNAs in the CTXand 39 in the MB (Figure 2A, B). We related MEs (see  Methods   ) toCIE paradigm phenotypic data (‘‘EoC’’ and drinking) used as traitthrough correlation analysis. In both regions, some modules arehighly correlated with the 2BC EoC trait and the average drinking (for the last 5-days 2BC session). miRNA modules CTX10, CTX1,MB29, MB1, and MB11 were highly positively correlated(corr.=0.58-0.69), while modules CTX30, CTX35, CTX22,MB5, MB20, and MB2 were highly negatively correlated(corr. , -0.6) to the EoC trait (Figure 2E, F). A list of the top 20significantly coexpressed miRNAs is shown in Figure 2C, D.Our previous protein WGCNA analysis was integrated with thepresent miRNA coexpression analysis. Data from individual CIE-related proteins, as well as information from 19 proteincoexpression modules in the CTX and 23 in the MB (Figure 2G,H) were used.Individual proteins whose coexpression was highly and signif-icantly correlated with the EoC were used as a trait for miRNAcoexpression modules, and top correlated individual miRNAswere used as a trait for protein modules. Several proteins werehighly negatively correlated with miRNA modules important(positively correlated) for the EoC trait (i.e., distinct isoforms of DNM1L, HBB1, HS90A, DYN1, Figure 2E, F). Similarly, severalmiRNAs were highly negatively correlated with protein modulesthat in turn are positively correlated to the EoC trait, and theyoften belonged to the same miRNA module (i.e., miR-3091-3pand miR-2861 in CTX and let-7a-2-3p and miR-763 in MB,Figure 2G, H). Validation by Real Time PCR analysis To validate results from miRNA microarray analysis, weperformed qRT-PCR. A representative subset of 10 significantlydifferentially regulated miRNAs was tested. One non-significantlyregulated miRNA was also tested as control. Following qRT-PCR,five miRNAs were significantly regulated in the same direction asshown by the microarray analysis (Table 1). Five other miRNAsdid not achieve statistically significant differences, but in 4 cases wewere still able to detect the expected change in expression levels.The non-significantly regulated miRNA did also not achievestatistical significance when tested by qRT-PCR.Furthermore, when comparing the qPCR expression patterns of the different miRNAs tested across the 14 samples used, miR-200a-3p showed a 0.98 Pearson correlation (p=3.6E-8) with miR-96-5p and 0.97 correlation (p=1E-7) with miR-141-3p. Indeed,these three miRNAs were identified as coexpressed by WGCNAanalysis, belonging to the same miRNA module (CTX23). Table 1.  Results of qRT-PCR analysis. Comparison microRNAExiqonprobe IDArrayp-value Array FCTaqManassay IDRT-PCRp-value RT-PCR FC Ref. genes CTX, CIE-2BC/Naı¨ve miR-488-3p 17316 6.47E-07 1.237 001659 3.36E-03 1.267 A, B, CCTX, CIE-2BC/Naı¨ve miR-410-3p 11102 1.07E-04 1.178 001274 4.66E-03 1.241 A, B, CCTX, CIE-2BC/Naı¨ve miR-3084-3p 148484 8.84E-07 1.284 461806_mat 7.19E-03 1.416 A, B, CCTX, CIE-2BC/Naı¨ve let-7a-2-3p 42530 1.95E-04 0.772 463508_mat 9.84E-01 1.002 A, B, CCTX, CIE-2BC/Naı¨ve miR-200a-3p 11000 1.27E-02 2.295 000502 1.40E-01 2.593 B, CCTX, CIE-2BC/Naı¨ve miR-140-3p 42630 4.96E-02 0.927 002234 1.09E-02 0.572 B, CCTX, CIE-2BC/Naı¨ve miR-141-3p 10946 8.56E-03 2.017 000463 2.85E-01 2.546 B, CCTX, CIE-2BC/Naı¨ve miR-96-5p 13147 3.17E-03 1.849 000186 1.40E-01 2.622 B, CCTX, CIE-2BC/Air-2BC miR-3107-3p 42946 4.58E-02 2.352 462536_mat 3.96E-02 1.404 A, BCTX, CIE-2BC/Air-2BC miR-34b-5p 29153 8.87E-02 1.153 002617 1.66E-01 1.537 A, BCTX, CIE-2BC/Air-2BC miR-410-3p 11102 1.83E-02 1.098 001274 6.99E-01 1.11 A, BConfirmation of differential expression for selected miRNA with real-time PCR. Total RNA from cortex samples was used, and all 7 samples for each experimental groupwere included (number of datapoints per subgroup, n=7). Array p-values are based on a Bayesian two-tailed t-test, and TaqMan assays p-values are based on anunpaired t-test, corrected for multiple testing. Data were normalized to the average of the endogenous control genes indicated, based on qbasePLUS software’s Mscores.  A , snoRNA142 (TaqMan assay ID: 001231);  B , snoRNA234 (001234);  C , U6 snRNA (001973).  ID , Identification number;  FC  , fold change.  P-values in italics , p , 0.05.doi:10.1371/journal.pone.0082565.t001 MiRNA Translational Control in Alcohol DependencePLOS ONE | www.plosone.org 4 December 2013 | Volume 8 | Issue 12 | e82565  Integrative networks To distinguish key molecules involved in the escalation of ethanol consumption to dependence in the CIE paradigm, weintegrated information from miRNA and protein differentialexpression and coexpression analyses with currently availablemiRNA target predictions (IPA database, Table S1), as well asknown and predicted PPIs (String database). The resulting networks involve 48 (CTX) and 60 (MB) molecular nodes,featuring several diverse regulatory mechanisms: coexpressedmiRNAs targeting the same regulated gene (e.g., miR-532-3pand miR-339-5p on  Pea15   in the MB), genes encoding physicallyinteracting or coexpressed proteins targeted by the same miRNA(e.g., miR-494-3p on both  Dpysl2 and Dpysl3 , and miR-140-3p oncoexpressed  Flot1  and  Dnm1  in the CTX), and isoform-specific Figure 1. Hierarchical clustering of differentially expressed miRNAs from CIE-2BC ( green  ), Air-2BC ( yellow  ), and Naı ¨ ve ( grey  ) mice. Top 10 significant differentially expressed miRNAs for each comparison and brain region analyzed are shown. A-C, CTX; D-F, MB. A, D: CIE-2BC vs. Air-2BC; B, E: CIE-2BC vs. Naı¨ve; C, F: Air-2BC vs. Naı¨ve.  Rows : individual miRNAs;  columns : individual samples. Red within the heatmap represents miRNAup-regulation, and blue within the heatmap represents miRNA down-regulation. Each heat map shown contains 10 miRNAs with significantdifferential expression (p , 5E-02, p , 5E-04, p , 5E-04, p , 5E-03, p , 5E-06, p , 5E-07, respectively). Refer to Dataset S1 for p-values and fold change forindividual miRNAs. The Venn diagrams indicate the number of shared and unique differentially expressed miRNAs among comparisons: between CIE-2BC versus Air-2BC, CIE-2BC versus Naı¨ve, and Air-2BC versus Naı¨ve groups in CTX (G), MB (H), and across the two brain regions (I). Only differencesgreater than 1.05 fold with p , 0.05 (Bayesian two-tailed t-test) on mapped miRNAs eligible for IPA dataset filter are listed in Venn diagrams.doi:10.1371/journal.pone.0082565.g001MiRNA Translational Control in Alcohol DependencePLOS ONE | www.plosone.org 5 December 2013 | Volume 8 | Issue 12 | e82565
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x