SEARCH

SEARCH BY CITATION

Keywords:

  • cocaine;
  • lncRNA;
  • nucleus accumbens;
  • transcriptome

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Cocaine dependence involves in the brain's reward circuit as well as nucleus accumbens (NAc), a key region of the mesolimbic dopamine pathway. Many studies have documented altered expression of genes and identified transcription factor networks and epigenetic processes that are fundamental to cocaine addiction. However, all these investigations have focused on mRNA of encoding genes, which may not always reflect the involvement of long non-coding RNAs (lncRNAs), which has been implied in a broad range of biological processes and complex diseases including brain development and neuropathological process. To explore the potential involvement of lncRNAs in drug addiction, which is viewed as a form of aberrant neuroplasticity, we used a custom-designed microarray to examine the expression profiles of mRNAs and lncRNAs in brain NAc of cocaine-conditioned mice and identified 764 mRNAs, and 603 lncRNAs were differentially expressed. Candidate lncRNAs were identified for further genomic context characterization as sense-overlap, antisense-overlap, intergenic, bidirection, and ultra-conserved region encoding lncRNAs. We found that 410 candidate lncRNAs which have been reported to act in cis or trans to their targeted loci, providing 48 pair mRNA-lncRNAs. These results suggest that the modification of mRNAs expression by cocaine may be associated with the actions of lncRNAs. Taken together, our results show that cocaine can cause the genome-wide alterations of lncRNAs expressed in NAc, and some of these modified RNA transcripts may to play a role in cocaine-induced neural plasticity and addiction.

Abbreviations used
CPP

conditioned place preference

GO

gene ontology

lncRNAs

long non-coding RNAs

NAc

nucleus accumbens

UCR

ultra-conserved region

Cocaine is one of the most widely abused drugs, which pose serious social, medical, and economical problems (Nestler 2004). Repeated use of cocaine causes long-lasting changes in the brain's reward circuitry, a crucial component of which is the nucleus accumbens (NAc) (Nestler 2005; Hyman et al. 2006). Cocaine triggers cellular and molecular alterations, and produces long-lasting effects, leading to stable changes in neuroplasticity in NAc (Kauer and Malenka 2007; McClung and Nestler 2007). Recently, many studies have documented altered expression of genes (McClung and Nestler 2003; Albertson et al. 2004; Mash et al. 2007), and identified transcription factor networks and epigenetic processes that are fundamental to depress cocaine addiction (Renthal et al. 2009; Hollander et al. 2010; Malvaez et al. 2010).

Although only a small proportion (~2%) of mammalian genome encodes proteins, a much greater proportion is transcribed as non-coding RNAs (ncRNAs) (Ersona and Pettyb 2008; Mercer et al. 2008a; Wilusz et al. 2009). Over the past decade, most attention has been devoted to small ncRNAs like microRNAs (~21–25 nucleotides), which regulate gene expression at post-transcriptional or/and transcriptional level. For example, striatal miR-212 protects against the development of compulsive drug-taking, and plays a key role in determining vulnerability to cocaine addiction (Hollander et al. 2010; Im et al. 2010).

The mammalian genome encodes thousands of long non-coding RNAs (lncRNAs), which have been shown to play an important role in development, cellular processes, and chromatin regulation. The lncRNAs are defined as > 200 nt long transcripts that do not contain a functional open reading frame, they do not encode any protein product, but rather seem to act as a RNA molecule. A role of lncRNAs controls the expression of genes through both cis- and trans-acting pathways (Johnson 2011). So, lncRNAs have been categorized in terms of localization relative to protein-coding genes, and likely influence brain development, neurotransmission, and neuropsychiatric disorders (Mercer et al. 2008b; Qureshi et al. 2010). In neuronal cells in particular, there stricted expression of ncRNAs especially during development has previously been suggested to have important functions including human brain evolution, synaptogenesis, and other regulatory functions (Pollard et al. 2006; Mehler and Mattick 2007). For instance, one lncRNA, namely BACE1-AS, has a role in regulating BACE1 and in driving Alzheimer's disease (Faghihi et al. 2008); another lncRNA, Malat1, regulates synapse formation by modulating the expression of genes involved in synapse formation and/or maintenance (Bernard et al. 2010). However, lncRNAs are largely unknown in the field of cocaine-induced addiction and neuroplasticity.

We adopted a transcriptomics-based approach to investigate the expression patterns of lncRNAs and mRNAs of the NAc in cocaine-conditioned mice. Using conservative thresholds for statistical and biological significance, we identified 764 mRNAs and 603 lncRNAs, providing 48 pair mRNA-lncRNAs significantly modified. We propose that these lncRNAs are likely to confer important functions cocaine-evoked neuroplasticity and addiction. Our observations will be informative and serves as a resource for future studies.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Animals and drugs

Animals used in these experiments were 8-week-old male C57BL/6J mice. Mice were maintained in a temperature-controlled vivarium (21 ± 2°C) under a 12 : 12 h light-dark cycle and housed four per cage. Food and fluid were available ad libitum. The animals were acclimatized for 7 days before experiment and habituated to handling for at least 3 days before behavioral testing. All animal protocols in this study were in accordance with the guidelines established by the Association for Assessment and Accreditation of Laboratory Animal Care. The protocols were approved by the Institutional Animal Care and Use Committee of the Institute (protocol number IACUC-S200904-P001).

The test substance cocaine-HCl was provided from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Cocaine was dissolved in 0.9% saline.

Conditioned place preference procedure

Conditioned place preference (CPP) test in mice was done as described by Valjent et al. (2006). Briefly, place conditioning studies were conducted using a shuttle box that consisted of three chambers: two distinct large conditioning chambers and a small neutral start chamber. During pre-conditioning phase (day 1, pre-test), mice were placed in the central area and allowed to explore both compartments freely for 15 min. The time spent in each chamber was recorded, and unbiased animals were randomly assigned to two groups, control group, and cocaine group. After injection of cocaine in cocaine group (20 mg/kg, i.p.) and injection of saline in control group (0.9% sodium chloride) on days 2, 4, and 6, mice were confined to the corresponding conditioning compartment by closing the removable wall for a period of 15 min. After injection of saline in both cocaine group and saline group on days 3, 5, and 7, mice were confined to the opposite conditioning chamber for the same time. In post-conditioning phase (day 8), mice were placed as in the central area with free access to both compartments for 15 min, and the time spent in each compartment was measured (see Fig. 1a). Results were expressed as the time spent in the cocaine-paired chamber minus the time spent in the saline-paired chamber during CPP testing.

image

Figure 1. Exposure to cocaine in drug-paired compartment inducing Conditioned place preference. (a) Experimental design: S, saline; C, cocaine. Each group included four mice. (b) After conditioning, mice developed a significant preference for the cocaine-paired side. *Significant Conditioned place preference scores were defined as with p < 0.05.

Download figure to PowerPoint

Preparation of brain samples and RNA isolation

After the end of each CPP testing, the mice were killed by rapid decapitation without anesthesia 2 h. The NAc was removed from the brain, snap-frozen in liquid nitrogen, and stored at −80°C until analysis. The NAc tissues were washed three times with cold phosphate-buffered saline and were scraped into Trizol reagent (Invitrogen, Carlsbad, CA, USA). Total RNA was isolated using the RNeasy kit (Qiagen, Valencia, CA, USA) according to manufacturer's instructions, including a DNase digestion step. After having passed RNA measurement on the Nanodrop ND-1000 (Nanodrop Technologies, Wilmington, DE, USA) by measuring absorbance (A280, A260, A230), RNA concentrations and quantity, and then denaturing gel electrophoresis. The samples were amplified and labeled using the Agilent Quick Amp (Agilent Technologies, Santa Clara, CA, USA) labeling kit and hybridized with Agilent whole genome oligomicroarray in Agilent's SureHyb Hybridization Chambers. After hybridization and washing, the processed slides were scanned with the Agilent DNA microarray scanner (part number G2505B) using settings recommended by Agilent Technologies.

Microarray analysis

Raw image data were converted to CEL and pivot files using Agilent Feature Extraction Software (version 10.5.1.1). All downstream analysis of microarray data was performed using the Agilent GeneSpring GX software (version 11.0). The eight microarray data sets were normalized in GeneSpring GX using the Agilent FE one-color scenario (mainly median normalization), The positive effect of this median normalization is illustrated in Box-plot (see example in Fig. 1 below), and genes marked present in one of eight samples (‘All Targets Value’) were chosen for data analysis. Genes were considered ‘present’ with all present using Agilent Feature Extraction Software, an intensity > 80. Log 2-transformed data were normalized using quantile normalization and used for comparisons. Differentially expressed lncRNAs and mRNAs were identified through fold change and t-test screening.

The profiling identified a subset of the total number of probes analyzed by Agilent whole genome oligomicroarray that are differentially expressed. Unsupervised hierarchical clustering suggests that samples of different biological origin have a distinguishable gene expression profiling relative to the other samples.

Real-time quantitative PCR validation

Real-time quantitative (RT)-PCR was carried out as reported previously (Albertson et al. 2006). RNA from the subjects used in the initial microarray study was used for verification of the microarray data. Reverse transcription was performed using Superscript III Reverse Transcriptase (Invitrogen). 1 μg RNA, 500 ng Oligo (dT), and 10 mM each dNTP were incubated for 5 min at 65°C and then chilled on ice for 2 min. 5 × First Strand Buffer (250 mM Tris-HCl (pH 8.3), 375 mM KCl, and 15 mM MgCl2), 5 mM DTT (final concentration), 40 U RNase Out, and 200 U Superscript III RT were then added. The 20 μL reaction was incubated for 60 min at 50°C followed by a final incubation at 70°C for 15 min for termination. Quantitative PCR was carried out on a real-time detection instrument ABI PRISM7900 system (Applied Biosystems, Foster City, CA, USA) in 384-well optical plates using 2 × PCR master mix (SuperArray Bioscience, Frederick, MD, USA) at the following conditions: 10 min at 95°C, 40 cycles:10 s at 95°C and 60 s at 60°C. They were used to quantitate relative amounts of product using β-actin as an endogenous control. Expression ratios were subjected to a log 2 transform to produce fold change data. Student's t-test was used to test for significant differences between control and cocaine groups. anova was used to compare gene expression with a Tukey's post hoc comparison (spss version 17, spss lnc., Chicago, IL, USA). The primers used are listed in Table S1.

Ontological and pathway analysis

Ontological analysis used Gene Ontology (GO) categories to determine processes or functional categories that were differentially expressed, as described previously (Beißbarth and Speed 2004) using GeneSpring GX software. This analysis determined the number of genes in a category present on the array and the number of expression changes that would be part of that category by random chance given the number of differentially expressed genes.

Cis and trans analysis

Identify the targets of differentially expressed lncRNAs via cis- or trans-regulatory effects. Differentially expressed lncRNAs were selected for target prediction.

For cis-acting analysis, the genomic position of the lncRNAs relative to the genomic positions of its nearest known genes, which were determined from the RefSeq and UCSC Known Genes databases (Jia et al. 2010). Briefly, the lncRNAs were divided into three categories: (i) lncRNAs overlapping with known genes on the same strand; (ii) lncRNAs not overlapping with known genes on the same strand, but within 10 kb of known genes; and (iii) lncRNAs more than 10 kb away from known genes on the same strand. Distance stratification was performed because recent data suggest that lncRNAs near known genes are more likely to be functional (Mitchell Guttman et al. 2009).

For trans-acting analysis, we determined the target genes prediction of lncRNAs by program RNAplex, which reduces the time needed to localize putative hybridization sites, mainly by neglecting intramolecular interactions and by using as lightly simplified energy model (Tafer and Hofacker 2008). We compared the lncRNAs with the known genes by BLAST (e < 1E-5), and then by using program RNAplex to choose trans-acting target genes, RNAplex parameters were set as –e-20.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Cocaine-induced conditioned place preference

To examine the expression profiles of transcriptome in the NAc, mice were conditioned to cocaine in a 6-day unbiased CPP protocol (Fig. 1a, group II) and compared with mice receiving saline injections (group I). The rate of time spent in drug-paired compartment during the place preference test was presented (Fig. 1b). Each group spent about 50% time on the drug-paired side during the pre-conditioning phase, confirming the unbiased nature of our place conditioning procedure. The calculated CPP score for the saline group was 55 ± 8 s and the cocaine-paired groups was 360 ± 32 s (Fig. 1b). The group received 20 mg/kg of cocaine spent significantly more time in the cocaine-paired chamber on the post-test day than the pre-test day (p < 0.05).

Cocaine-induced CPP is associated with changes in mRNAs and lncRNAs expression in NAc

The mRNAs and lncRNAs expression profiles were assessed 24 h after the last conditioning session of CPP test. We measured the profiles of mRNAs and lncRNAs expressions with microarrays (Arraystar, Rockville, MD, USA). On the basis of the initial mRNAs and lncRNAs set, we utilized Principal Components Analysis to perform an unsupervised examination discriminated cocaine-conditioned mice from controls (Fig. 2a). The first two components accounted for 62.4% of the total variance with components 1 and 2 accounting for 47.7% and 14.7% of the variance, respectively. As shown in Fig. 2a, the cocaine-conditioned animals were distinctly separated from the saline controls in Principal Components Analysis plot, suggesting that mRNAs and lncRNAs expressions in NAc were less affected by cocaine CPP treatment.

image

Figure 2. Gene expression in NAc for cocaine-induce Conditioned place preference and control subjects. (a) Principal Components Analysis (PCA) reveals separation between groups for the 764 differentially expressed genes and 603 lncRNAs. The red represents the cocaine-induce Conditioned place preference (n = 4) and the black illustrates the control subjects (n = 4). (b) Volcano plot providing alteration of fold change of two-fold up and down, respectively [Log2 (fold change), x-axis] and the p values of 0.05 [−Log10 (p value), y-axis] for genes (left) and lncRNAs (right). The red point in the plot represents the differentially expressed genes or lncRNAs with statistical significance.

Download figure to PowerPoint

We found that 64.2% of RNAs were confidently detected above background levels in NAc (see Methods; Table 1), and that differentially expressed RNAs were identified through a combination of statistical significance (p < 0.05 and fold change > 2) in NAc of cocaine-conditioned mice. A volcano plot illustrates the variance in mRNAs and lncRNAs numbers at different p-values and fold change (Fig. 2b).

Table 1. Summary of microarray expression profile
Probe classTotalExpressed above backgroundDifferentially expresseda
  1. a

    Significant differential expression was defined as probes with p < 0.05 and fold change > 2.

mRNAs1781412939 (72.6%)764 (7.3%)
LncRNAs96794708 (48.6%)603 (13%)
Combined2748317647 (64.2%)1553 (8.8%)

Additional confirmation of these selected mRNAs and lncRNAs was obtained with hierarchical clustering and shown as a Clustered Image Map. We compared the mRNAs and lncRNAs expression profiles of cocaine-paired animals to those of saline controls as heap maps (Fig. 3a), and the Clustered Image Map clearly illustrates the separation of these two groups with a defined cluster order for the individual subjects (Fig. 3b). This illuminated 764 mRNAs expression changes in the NAc (representing 6.7% of the total mRNA species detected), including 325 up-regulated mRNAs and 439 down-regulated mRNAs (Table S2). From the lncRNAs subset, 603 lncRNAs were expressed (2.2% of the total detected), 13% (603 of 4708) of which were significantly differentially expressed (p < 0.05 and fold change > 2), including 129 up-regulated lncRNAs and 474 down-regulated lncRNAs between cocaine group and control (Table S3). Generally, a smaller up-regulation and a large down-regulation pattern of lncRNAs expressions of NAc were revealed in cocaine-treated mice. To determine if lncRNAs expression is associated with cocaine CPP-related behavior, Pearson's correlation coefficients between the expression of the lncRNAs and CPP scores were calculated. The expression of majority of lncRNAs was highly correlated (587/603, r > 0.5 or r < −0.5) (Table S13).

image

Figure 3. Heat map and hierarchical cluster analysis of NAc samples from cocaine-conditioned mice. (a) Heat map representing expression values of a panel of cocaine group relative to control. Each column represents the indicated one sample. Each row indicates mRNAs or long non-coding RNAs (lncRNAs). The color change reflects relative change according to the scale shown; red indicates positive fold change and green indicates negative fold change. (b) Clustered Image Map of the relative change in mRNAs and lncRNAs expression between the control and cocaine exposure groups. n = 4, each group.

Download figure to PowerPoint

Quantitative RT-PCR

To validate the microarray analysis findings, we randomly selected 11 mRNAs and 13 lncRNAs among the differential transcripts and analyzed their expression by quantitative RT-PCR in the NAc of individual mice. The results of quantitative RT-PCR were similar to those obtained from microarray analyses (p < 0.05), showing the good reproducibility of expression changes of lncRNAs and mRNAs using an independent method (Table 2 and Table 3).

Table 2. Quantitative RT-PCR confirmation for selected mRNAs changed after cocaine CPP treatment in NAc
GeneGenbankMicroarrayRT-PCR
Fold changep-valueFold changep-value
  1. Values indicate fold difference compared with control as detected on microarray and by quantitative RT-PCR.

BdnfNM_0075403.32<0.014.64<0.01
Grm8NM_0081742.01<0.012.26<0.05
JunbNM_008416−2.01<0.001−1.83<0.05
NovNM_0109303.33<0.052.74<0.05
Tbr1NM_0093222.92<0.052.39<0.05
Ppp1r1bNM_144828−2.40<0.001−1.21>0.05
Fkbp1aNM_008019−2.58<0.01−3.22<0.05
Nsd1NM_0087392.92<0.052.34<0.05
Mbd3NM_0135952.26<0.011.47>0.05
Pcgf2NM_009545−2.35<0.001−2.61<0.01
Chd1NM_007690−2.17<0.001−2.64<0.01
Table 3. Quantitative RT-PCR confirmation for selected lncRNAs changed after cocaine CPP treatment in NAc
GenbankChromosomencRNAtypeMicroarrayRT-PCR
Fold changep-valueFold changep-value
  1. Values indicate fold difference compared with control as detected on microarray and by quantitative RT-PCR.

uc008nebchr2sense_overlap−2.04<0.001−3.64<0.05
AK037356chr8intergenic−2.07<0.001−1.64<0.01
AK047779chr16intergenic−2.97<0.001−2.46<0.05
AK016381chr3intergenic3.81<0.052.38>0.05
AK044606chr14antisense_overlap−2.10<0.001−3.74<0.05
AK031173chr8antisense_overlap3.45<0.0014.38<0.01
uc007fjqchr10bidirection−2.22<0.001−2.47<0.05
AK017009chr9bidirection−2.25<0.001−2.83>0.05
AK162927chr12bidirection2.41<0.0012.85<0.01
AK050225chr9others−2.02<0.001−1.47<0.05
AK078200chr9sense_overlap−2.22<0.001−3.84<0.05
AK036076chr10sense_overlap−2.38<0.001−1.73<0.01
AK018530chr19sense_overlap−2.16<0.001−2.23<0.05

Ontological and pathway analysis of mRNAs expression in NAc

To investigate whether the clustering of genes correlates with functional groupings, we performed GO functional enrichment analysis on all of the genes as a whole, and on each cluster of temporally co-expressed genes separately. The GO annotation system uses a controlled and hierarchical vocabulary to assign function to genes or gene products in any organism (Ashburner et al. 2000). Analysis of GO categories of changes in NAc identified a number of mRNAs significantly regulated by cocaine. Interestingly, these mRNAs are significantly enriched in several pathways, including calcium signaling pathway, MAPK signaling pathway, regulation of transcription, DNA-dependent, calcium ion binding, and transcription factor activity, which are processes generally linked to neuroplasticity, learning, and memory. Therefore, we subjected the differentially expressed mRNAs to GO analysis to detect molecular and cellular functional domains impacted by cocaine (Table 4).

Table 4. GO terms identified in NAc as overrepresented in mice exposed to cocaine compared with the control
 Number of genesp-Valueaq-Valuea
  1. a

    p-values and q values were calculated using Gene Go software with a false discovery rate cutoff of 0.01. Overall, 187 GO categories showed p-values < 0.01 and q values < 0.01.

  2. All 764 genes called present on at least one array were included in the analysis. The number of genes for each GO term is provided, and GO terms grouped into families are part of the same GO tree.

Pathway
Neuroactive ligand-receptor interaction189.91E-171.35E-14
Calcium signaling pathway133.55E-122.41E-10
Regulation of actin cytoskeleton136.98E-123.16E-10
Cytokine–cytokine receptor interaction135.54E-111.88E-09
Focal adhesion117.15E-101.94E-08
Axon guidance94.51E-099.44E-08
Cell Communication94.51E-099.44E-08
MAPK signaling pathway93.40E-085.78E-07
Biological process
GO:0006355 regulation of transcription, DNA-dependent1322.05E-226.08E-20
GO:0045944 positive regulation of transcription from RNA polymerase II promoter445.19E-168.81E-14
GO:0007156 homophilic cell adhesion248.37E-139.95E-11
GO:0031000 response to caffeine71.54E-101.41E-08
Cellular component
GO:0005634 nucleus2581.77E-402.10E-37
GO:0016021 integral to membrane2161.62E-183.22E-16
GO:0005886 plasma membrane1256.10E-126.59E-10
Molecular function
GO:0005515 protein binding3309.91E-365.89E-33
GO:0008270 zinc ion binding1403.88E-261.54E-23
GO:0005509 calcium ion binding691.53E-183.22E-16
GO:0003700 transcription factor activity696.42E-169.53E-14
GO:0043565 sequence-specific DNA binding507.58E-141.00E-11
GO:0046982 protein heterodimerization activity218.49E-118.41E-09

Expressed lncRNAs share loci with protein-coding genes

The functional characterization of hundreds of lncRNAs shown here presents a formidable task, and lncRNAs have previously been shown to originate from complex loci that contains interlaced networks of long non-coding and protein-coding transcripts (Kapranov et al. 2005; Engström et al. 2006). For examples, a number of previously characterized lncRNAs regulate the expression of adjacent or overlapping protein-coding genes (Feng et al. 2006; Martianov et al. 2007), analysis of the genomic context of those lncRNAs modified significantly by cocaine helps to predict their functional role, at least at the biological if not the mechanistic level. Therefore, we identified all potential lncRNAs as sense-overlap, antisense-overlap, intergenic, bidirection, ultra-conserved region (UCR), and others (Tables S4–8). These included 361 sense sequences (56 up-regulated and 305 down-regulated), 52 antisense-overlap sequences (18 up-regulated and 34 down-regulated), 64 intergenic lncRNAs (12 up-regulated and 52 down-regulated), 41 bidirectional lncRNAs (15 up-regulated and 26 down-regulated), 16 UCR encoding lncRNAs, and 69 others.

Sense-overlap and antisense-overlap sequences

Transcriptional profiling has shown that sense and antisense transcriptions are prevalent in the mammalian genome (Carninci et al. 2005); moreover, several studies indicate their importance in regulating diverse neurological processes (Yamasaki et al. 2005; Faghihi et al. 2008). We identified 361 sense lncRNAs and 52 lncRNAs that are antisense to the exons of protein-coding genes (Table S4 and Table S5). These sense and antisense lncRNAs often share varied and complex expression relationships with their protein-coding transcripts.

Intergenic lncRNAs

It has been recently reported the identification of more than 1000 intergenic lncRNAs in the mouse genome (Carninci 2008; Khalil et al. 2009). In an attempt to understand the potential biological roles of intergenic lncRNAs, a method to infer putative function based on correlation in expression between intergenic lncRNAs and protein-coding genes was developed. These studies led to preliminary hypotheses about the involvement of intergenic lncRNAs in diverse biological processes, from stem cell pluripotency to cell-cycle regulation (Loewer et al. 2010). We observed 64 intergenic lncRNAs that were changed (Table S6).

Bidirectional lncRNAs

A major organizational theme with in both mouse and human genomes that encompass ~10% of known genes is the prevalence of bidirectional transcript pairs (Trinklein et al. 2004; Engström et al. 2006), where expressions of two transcripts are initiated in close proximity, but in opposite directions. We identified 41 lncRNAs that form bidirectional pairs with protein-coding genes (Table S7).

Cis and trans analysis

A major function of lncRNAs prevalent appears to be to regulate expression of neighboring mRNAs through cis- and trans-acting mechanisms throughout the mammalian genome (Petruk et al. 2006; Rinn et al. 2007). Within our data sets, we found that 44.3% (267 of 603) of expressed lncRNAs which have been reported to act in cis to their targeted loci, and 23.7% (143 of 603) of expressed lncRNAs which have been reported to act in trans to their targeted loci (Table S10 and Table S11). Next, we compared the mRNAs, which are acted in cis or trans to their lncRNAs, with those significantly changed in the NAc of cocaine-conditioned mice (Table S12). We found that cocaine administration led to the induction of 48 mRNAs, and the subset of those also induced to act in cis (36 of 267) and in trans (12 of 143). Furthermore, we analyzed the correlation of the expression between the mRNAs and associated lncRNAs, and found that the expression profiles of most of those lncRNAs correlate significantly and positively with associated mRNAs. These results suggest that mRNAs expression by cocaine may be associated with the actions of lncRNAs. However, few pairs like uc008ydj-Nup54, uc007gbv-Dos, and AK048353-Mfsd4 are reverse tendency.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Recent transcriptional profiling has provided an unprecedented level of detail regarding drug addiction (Albertson et al. 2006; Freeman et al. 2010), and several reports regarding aberrant expression of microRNAs have been reported (Hollander et al. 2010; Eipper-Mains et al. 2011). However, the expression of lncRNAs and their potential involvement in cocaine addiction remain unknown so far.

A computational analysis of in situ hybridization data from the Allen Brain Atlas identified 849 lncRNAs showing specific expression patterns in adult mouse brain (Mercer et al. 2008b). Moreover, a great diversity of lncRNAs is expressed and regulated in the brain, where they are thought to have fundamental functions (Mercer et al. 2008b). We were therefore interested in studying lncRNAs in cocaine-conditioned mice in the hope of gaining new insights into cocaine addiction.

In this study, we have taken advantage of high throughput microarrays and several data mining tools to identify and characterize the transcripts of both protein-coding and non-coding transcripts in NAc from cocaine-induced CPP subjects (Table S1 and Table 2). Using conservative thresholds for statistical and biological significance, we identified and characterized 764 mRNAs and 603 lncRNAs, providing 48 pair mRNA-lncRNAs, which were meanwhile significance changed.

In comparison with control, cocaine CPP led to a stronger shift in mRNA transcription of NAc, as manifested by the larger number of differentially expressed genes and by changes in molecular functions defined by GO and pathway including MAPK signaling, axon guidance, actin cytoskeleton regulation, and calcium signaling (Table 4). In fact, numerous studies have shown that acute and chronic exposure to abused drugs, such as cocaine and morphine, can induce MAPK activation (Zhang et al. 2004; Mattson et al. 2005), and modulate axon guidance molecules (Bahi and Dreyer 2005) in mediating the rewarding effects. It has been known that these signalings contribute to the development of drug-induced persistent changes in the brain.

The majority of lncRNAs we identified have not previously been characterized beyond their original cloning and sequencing, and this study places them for the first time in a defined biological context in NAc of cocaine-addicted mice. It has been known that epigenetic alterations in the accessibility of genes within their native chromatin structure can be induced by histone tail modifications and DNA methylation, and the regulation of gene expression by ncRNAs (Robison and Nestler 2011; Wong et al. 2011). Some recent study demonstrated that the lncRNAs, RNCR2, plays a critical role in regulating mammalian retinal cell fate specification (Rapicavoli et al. 2010); Evf2 recruits DLX and MECP2 transcription factors and controls the expressions of Dlx5, Dlx6, and Gad1 through trans- and cis-acting mechanisms in the embryonic brain (Bond et al. 2009). Given these documented roles of lncRNAs in epigenetic mechanisms and neurological disorders (Qureshi et al. 2010), our findings in this study that hundreds of lncRNAs in NAc are modified by cocaine should not be surprising.

LncRNAs have been categorized in terms of localization relative to protein-coding genes, splicing, and chromatin modification (Yan et al. 2005; Rinn et al. 2007; Mercer et al. 2008b), although it remains a matter of debate whether the majority are biologically meaningful or merely ‘transcriptional noise’ (Huttenhofer et al. 2005; Koerner et al. 2009). We then examined the genomic context of those altered lncRNAs from complex genomic loci that encompasses protein-coding genes with defined roles. Interestingly, our results are consistent with the likelihood that these pairs of lncRNAs and associated protein-coding genes often exhibit in a variety of ways, including as sense-overlap, antisense-overlap, intergenic, bidirection, UCR, and others. Viewing the lncRNAs in their genomic context, our findings suggest the potential function implications of their expression profiles, particularly with respect to lncRNAs associated with cocaine abuse. We speculate that these lncRNAs may in fluence the expression of associated protein-coding genes, which is similar to previously characterized examples such as BACE1-AS (Faghihi et al. 2008), Sox2ot (Amaral et al. 2009), and HOTAIR (Gupta et al. 2010).

A major function of lncRNAs appears to control the expression of the proximal and distal mRNAs expressed through cis- and/or trans-acting pathways (Berretta and Morillon 2009). For example, the transcription of lncRNAs regulates expression of homeotic genes of Drosophila in cis action (Petruk et al. 2006), and the lncRNAs, Jpx, as an RNA-based activator of Xist, is a molecular switch for X chromosome inactivation in cis-acting (Tian et al. 2010). However, the cis-acting lncRNA, Xsit, is shown to be capable of functioning in trans, suggesting that the cis/trans distinction may reflect differences in targeting of lncRNA rather than differences in mechanism (Jeon and Lee 2011). Those lncRNAs modified by cocaine in this study have been reported to act in cis or trans to targeted loci (Table S10 and Table S11).

In summary, we report for the first time the global transcriptional profiling (including mRNAs and lncRNAs) in NAc and how they change significantly by cocaine CPP. We have identified and characterized 764 mRNAs and 603 lncRNAs using conservative thresholds for statistical and biological significance, provided 48 pair candidates for further functional experimentation, such as RNA binding assays and expression knockdown experiments. These data provide a large spectrum of lncRNAs expression associated with cocaine-induced neuroplasticity and an important resource for future research. Nevertheless, it is important to note that the findings described in this article are merely a starting point for the study of lncRNAs in the field of drug addiction, and much exciting work lies ahead to functionally characterize the identified lncRNAs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

This work was supported by the Project of the National Natural Sciences Foundation of China (30870888, 30970938) and National Innovative Drug Development (2012ZX09302-004).

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Conflict of interest
  8. References
  9. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
jnc12006-sup-TableS1-S13.pdfapplication/PDF390K

Table S1. List of RT-PCR primers used to validate the microarray data.

Table S2. Full list of mRNAs was significant differential expression in NAc by in cocaine-conditioned mice.

Table S3. Full list of lncRNAs was significant differential expression in NAc by in cocaine-conditioned mice.

Table S4. List of sense-overlap lncRNAs was significant differential expression in NAc in cocaine-conditioned mice.

Table S5. List of antisense-overlap lncRNAs was significant differential expression in NAc in cocaine-conditioned mice.

Table S6. List of intergenic lncRNAs was significant differential expression in NAc in cocaine-conditioned mice.

Table S7. List of bidirectional lncRNAs was significant differential expression in NAc in cocaine-conditioned mice.

Table S8. List of lncRNAs form UCR database was significant differential expression in NAc in cocaine-conditioned mice.

Table S9. List of other lncRNAs was significant differential expression in NAc in cocaine-conditioned mice.

Table S10. List of the lncRNAs was predicted to act in cis to their targeted loci.

Table S11. List of the lncRNAs was predicted to act in trans to their targeted loci.

Table S12. List of the lncRNAs, which acted in cis or trans to their mRNAs, to those significantly changed in the NAc of cocaine-conditioned mice.

Table S13. LncRNAs with highly correlated expression profiles to cocaine CPP scores. Correlation co-efficients were determined by Pearson's correlation.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.