Cocaine induces paradigm-specific changes to the transcriptome within the ventral tegmental area

Abstract

During the initial stages of drug use, cocaine-induced neuroadaptations within the ventral tegmental area (VTA) are critical for drug-associated cue learning and drug reinforcement processes. These neuroadaptations occur, in part, from alterations to the transcriptome. Although cocaine-induced transcriptional mechanisms within the VTA have been examined, various regimens and paradigms have been employed to examine candidate target genes. In order to identify key genes and biological processes regulating cocaine-induced processes, we employed genome-wide RNA-sequencing to analyze transcriptional profiles within the VTA from male mice that underwent one of four commonly used paradigms: acute home cage injections of cocaine, chronic home cage injections of cocaine, cocaine-conditioning, or intravenous-self administration of cocaine. We found that cocaine alters distinct sets of VTA genes within each exposure paradigm. Using behavioral measures from cocaine self-administering mice, we also found several genes whose expression patterns corelate with cocaine intake. In addition to overall gene expression levels, we identified several predicted upstream regulators of cocaine-induced transcription shared across all paradigms. Although distinct gene sets were altered across cocaine exposure paradigms, we found, from Gene Ontology (GO) term analysis, that biological processes important for energy regulation and synaptic plasticity were affected across all cocaine paradigms. Coexpression analysis also identified gene networks that are altered by cocaine. These data indicate that cocaine alters networks enriched with glial cell markers of the VTA that are involved in gene regulation and synaptic processes. Our analyses demonstrate that transcriptional changes within the VTA depend on the route, dose and context of cocaine exposure, and highlight several biological processes affected by cocaine. Overall, these findings provide a unique resource of gene expression data for future studies examining novel cocaine gene targets that regulate drug-associated behaviors.

Fig. 1. Conventional cocaine behavioral paradigms used to identify paradigm-specific changes to the VTA Transcriptome.

a Schematic timeline of home cage injections. Mice either underwent acute I.P. injections with cocaine (20 mg/kg) or saline (HC Acute) or were (b) chronic I.P. injections with cocaine or saline (HC Chronic). Tissue was collected 1 h following the last injection. c Schematic timeline of animals undergoing modified contextual conditioning. Mice underwent one conditioning session of either cocaine (20 mg/kg) or saline (Conditioned). Tissue was collected 1 h following the I.P. injection of either cocaine of saline. d Schematic of IVSA paradigm. Mice were mildly food restricted (85–90% free-feeding body weight) and following food training at an FR1 schedule (Supplementary Fig. 1), mice underwent either 7 days of IVSA cocaine or saline. Tissue was collect 1 h following completion of the last IVSA session. e, f Cocaine-associated Memories are formed within One Cocaine Contextual Conditioning Session and Acquisition of Cocaine-self-administration occurs within 1 Week of IVSA In Adult Male Mice. e Mice underwent either one of the following using contextual conditioning boxes: one conditioning session of cocaine and saline; two conditioning sessions of cocaine and saline; or four conditioning sessions of only saline. f Conditioning with either one pairing or two pairings of cocaine and saline induced a higher contextual place preference score than saline alone (One-way ANOVA with Tukey’s multiple comparisons *p < 0.05, **p < 0.0001). g–k Cocaine is reinforcing under FR1 conditions. g Schematic of IVSA paradigm. While mildly food restricted, mice underwent food training at a FR1 schedule and subsequently underwent either 7 days of IVSA cocaine or saline. h Cumulative record of cocaine or saline infusions of mice. (Two-way ANOVA repeated measure Cocaine effect (F1,21) = 2.784, p = 0.112; Session effect F(6,126) = 134.5, p < 0.0001; Interaction effect F(6,126) = 3.634, p = 0.0023). Mice self-administered for either cocaine or saline across 7 days. i. Cocaine self-administering mice discriminate more between active levers than saline (Two-way ANOVA repeated measure: Session effect F(6,126) = 3.235, p = 0.005; Cocaine effect F(1,21) = 25.67, p < 0.001; Interaction effect F(6,126) = 2.556, p = 0.0227). j Mice with access to IV cocaine underwent reinforcement learning, shown by the discrimination between active lever pressing vs. the inactive lever (Two-way ANOVA repeated measure Session effect (F1,14) = 52.98, p = 0.3224; Lever effect F(6, 1,14) = 52.98, p < 0.0001; Interaction effect F(6,84) = 0.9885, p = 0.4384). k Mice with access to IV saline failed to discriminate between levers (Two-way ANOVA repeated measure Lever effect (F1,14) = 25.88, p = 0.0002; Session effect F(6,84) = 1.144, p = 0.3442; Interaction effect F(6,84) = 0.6, p = 0.2165).

Fig. 2. Cocaine alters gene expression within the VTA in a paradigm-specific manner.

a–d Volcano plots illustrating significance (Y-axis) and magnitude (X-axis) of cocaine-induced changes from each paradigm. e Expression changes for a semi-random subset of genes was analyzed by qPCR for each paradigm from RNA-Seq samples. Cocaine-induced expression change is represented by Log2(fold-change) vs. saline group from each paradigm and compare to the fold change found by RNA-Seq. * = p < 0.05; ^# = p < 0.1. f Heat map comparing number and magnitude of cocaine-induced gene expression changes in each group.. g Charts showing the percentage of upregulated and downregulated genes by cocaine from each paradigm and listed total number of differentially expressed genes (genes that were up- and downregulated from each cocaine-treated group relative are relative to the saline counterparts within each paradigm (FC < 0.5, uncorrected p value < 0.05). h Cocaine induces paradigm-specific changes in gene expression within the VTA. All Venn Diagrams show upregulated or downregulated genes (FC > 0.5, p < 0.05) by cocaine relative to the saline controls within that specific paradigm. Any genes upregulated or downregulated by both paradigms being compared in Venn Diagram are listed. Number of common and/or distinct upregulated DEGS and downregulated DEGs (FC, 0.5 p < 0.05) relative to saline-controls from each paradigm shown with Venn Diagrams.

Fig. 3. Predicted biological processes altered by each cocaine exposure paradigm.

a–d Top gene ontology (GO) terms of both upregulated and downregulated DEGs within each paradigm are listed. e Common GO terms affected by cocaine across paradigms.

Fig. 4. Top predicted upstream regulators of both upregulated genes and downregulated genes induced by cocaine in each paradigm.

a–h Top predicted upstream regulators of differentially expressed genes by each cocaine exposure paradigm. Heat maps show the log-fold changes of the genes targeted by the top upstream regulator from each cocaine exposure paradigm. Below each comparison is a DNA motif of the binding sequences of a top upstream regulator in a specific paradigm.

Fig. 5. Gene expression levels correlating with one behavioral measure of cocaine IVSA.

a Schematic of IVSA. b Number of genes upregulated in the VTA by cocaine self-administration compared to saline self-administration. c Correlation heat map of the fold changes from the top 50 genes with the average daily cocaine intake (Tnfrsf12a, Hip1, Idnk) (left) and total cocaine consumption (right). Exact r values for each gene and exact p values are available in Supplementary Data 2. d Fold changes of Idnk correlate with average daily cocaine intake (r = 0.5437, p = 0.03). e Number of genes downregulated in the VTA by cocaine self-administration compared to saline self-administration. f Correlation heat map of the fold changes from the top 50 genes with the average daily cocaine intake (Zbtb16, Mef2a) (left) and total cocaine consumption (Rn45s, Zbtb16) (right). g Zbtb16 fold changes correlate with average daily cocaine intake (r = 0.7137, p = 0.002) and (h) total cocaine consumption (r = 0.5366, p = 0.03). (* = p < 0.05, ** = p < 0.01).

Fig. 6. Coexpression analysis identifies gene networks altered by cocaine implicated in gene regulation and synaptic processes within specific cell types of the VTA.

a WGCNA identified gene coexpression relationships, formed into modules, within each cocaine exposure paradigm. The modules with significant enrichment (* = p < 0.05) after Bonferroni correction are plotted. b Cell type enrichment analysis on the top modules shows that gene networks are altered within various cell types, primarily glial cells. c GO term analysis conducted on the top module in each paradigm shows that gene networks are altered within many biological processes, including those related to synaptic processes and gene regulation. d 30 most connected “hub” genes are identified from the top module in each paradigm.