Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Cerebellum Transcriptome of Mice Bred for High Voluntary Activity Offers Insights into Locomotor Control and Reward-Dependent Behaviors

  • Kelsey Caetano-Anollés,

    Affiliation Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America

  • Justin S. Rhodes,

    Affiliations Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America, Center for Nutrition, Learning and Memory, University of Illinois at Urbana-Champaign, Urbana, IL, 61801, United States of America

  • Theodore Garland Jr,

    Affiliation Department of Biology, University of California Riverside, Riverside, CA, United States of America

  • Sam D. Perez,

    Affiliation Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America

  • Alvaro G. Hernandez,

    Affiliation High-Throughput Sequencing and Genotyping Unit, Roy J. Carver Biotechnology Center, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America

  • Bruce R. Southey,

    Affiliation Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America

  • Sandra L. Rodriguez-Zas

    rodrgzzs@illinois.edu

    Affiliations Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America, Department of Statistics, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America, Carle Woese Institute for Genomic Biology, University of Illinois at Urbana-Champaign, Urbana, IL, United States of America

Cerebellum Transcriptome of Mice Bred for High Voluntary Activity Offers Insights into Locomotor Control and Reward-Dependent Behaviors

  • Kelsey Caetano-Anollés, 
  • Justin S. Rhodes, 
  • Theodore Garland Jr, 
  • Sam D. Perez, 
  • Alvaro G. Hernandez, 
  • Bruce R. Southey, 
  • Sandra L. Rodriguez-Zas
PLOS
x

Abstract

The role of the cerebellum in motivation and addictive behaviors is less understood than that in control and coordination of movements. High running can be a self-rewarding behavior exhibiting addictive properties. Changes in the cerebellum transcriptional networks of mice from a line selectively bred for High voluntary running (H) were profiled relative to an unselected Control (C) line. The environmental modulation of these changes was assessed both in activity environments corresponding to 7 days of Free (F) access to running wheel and to Blocked (B) access on day 7. Overall, 457 genes exhibited a significant (FDR-adjusted P-value < 0.05) genotype-by-environment interaction effect, indicating that activity genotype differences in gene expression depend on environmental access to running. Among these genes, network analysis highlighted 6 genes (Nrgn, Drd2, Rxrg, Gda, Adora2a, and Rab40b) connected by their products that displayed opposite expression patterns in the activity genotype contrast within the B and F environments. The comparison of network expression topologies suggests that selection for high voluntary running is linked to a predominant dysregulation of hub genes in the F environment that enables running whereas a dysregulation of ancillary genes is favored in the B environment that blocks running. Genes associated with locomotor regulation, signaling pathways, reward-processing, goal-focused, and reward-dependent behaviors exhibited significant genotype-by-environment interaction (e.g. Pak6, Adora2a, Drd2, and Arhgap8). Neuropeptide genes including Adcyap1, Cck, Sst, Vgf, Npy, Nts, Penk, and Tac2 and related receptor genes also exhibited significant genotype-by-environment interaction. The majority of the 183 differentially expressed genes between activity genotypes (e.g. Drd1) were under-expressed in C relative to H genotypes and were also under-expressed in B relative to F environments. Our findings indicate that the high voluntary running mouse line studied is a helpful model for understanding the molecular mechanisms in the cerebellum that influence locomotor control and reward-dependent behaviors.

Introduction

The role of cerebellum in the control of movement has been extensively studied. However, the roles of the cerebellum in motivation, executive control, working memory, learning, and addictive behaviors are starting to be understood [1, 2].For example, the cerebellum has been associated with cocaine-related behaviors [3] as well as motor skills, object manipulation, knowledge, and their automatization [4]. Also, the cerebellum is activated by drug-associated cues [57] and during cognitive tasks such as language and memory in humans [8], and has been linked to reward-based learning [9, 10]. The involvement of the cerebellum in motivation or the internal drive of an organism may be established through its interactions with the endocrine system [11]. Indicators of exploratory behavior and spatial orientation in cerebellectomized rodents indicate that the cerebellum is involved not only in cognitive but also in motivational processes, spatial memory, and in cognitive processes of the motor program elaboration [12, 13].

Mouse lines selectively bred for high physical activity, such as the High Runner lines, are offering insights into the neurobiology of increased voluntary wheel running behavior [1416]. Mouse lines selected for high voluntary wheel running exhibit significant behavioral and physiological differences relative to control lines as early as 10 generations after selective breeding. Moreover, studies of these lines are characterizing the role of brain regions in locomotor control [14, 15, 1719]. Mice from the High Runner lines show significantly lower monoamine concentrations than mice from the control lines in the substantia nigra pars compacta and dorsolateral striatum regions of the brain, both of which are involved in locomotor control [20]. Also, blocked access to a wheel elicits neurobiological profiles similar to narcotic withdrawal in High Runner lines [16, 21].

Studies of high and low voluntary wheel-running rat and mouse lines have led to the proposition that this physical activity model can support the understanding of genes related to the motivation to run and to develop and maintain addictive behaviors in addition to locomotor activity [16, 22]. Physical activity and drugs of abuse have rewarding effects supported by similar brain pathways. High Runner lines also exhibit dysregulation in dopamine signaling [23] and endocannabinoid system involved in brain reward processes [16, 24]. High running can be a self-rewarding behavior exhibiting addictive properties [15, 17] and significant departures from average home cage activity levels have been associated with other behavioral disorders [25]. High Runner mouse lines also exhibit high home cage activity in the absence of wheels and high withdrawal behavior of despairity in a forced-swim test after removal of wheels following 6 days of access [26]. Likewise, when High Runner and control mice are rewarded with running time after pressing a lever [27], most High Runner and control mice press the lever when the reward was 30 minutes of running time but few High Runner mice press the lever when the reward was reduced to 1.5 minutes. The weaker response of the High Runner mice for limited reward suggests higher motivation for extended physical activity rather than potentially shorter attention because the latter would have been expressed at both reward levels.

Research on the High Runner mouse lines has concentrated on the striatum because of the association of this brain region with motor behavior [2830], reward processing and addiction. However, changes in the cerebellum in response to selection are also expected because of the involvement of this brain region in the control of movement and motivational processes. Although the size of the cerebellum does not differ significantly between High Runner and control mice [15], selection may alter the molecular profiles and thus impact on behavior and locomotor control. Potential differences in the cerebellum transcriptome between mice from a high activity and unselected control lines and in consideration of environmental opportunities for activity have not been assessed.

A study of cerebelloctomized mice concluded that the cerebellum is involved in motor and spatial capabilities, and in cognitive and motivational processes that are key for exploration behavior [12]. Likewise, a study of the capability of cerebelloctomized rats to locate a reward concluded that the cerebellum is involved in the mechanisms sustaining focused spatial memory and in the cognitive processes of the motor program elaboration [13]. A similar association between the cerebellum and motor control, motivation, and reward-focus may have been developed in the High Runner mice line and a transcriptome analysis can offer insights into these roles.

The goals of this study were: 1) to compare the cerebellum transcriptome and associated networks of a High Runner mouse line relative to mice from a control line and, 2) to evaluate the modulation of potential transcriptome differences between genotypes and by the environmental opportunity for physical activity. We profiled the cerebellum transcriptome of mice using a 2-by-2 factorial design comparing 2 activity genotypes, a line selected for high wheel running versus an unselected control line, in either blocked or free access to wheel running environments. The comparison of both activity genotypes blocked and free access to wheel running environments permits us to disentangle the transcriptome profiles and networks related to the predisposition for high running as well as gene expression patterns correlated with actual running itself. Transcriptomic profiles in the brains of mice selectively bred for high voluntary wheel running in environments where the potential reward of wheel running is present or absent revealed molecular mechanisms underlying reward-dependent behaviors, motivation, and coordination of increased physical exercise. This study supports. efforts to understand the role of the cerebellum in cognition, motivation, and addictive behaviors.

Materials and Methods

Subjects

Mice from generation 66 of a replicated selective breeding experiment for high voluntary wheel-running behavior were used. As previously described [18], the original progenitors were outbred, genetically variable laboratory house mice (Mus domesticus) of the Hsd:ICR strain. After 2 generations of random mating, mice were randomly paired and assigned to 8 closed lines (10 pairs in each). In each subsequent generation, ~6 to 8 week-old offspring of these pairs were housed individually with access to a running wheel for 6 days (except during generations 32–35 when selection was relaxed as the colony was moved from Wisconsin to California). The highest-running male and female from each family within 4 High-Runner lines were selected as parents of the next generation. The selection criterion was total number of revolutions run on days 5 and 6 of the 6-day test. A male and a female were randomly chosen from within each family within 4 Control lines. Within all lines, the selected parents were randomly paired while avoiding sibling matings. In the present study, mice from 1 Control line (line 1) and 1 High-Runner line (line 7) chosen at random among all 8 lines available were evaluated in 2 activity environments. A total of 16, 7-week old adult males were used (n = 4 per line and environment combination). Power calculations based on gene expression measurements reported in prior studies [23, 31] involving these High-Runner lines indicated that a sample size of 4 provided 75% statistical power for a type I error rate corresponding to an FDR-adjusted P-value equal to 0.05 and log2(fold change) equal to 2.

Animal procedures were approved by the Illinois Institutional Animal Care and Use Committee and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. As described previously [18], mice were weaned at 21 days of age, and individually marked. At approximately 6 weeks of age, mice were placed individually with access to wheels for 7 days. Mice were kept on a 12-hour light/dark cycle with lights on at 06:00 hours and lights off at 18:00 hours [32]. On day 7, half of the mice were blocked from entering the running wheel by placing a barrier in the tunnel connecting the wheel to the cage [21] During the mid-portion of the dark cycle, when mice are normally most active on running wheels, mice were removed from their cages and immediately decapitated. The Control and High Runner lines used in this study are denoted as activity genotypes C and H, respectively. The conditions blocked from running and free access to the wheels following a period of wheel availability are denoted environments B and F, respectively.

RNA profiling

Brain was extracted and the whole cerebellum was dissected on a chilled aluminum block, snap frozen on dry ice and stored in a centrifuge tube at -80°C following published protocols [33]. Subsequently, tissue was homogenized with an RNase-free disposable pellet pestle (Fisher) and RNA was extracted using the commercially available RNeasy® Lipid Tissue Mini Kit (Qiagen, Valencia, CA). Purification of the isolated RNA included treatment with DNase I (Qiagen, Valencia, CA), accordingly to the manufacturer's instructions. For assessing total RNA yield, aliquot samples were measured with the Qubit® 2.0 (Life Technologies, Carlsbad, CA). Quality and integrity of isolated RNA samples were determined by 28S/18S rRNA analysis with the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara CA). RNA integrity was assessed using the Agilent 2100 Bioanalyzer with RNA Pico chip (Agilent Technologies, Palo Alto, CA). All samples had RNA Integrity Numbers (RIN) > 9. Libraries from individual mouse cerebellum samples were sequenced [34] and 100nt-long paired-end reads from each mouse separately were obtained using Illumina HiSeq 2000 (Illumina, San Diego, CA, USA).

Differential expression analysis

Using the FastqGroomer tool, FastQ data files were converted to FastqSanger format following proven protocols [34, 35]. A quality control check was performed on the reads using FastQC, and the threshold for end position removal due to low quality using the Fastq Quality Trimmer tool was Phred < 20. The reads were then mapped to the mouse mm10 genome assembly (UCSC; http://genome.ucsc.edu) using Tophat (v2.0.13) [36]. Tophat settings were expected (mean) inner distance = 200nt, anchor length junctions spanned by reads with at least 8bp on each side of the junction with 0 mismatch, independent mapping of read segments of 25nt long, and default intron length specifications. Cufflinks (v2.2.0) [36] was used to assemble the transcripts and estimate gene abundances. Differential transcript abundance between experimental conditions was tested using Cuffdiff using a geometric library normalization [36]. Differential expression was tested between genotype-environment combination groups, between activity genotype groups, and between environment groups. Evaluation of the interaction between activity genotypes (H or C) and environments (B or F) allowed the identification of differential profiles between high activity and control genotypes that were environment-dependent [35].

The nomenclature used in this study to identify the genotype-environment combination uses 2 letters, the first letter specifies genotype and the second letter specifies environment. Four genotype-environment combination groups are available in this study: High runner genotype-Free environment (HF); High runner genotype-Blocked environment (HB); Control genotype-Runner environment (CF); and Control genotype-Blocked environment (CB). The 4 genotype-environment groups allowed the evaluation of 6 distinct pairwise contrasts: CB-HB, CB-CF, CB-HF, HB-CF, HB-HF, and CF-HF. While information on all these 6 contrasts is provided, the interpretation focuses on 3 orthogonal contrasts: CF-HF, CB-HB, and HB-HF. The first 2 orthogonal contrasts allowed the identification of differences between H and C activity genotypes within each environment and the third contrast allowed us to identify differences between B and F environments within the H genotype. Commonalities between the first 2 orthogonal contrasts support the identification of genes associated with activity genotype regardless of environment. Commonalities between all these orthogonal contrasts identify genes that may respond to both activity genotype and environment. Discussion based on these 3 orthogonal contrasts is complemented with information from selected other contrasts to facilitate the interpretation of results supporting the interaction between genotype and environment. Testing for the main effects of genotype and environment followed the testing for interaction. False discovery rate (FDR) adjusted P-values were used for multiple test adjustment of differential expression across all genes tested [37].

Functional and network analysis

Enrichment of functional terms among the genes in 3 orthogonal interaction contrasts (CF-HF, CB-HB, and HB-HF), the activity genotype contrast, and the environment contrast were investigated. The comparison of the enriched functional terms would reveal shared and distinct molecular processes across contrasts that are not obvious from the analysis of individual genes. Two complementary functional enrichment analyses of the transcriptome profiles were undertaken. First, among genes differentially expressed (FDR-adjusted P-value < 0.05) enrichment of functional categories and pathways was performed using the Database for Analysis, Validation, and Integrated Discovery system (DAVID) [38]. Gene Ontology (GO) functional categories investigated included biological process (BP), molecular function (MF), and pathways defined by the Kyoto Encyclopedia of Genes and Genomes (KEGG). Enrichment of Functional Annotation Tool (FAT) categories was studied in recognition of the redundancy between GO terms. The FAT categories comprise informative BP and MF terms while filtering broad and less informative terms. EASE scores (modified Fisher Exact) were used to assess the statistical significance of the enrichment of the individual terms. Clustering of categories based on shared genes further minimized redundancy between categories. The statistical significance of these clusters was summarized using the enrichment score that corresponds to the geometric mean of the EASE scores of the categories in the cluster [3942]. DAVID analysis was complemented with Gene Set Enrichment Analysis (GSEA) [43]. This analysis was performed on the gene expression profiles of all genes profiled to uncover enriched categories within the up- and down-regulated genes within orthogonal contrasts [44]. Maximum and minimum cutoffs for gene set size within up- and down-regulated lists were 1000 and 5 genes respectively. The statistical significance of the enrichment was assessed using 1000 permutations and FDR adjustment for multiple testing.

Gene networks were reconstructed within the Cytoscape environment [45] using the Bisogenet package [46]. Networks encompass genes exhibiting significant (FDR-adjusted P-value < 0.05) differential expression in at least one of the 3 orthogonal contrasts (CF-HF, CB-HB, HB-HF) to facilitate visual comparison [47]. Networks connecting at least 15 differentially expressed genes directly or indirectly through intermediate genes not differentially expressed were considered. The comparison of the expression profiles within networks across contrasts reveal shared and distinct co-regulation patterns and complement the information from functional analysis. Networks depict gene relationships based on protein-protein interactions annotated in BIOGRID, HPRD, DIP, BIND, INTACT, and MINT databases [40, 46, 48].

Results and Discussion

Summary of RNA-Seq measurements

The quality and quantity of the RNA sequence reads was evaluated across samples. The number of reads and quality scores along the reads were comparable across samples from all 4 activity genotype-environment groups. The average quality score Phred of the reads was 30 indicating a 99.9% base call accuracy [49]. The percentage of reads mapped to the mouse genome was approximately 99% corresponding to an average of 129,718,676.5 total reads mapped per sample. Differential expression between genotype and environment groups was tested on 9691 genes.

Activity genotype-by-environment interaction effects on the transcriptome

The differential expression of individual genes uncovered in the present study supports accumulating evidence that the cerebellum is involved in motor control, regulation of neurotransmitter and hormone pathways, addictive behaviors [1], and neuropeptide and prohormone processes. Overall, 457 genes exhibited a significant (FDR-adjusted P-value < 0.05, comparable in this study to an unadjusted P-value < 0.001) genotype-by-environment interaction effect. A list of genes exhibiting the most significant (average log2(fold change) > |2|, FDR-adjusted P-value < 0.005) interaction effects is presented in Table 1. The complete list of genes significantly (FDR-adjusted P-value < 0.05) differentially expressed for the activity genotype-by-environment interaction effect is provided in Table A in S1 File. Annotation of many of these genes to signaling, sensory, and motor processes orchestrated in the cerebellum and involved in addictive behaviors [4] was noted.

thumbnail
Table 1. Genes exhibiting significant (FDR-adjusted P-value < 0.005, average contrast log2(fold change) > |2|) activity genotype-by-environment interaction.

https://doi.org/10.1371/journal.pone.0167095.t001

Among the genes exhibiting significant interaction, the number of significantly (FDR-adjusted P-value < 0.05) differentially expressed genes between specific contrasts were: 233 (CB-HB), 225 (CB-CF), 227 (CB-HF), 297 (HB-CF), 25 (HB-HF), and 248 (CF-HF). Transcriptome differences are more prevalent between activity genotypes (regardless of wheel-running availability in the environment), and between environments in the C genotype. Environmental differences were associated with substantially more differentially expressed genes (almost tenfold) in C genotype mice (CB-CF, 225 genes) than in H genotype mice (HB-HF, 25 genes). This result suggests that the environmental access to a running wheel, following previous opportunity to run on wheels, has limited impact on the transcriptome profiles beyond the impact of selection for high running at the genetic level. The environment that allows wheel running elicits differential gene expression in the C genotype that may already be activated in the H genotype.

Another finding from the evaluation of the 6 genotype-environment contrasts is that HB-CF exhibited the highest number of differentially expressed genes among all contrasts. This contrast compares two groups pairing “opposite” activity genotypes and environments (i.e. H genotype in B environment and C genotype in F environment). The abundance of differentially expressed genes in this contrast suggests that there may be synergistic effects between opposite levels of activity genotype and environment activity resulting in the highest number of gene differences. A possible hypothesis is that these transcript profiles support mechanisms enabling H genotype mice in the blocked-from-wheel running environment to experience exacerbated withdrawal behaviors whereas these mechanisms and associated withdrawal behaviors are not apparent in the C genotype mice in the F environment [16, 21].

Considering all genes differentially expressed at FDR-adjusted P-value < 0.05, the correlation between log2(CB-HB fold change) and log2(CB-CF fold change) is 0.77 (Table A in S1 File). In other words, the transcriptome comparison between control and high activity genotypes in this study within the B environment offers insights into the comparison between blocked and free activity environments within the C genotype. This finding supports our hypothesis that the physical activity genotypes studied are an effective model to understand the impact of availability of activity reward in the environment on the cerebellum transcriptome.

The contrast between environments within the H genotype (HB-HF) was selected a priori because this comparison would offer information on the lesser studied H genotype relative to the C genotype that is expected to resemble the baseline mouse population. A posteriori review of results indicated that environment differences elicited more extreme differential expression on the C relative to the H genotype among the genes exhibiting an interaction effect (Table 1 and Table A in S1 File). This result suggests that the environment reward conditions studied elicit comparable transcriptome responses in the H genotype whereas the cerebellum transcriptome of the C genotype is more susceptible to differences between the environments studied.

The profiles of the genes exhibiting significant genotype-by-environment interaction effect are consistent with the present understanding of the role of the cerebellum in locomotor regulation and the budding recognition of the role of cerebellum in reward-processing, goal focused- and reward-dependent behaviors. One common interaction profile among these genes was their simultaneous negative differential expression in the contrasts CB-HB, CB-CF, and CB-HF. This trend of under-expression in C relative to H genotype in the B environment, under-expression in the B relative to the F environment in the C genotype, and under-expression in the CB relative to the HF group was observed for P21 Protein Cdc42/Rac-Activated Kinase 6 (Pak6). This finding is consistent with a report of Pak6 deficiencies in mouse cerebellum associated with deficits in learning and locomotion [50]. The positive correlation between Pak6 abundance and activity genotype or environment detected in our study further validates the role of this gene on locomotor control, regardless of the genetic or environmental prompt.

Adenosine A2a receptor (Adora2a) had negative differential expression in the contrasts CB-HB, CB-CF, and CB-HF. This profile indicates that either H genotype or F environment alone or the combination of H genotype and F environment can trigger over-expression of this gene relative to the C genotype in the B environment. Consistent with our findings, studies have reported that adenosine receptor antagonists attenuate while adenosine uptake inhibitors augment ethanol-induced motor impairment in rodents via receptor populations in the cerebellum [51]. Adora2a is also involved in goal-focused behaviors and motor functions used to carry out those behaviors and in the absence of this receptor, mice display reduced exploratory behaviors and less motivation towards potentially addictive behaviors such as sexual activity and eating [52]. Moreover, Adora2a agonists significantly decrease cocaine induced hyperactivity in mice [53] and genetic variations on Adora2a affect psychomotor response speed and modulate the effect of the caffeine stimulant in humans [54]. The positive association between Adora2a mRNA abundance and the activity genotype motivated to run (H genotype) and environmental capability to run (F environment) uncovered in this study and with comparable phenotypes reported in previous studies reinforce the notion that this gene is a fundamental in the coordination of locomotion and reward-dependent behaviors.

Dopamine D2 receptor (Drd2) exhibited the same expression pattern as Adora2a. Drd2 had negative differential expression in the contrasts CB-HB, CB-CF, and CB-HF indicating that either H genotype or F environment alone or the combination of H genotype and F environment can trigger over-expression of this gene relative to the C genotype or B environment. The similarity between Drd2 and Adora2a expression patterns observed in this study supports prior implications of antagonistic activity between these genes. Drd2 knockout mice exhibit impaired locomotion and coordinated movements and treatment with Adora2a antagonists rescues the behavioral parameters [55]. Both the Drd2 knockout mice and the Adora2a antagonist treatment are extreme cases of the low Drd2 and Adora2a expression levels observed in the C genotype and B environment. The phenotypic associations observed on the former empirical studies support the differential transcriptional patterns between running genotype-environment groups uncovered in the present study.

The positive association between Drd2 level and motivation to run supports prior work on the effect of Drd2 in addiction and reward-dependent behaviors. Mice selectively bred for excessive exercise or obesity exhibit dopaminergic dysregulation [23]. Exercise elevates striatal dopamine D2 receptors in a mouse model of Parkinson’s disease [56] and dopamine receptors have been repeatedly detected in the cerebellar cortex [5]. Drd2 has been associated with addiction to opioids, nicotine, and cocaine [57] and mRNA encoding for Drd2 increase in response to cocaine treatment [58]. Studies of cerebellum in mice administered cocaine determined dysregulation in cerebral regions that have dopamine-signaling proteins [6, 59, 60]. In mice lacking D2 receptors, the rewarding properties of morphine were reduced [61]. The previous reports of positive associations between Drd2 levels and exercise and drugs of abuse are consistent with our findings of positive associations with high activity genotype and environment that may initiate, activate, and facilitate an inclination towards activity.

Copine 4 and 5 (Cpne4 and Cpne5, respectively) exhibited profiles similar to those observed in Adora2a and Drd2 (Table 1). Albeit less studied than Adora2a and Drd2, reports that genetic variants in Cpne5 were associated with alcohol dependence and obesity [62]. Cadherin 13 (Cdh13) exhibited a genotype-by-environment interaction effect and has been associated with d-amphetamine response [63] and with hyperactive or impulsive symptoms in attention deficit hyperactivity disorder cases [64]. The previous findings confirm the premise of a molecular link between high voluntary activity and goal- and reward-motivated behaviors through the cerebellum suggested in previous reports [22] and investigated in this study.

Another recurrent interaction profile displayed by transcripts such as Rho GTPase activating protein 8 (Arhgap8) and Stearoyl-coenzyme A desaturase 4 (Scd4) encompassed simultaneous gene over-expression in the CB-HB and CF-HF contrasts. Rho GTPase signaling is impaired in the cerebellum of a mice model of fetal alcohol syndrome [65] and several RhoGAPs are coded by the white rabbit gene in Drosophila, a gene that plays a key role in the regulation of both stimulant and sedative behaviors such as motor incoordination induced by ethanol [66]. Moreover, a study of the brain from mice with exercise-induced fatigue uncovered differential expression of Scd4 between fatigued and control mice [67]. The previous trends are consistent with our observations that Arhgap8 and Scd4 were over-expressed in mice from the C genotype compared to the H genotype in both activity environments.

Cytochrome P450, family 11, subfamily a, polypeptide 1 (Cyp11a1) was also over-expressed in the C genotype compared to the H genotype in both environments. The cerebellum hosts enzymatic process that allow neurosteroid synthesis and within this process the cholesterol side chain cleavage enzyme Cyp11a1 enables the production of pregnenolone. Alterations of the neurosteroidogenesis-related proteins are speculated to participate in dysregulated locomotor coordination involving cerebellar functions [68]. Also related to hormone processes, Peptidyl arginine deiminase, type IV (Padi4) exhibited the same pattern as Arhgap8 and Cyp11a1. Padi4 plays a role in hormone-induced transcription [69] and is a maker of oxidative stress, a process associated with drug addiction and psychiatric disorders [70, 71].

The expression of various neuropeptide prohormones and related receptor genes exhibited significant genotype-by-environment interaction. Prohormones included: adenylate cyclase activating polypeptide 1 (Adcyap1), cholecystokinin (Cck), somatostatin (Sst), VGF nerve growth factor inducible (Vgf), neuropeptide Y (Npy), neurotensin (Nts), proenkephalin (Penk), and tachykinin 2 (Tac2). The profile of these genes was characterized by significant under-expression in the CB group relative to the HB, CF and HF groups-HB and CB-CF contrasts. This profile suggests that the H genotype and F environment, individually or in combination, result in an elevation of the expression of these neuropeptide genes relative to the combination of baseline activity genotype and environment. Also, for neuropeptide genes, selection-control line differences in a non-reward environment elicited a transcriptome response similar to reward availability among unselected mice. This comparable impact of H genotype and F environment supports the paradigm that the running genotypes studied can serve as model for running availability environment and associated addiction and reward-dependent behaviors.

The gene expression trends observed in this study were consistent with previous studies of other strains of mice that considered neuropeptides in the cerebellum and addiction processes. The expression in the cerebellum of neuropeptides encoded in prohormone genes such as Penk, Npy, Nts, Sst, and Cck is well-reported [72] and associations with reward-dependent behaviors has been reported. Penk knock-out mice exhibit lower cannabinoid withdrawal syndrome than the corresponding wild type mice [73]. Npy, a peptide widely expressed in the cerebellum, modulates many functions and behaviors including anxiety that is associated with reward withdrawal conditions [74]. Moreover, the expression of Npy in the hypothalamus has been correlated with increased wheel running in rats [75]. A study of mice deficient in Sst demonstrated motor learning impairment, a function controlled by the cerebellum [76].

Transcripts of neuropeptide receptors, such as glucagon receptor (Gcgr), vasoactive intestinal peptide receptors 1 and 2 (Vipr1 and Vipr2), corticotropin releasing hormone receptor 2 (Crhr2), Npy receptor Npy1r, and opioid receptor delta 1 (Oprd1) also exhibited significant genotype-by-environment interaction effects and patterns similar to those detected in neuropeptide genes. Oprd1 modulates substance dependence risk [77]. Cerebellar astrocytes secrete neuroactive and vasoactive peptides including Npy in response to oxidative stress conditions [78]. Npy receptor knockout mice exhibited faster disappearance and reappearance of cocaine-induced conditioned place-preference and these receptors are expressed in the cerebellum [79]. Glucagon, vasoactive intestinal peptide, and secretin belong to the same family of bioactive peptides that have high homology and secretin receptors are abundant in the cerebellum [80]. Hemogloblin (Hba) A and B also exhibited significant genotype-by-environment interaction and prior studies have reported hemoglobin in mouse cerebellum [81]This result may imply the presence of peptides derived from hemoglobin, such as hemopressin, that have an important role in binding cannabinoid receptors [8284]. These genes were over-expressed CF-HF indicating that selection has resulted in lower levels of hemogloblin gene expression in an environment where the activity reward is available.

Fig 1 presents a Venn diagram illustrating the differentially (FDR-adjusted P-value < 0.05) expressed genes shared between CF-HF, CB-HB, and HB-HF contrasts. The limited overlap between contrasts depicted in this diagram confirms that many of the differences in gene expression between activity genotypes are dependent on the environment. While 5 genes were differentially expressed in all 3 contrasts, 68 genes were differentially expressed in both activity genotype contrasts (CB-HB and CF-HF). This overlap offers insights into molecular mechanisms affected by activity genotype that are shared across activity availability environments. Among the genes differentially expressed between activity genotypes in both environments, Neurogranin (Nrgn) and Dendrin (Ddn) appear in the list of genes exhibiting the most significant interaction effect (FDR-adjusted P-value < 0.005, average log2(fold change) > |2|) presented in Table 1. Both Nrgn and Ddn displayed the previously described frequent interaction profile including negative differential expression in the contrasts CB-HB and CB-CF.

thumbnail
Fig 1. Number of significantly differentially (FDR-adjusted P-value < 0.05) expressed genes that overlap between low activity control genotype-high activity environment and high activity genotype- high activity environment (CF-HF), low activity control genotype-low activity environment and high activity genotype- low activity environment (CB-HB), and high activity genotype-low activity environment and high activity genotype- high activity environment (HB-HF).

https://doi.org/10.1371/journal.pone.0167095.g001

The Ddn profile observed in this study is consistent with reports that this gene is over-expressed in response to 3,4-Methylenedioxymethamphetamine (‘ecstasy’) in the dorsal raphe nucleus of mice [85]. The neuropeptide-like peptide Nrgn is expressed in murine Golgi cells that are GABAergic interneurons [86] and ablation of these Golgi cells results in severe motor disorders and inability to perform compound movements [87]. Nrgn is also associated with alcohol addiction through the role in synaptic plasticity and signal transduction processes [88]. The under-expression of Nrgn in the C genotype and B environment is consistent with limited motor coordination and movements observed in mice without Nrgn-rich Golgi cells previously reported. This finding, together with the association between addiction behaviors and Nrgn and Dnd, support the hypothesis that the genotypes and environments compared in this study offer insights into transcriptome underlying locomotor and reward-dependent phenotypes.

The consideration of the 3 genotype-by-environment pairwise contrasts that include the baseline CB group offers a complementary view of the impact of activity interaction effects. Fig 2 illustrates the overlap in differentially (FDR-adjusted P-value < 0.05) expressed genes between all 3 contrasts that include the CB baseline group. Overall, 164 differentially expressed genes were shared between all 3 contrasts involving the genotype-environment combination CB, suggesting that either H genotype, F environment, or the combination of both conditions can elicit similar transcriptome changes relative to the CB baseline. Furthermore, this complementary perspective confirms that genotypic make-up leading to high voluntary activity is associated with changes in gene expression comparable to environmental wheel running restriction following a period of availability.

thumbnail
Fig 2. Number of significantly differentially (FDR-adjusted P-value < 0.05) expressed genes that overlap between low activity control genotype-low activity environment and high activity genotype- low activity environment (CB-HB), low activity control genotype-low activity environment and low activity control genotype- high activity environment (CB-CF), and low activity control genotype-low activity environment and high activity genotype- high activity environment (CB-HF).

https://doi.org/10.1371/journal.pone.0167095.g002

Functional categories enriched among genes exhibiting activity genotype-by-environment interaction effects

The functional categories enriched among the genes exhibiting significant genotype-by environment interaction included processes associated with motor control and reward-dependent behaviors. This finding confirms the role of the cerebellum in locomotor regulation and offers evidence that the cerebellum also plays a role in molecular signaling connected with addictive behaviors [1]. Table 2 lists selected categories and functional clusters that were enriched among the genes exhibiting significant interaction (FDR-adjusted P-value < 0.05) by orthogonal contrast (CB-HB, CF-HF, and HB-HF) and that have an enrichment score > 1.5. A complete list of terms and enriched clusters is presented in Table B in S1 File.

thumbnail
Table 2. Functional clusters that have enrichment score > 1.5 and corresponding Gene Ontology (GO) biological process (BP), molecular function (MF) Functional Annotation Tool (FAT) categories, and KEGG pathways among the genes that exhibited significant genotype-by-environment interaction (FDR-adjusted P-value < 0.05) and differentially expressed (FDR-adjusted P-value < 0.05) in the 3 orthogonal genotype-by-environment contrasts (CB-HB, CF-HF, and HB-HF).

https://doi.org/10.1371/journal.pone.0167095.t002

Central themes among the genes differentially expressed in the CB-HB contrast are the enrichment of regulation of locomotor behavior, neurotransmitter and neuropeptide binding processes, cyclic adenosine monophosphate (cAMP) metabolic process, neuro-related process, and response to drug. Locomotor behavior and neurotransmitter and neuropeptide categories are consistent with our previous discussion at the individual gene level. Enrichment of cAMP is consistent with reports that dopamine 2-like receptors inhibit adenylate cyclase and thus reduce the concentration of cAMP whereas dopamine 1-like receptors have the opposite effect [89]. The results from the functional analysis of the CB-HB contrast are consistent with studies reporting association between physical activity and neurological system processes [90, 91]. Among the genes that exhibited a significant interaction effect, the high correlation of log2(fold change) between CB-HB and CB-CF and comparable enriched categories confirm that the processes impacted by the selection for high physical activity are similarly impacted by the availability of activity reward in the environment. Enrichment of neuropeptide signaling processes encompassed mostly under-expressed neuropeptide genes including Adcyap1, Oprd1, melanin-concentrating hormone receptor 1(Mchr1), Npy, Npy1r and Penk and over-expressed genes Npy2r and Tac2.

Weaker enrichment trends were observed among the genes differentially expressed in the CF-HF contrast meanwhile the limited number of differentially expressed genes prevented significant enrichment in the HB-HF contrast (Table 2). The enrichment oxygen transport processes with a predominance of genes under-expressed in the CF-HF contrast may be associated with regulation of locomotion. The functional analyses offer a glimpse of how genetic selection for high voluntary running has altered the cerebellum molecular processes and may be behind the behavioral inclination, motivation, or addiction towards high physical activity as well as impacting their physical characteristics [15].

Network of genes exhibiting activity genotype-by-environment interaction effects

Our previous findings offer insights into the expression profile of genes both individually and in functional clusters. Network analysis aids in understanding the relationship between genes and in assessing the connectivity among genes. The relationship among genes exhibiting significant (FDR-adjusted P-value < 0.05) activity genotype-by-environment interaction was depicted using networks. This visualization contributed to our understanding of the similarities in gene co-regulation between activity genotypes within the B and F environments (CB-HB and CF-HF in Figs 3 and 4, respectively) and between activity environments in the H genotype (HB-HF in Fig 5). Rectangular nodes represent the differentially expressed genes and edges represent the known associations between genes based on curated databases of molecular interactions.

thumbnail
Fig 3. Network of genes significantly differentially expressed (FDR-adjusted P-value < 0.05) in the Control genotype-Blocked environment and High activity genotype-Blocked environment (CB-HB) orthogonal contrast.

(Node Color: Red indicates under-expression, Green indicates over-expression, Gray indicates genes connecting differentially expressed gene but not differentially expressed in this study; Node Size: larger sizes indicates relatively higher fold changes, while smaller sizes indicates relatively smaller fold changes).

https://doi.org/10.1371/journal.pone.0167095.g003

thumbnail
Fig 4. Gene network of genes significantly differentially expressed (FDR-adjusted P-value < 0.05) in the Control genotype-Free activity environment and High activity genotype-Free environment (CF-HF) orthogonal contrast.

(Node Color: Red indicates under-expression, Green indicates over-expression, Gray indicates genes connecting differentially expressed gene but not differentially expressed in this study; Node Size: larger sizes indicates relatively higher fold changes, while smaller sizes indicates relatively smaller fold changes).

https://doi.org/10.1371/journal.pone.0167095.g004

thumbnail
Fig 5. Gene network of genes significantly differentially expressed (FDR-adjusted P-value < 0.05) in the High activity genotype-Blocked environment and High activity genotype-Free environment (HB-HF) orthogonal contrast.

(Node Color: Red indicates under-expression, Green indicates over-expression, Gray indicates genes connecting differentially expressed gene but not differentially expressed in this study; Node Size: larger sizes indicates relatively higher fold changes, while smaller sizes indicates relatively smaller fold changes).

https://doi.org/10.1371/journal.pone.0167095.g005

The distribution of profiles in the networks CF-HF (comparing genotypes within the F environment) and CB-HB (comparing genotypes within the B genotype) appear more extreme meanwhile the distribution of profiles in the network HB-HF (comparing environments within the H genotype) appears intermediate. Meanwhile a number of genes are under-expressed in the CB-HB network, many genes are over-expressed in the CF-HF network. Also, the differential expression level appears more consistent across the CF-HF network than in the CB-HB network. These topologies confirm the differential response of the cerebellum transcriptome from the H genotype relative to the C genotype between activity environments.

The distribution of high differentially expressed genes between genotypes in the periphery of the network in the B environment (Fig 3) and in the center of the network in the H environment (Fig 4) offered insights into the interaction between genotype and environment. This comparison of topologies suggests that selection for high voluntary running is linked to a predominant dysregulation of hub genes in an environment that enables running whereas a dysregulation of ancillary genes is favored in an environment that blocks running. The premise that core components of a network are usually conserved and central in terms of functions, while periphery components are usually more variable and specialized [92] may offer insights into our finding that core genes tended to be dysregulated in the F environment while periphery genes tended to be dysregulated in the B environment. Based on this premise we postulate that expression changes in highly connected genes in the cerebellum of H genotype mice may be associated with the capacity of these mice to act upon their motivation to run and to execute high wheel running in the F environment

The evaluation of network components indicates that Laminin, Gamma 2 (Lamc2), Fas Cell Surface Death Receptor (Fas), and Ribosomal Protein L35a (Rpl35a) exhibited a similar differential pattern in the genotype contrast CB-HB and CF-HF. These genes were significantly over-expressed in mice from the H relative to the C genotype in both activity environments. The association of these genes with genotype activity dependence, irrespectively of reward availability in the environment suggests a fundamental disruption in the expression level that the environment is not able to offset. These genes have been associated with addictive behaviors. Lamc2 is down-regulated by acute nicotine exposure [93], Fas receptor proteins are modulated during opiate addiction in the rat brain [94], and Rpl35a was dysregulated in cell lines treated with nicotine [95]. LIM Homeobox 9 (Lhx9) was also over-expressed in the CF-HF contrast. Lxhr9 regulates the neuropeptide orexin that is involved in arousal, appetite, wakefulness, and reward processes [96]. The involvement of orexin in arousal wakefulness and reward systems [97] makes this gene a good candidate to understand the molecular mechanisms underlying the motivation to run.

Further consideration of the network components uncovers 6 genes including Nrgn, Drd2, Retinoid X Receptor Gamma (Rxrg), Guanine Deaminase (Gda), Adora2a, and Ras-Related Protein Rab-40B (Rab40b) that display opposite patterns between the CB-HB and CF-HF gene networks. This pattern indicates that environmental activity availability impacted the change in expression level between C and H genotypes. These genes are over-expressed in CF-HF and under-expressed in CB-HB. These trends confirm that genotype differences must be evaluated in the environmental context and this context-dependency may be applicable to other addictive behaviors.

Among the genes that exhibited distinct patterns between the CB-HB and CF-HF networks, Gda and Rxrg were under-expressed in C relative to H genotype mice in the B environment but over-expressed in the H environment. Under the premise that the comparison of H and C genotypes can help us understand reward dependent behaviors, the differential expression of Gda between genotypes in the B environment is consistent with a report of Gdaunder-expression in the brain of monkeys that self-administered cocaine [98]. Likewise, the differential expression of Rxrg between genotypes in the B environment is consistent with reports that over-expression of Rxrg in humans is associated with sensation seeking [99] and that Rxrg knock-out mice show impaired locomotion along with reduced dopamine signaling [100]. Also, ablation/loss of Rxrg signaling in mice leads to depressive behaviors due to the decreased dopamine signaling in the striatum, supporting the association of this gene with reward and pleasure [101]. In agreement with dopamine receptor trends previously described, Rxrg is a transcription factor that regulates striatal D1r gene expression [102] and retinoid receptors directly up-regulate Drd2 [103]. The present study uncovered several genes that exhibited environment-dependent differential expression between genotypes. This finding demonstrates the need to account for reward availability context or environments in studies of molecular mechanisms associated with high voluntary running. This finding is likely applicable to other genes and reward-dependent behaviors.

Most of the genes over- or under- expressed in the networks of activity genotype contrasts (CB-HB and CF-HF) are only intermediately differentially expressed between activity environments in the H genotype (HB-HF). This result is consistent with the overlap in differentially expressed genes between the 3 orthogonal contrasts presented in Fig 1. This result suggests that environment is not a major factor confounding the results of our genotype comparisons. The only exception is Midline 1 (Mid1). This gene was significantly over-expressed in the CB-HB contrast yet was significantly under-expressed in the HB-HF. In fact, Mid1 was the only significantly under-expressed gene in this network. This result could be linked to reports that Mid 1 is repressed in daDREAM mice that also exhibit significant suppression of neuromotor development early after birth [104]. Our finding suggests that a reward-available environment can counteract the dysregulation in Mid1 expression associated with high voluntary activity genotype.

Effect of high voluntary activity genotype on the cerebellum transcriptome

In addition to environmental-dependent, differential expression patterns between activity genotypes that are environmental-independent were uncovered. Table 3 summarizes genes differentially expressed (log2(fold change) > |3| and FDR-adjusted P-value < 0.007) between low and high activity genotypes. A list of genes significantly differentially expressed at FDR-adjusted P-value < 0.05 is provided in Table C in S1 File. The majority of the 183 differentially expressed genes were under-expressed in the C relative to the H genotype mice. Selected genes that exhibited differential expression between the genotype groups and that can offer insights into high activity genotype differences in the cerebellum transcriptome are discussed.

thumbnail
Table 3. Genes significantly differentially expressed (FDR-adjusted P-value < 0.007 and log2(fold change) > |3|) between mice from Control (C) and High (H) activity genotypes.

https://doi.org/10.1371/journal.pone.0167095.t003

Several genes differentially expressed between the C and H genotype mice have been associated with locomotor control, reward-dependent behaviors, neurological systems and signaling pathways. Dopamine receptor D1 (Drd1) was over-expressed in the cerebellum of H relative to C genotype mice. This result is consistent with the over-expression of Drd1 in the caudate-putamen of cocaine-sensitized mice relative to control [105] and reports that dopamine receptors D1 and D2 systems mediate the effects of cocaine on cerebellar neurons of adult rats [106]. Also, endogenous drug administration simulates D1 receptor, which leads to the expression of D2 receptor-mediated reward-driven behaviors and gene expression changes in non-cerebellar brain regions [105, 107110]. Both D1-like and D2-like dopamine receptors are involved in the response of rats who are taught to perform a behavior in order to receive pleasurable electrical stimulation [111, 112]. All together, these results offer evidence of the involvement of Drd1 expression on reward-dependent behaviors and confirm that the High Runner line studied here is helpful to understand the transcriptome associated with these behaviors.

Muscarinic Acetylcholine Receptor M1 (Chrm1) was also over-expressed in mice from the H relative to C activity genotype. A study of cerebellar samples reported associations between polymorphisms in Chrm1 and modifications in aspects of executive function in humans [113]. This gene is related to dopamine transmission [114, 115] and further supports the notion that mice from the studied high activity genotype share disruption of dopamine pathways with drug and other reward-dependent behaviors. Other genes over-expressed in mice from the H relative to C activity genotype include the transcription factor Forkhead Box G1 (Foxg1). Studies of cerebellar granule neurons report that this gene promotes neuronal survival [116]. Differential expression of endothelial Lipase (Lipg) between the H and C genotypes was observed. Lipg is associated with neurite pathology and is detected in the cerebellum [117]. The association with addictive behaviors of these genes differentially expressed between H and C activity genotypes highlights the genetic pathways shared between high voluntary activity and neurological, cognitive, and addiction processes.

Results from the functional enrichment analysis of the genes differentially (FDR-adjusted P-value < 0.05) expressed between activity genotypes are consistent with previous findings about processes in the cerebellum associated with locomotor control and offer additional evidence on the link between high activity genotype and reward-dependent behaviors [1]. Terms enriched included voltage-gated cation channel activity (Enrichment Score = 5.36), GTPase regulatory activity (Enrichment Score = 2.04), neuroactive ligand-receptor interaction (Enrichment Score = 1.73), metal ion binding (Enrichment Score = 1.59), regulation of cAMP metabolic process (Enrichment Score = 1.48), neurotransmitter receptor activity (Enrichment Score = 1.25), and neuron projection morphogenesis (Enrichment Score = 1.18). The most significantly enriched cluster is highlighted in Table 4. The majority of the terms were related to ion binding and activity. This result is consistent with a study of working memory-related activity in the cerebellum that uncovered enrichment of voltage-gated cation channel activity encompassing genes related to neuronal excitability [118]. Working memory is the active maintenance of task-relevant information during a cognitive task and is expected to be important in activities such as wheel running. Our result suggests that selective breeding for high activity has changed the expression of a number of transcripts, many of which are not environment-dependent (e.g. Drd1 and Chrm1). These genes have been annotated to processes supporting motivation or addiction-like behavior towards activity, including signaling and locomotor control, thus enabling the mice to execute high intensity exercise. Some of these processes are shared with genes that exhibit environment-dependent differential expression between genotypes. Therefore, conclusions from systems biology studies of high voluntary running lines should consider the context or environment of the study.

thumbnail
Table 4. Enriched (enrichment score = 5.36) cluster of Gene Ontology (GO) biological process (BP), molecular function (MF) Functional Annotation Tool (FAT) categories among the genes significantly differentially expressed (FDR-adjusted P-value < 0.05) between mice from Control (C) and High (H) activity genotypes.

https://doi.org/10.1371/journal.pone.0167095.t004

Effect of activity environment on gene expression in the cerebellum

Differential expression patterns between activity environments that are genotype-independent were uncovered in addition to genotype-dependent patterns. Table 5 summarizes genes differentially expressed (log2(fold change) > |4|, FDR-adjusted P-value < 0.007) between the B and F environments. A more complete list of genes differentially expressed at FDR-adjusted P-value < 0.05 is provided within Table D in S1 File. The majority of the 148 differentially expressed genes in the cerebellum were under-expressed in the B relative to the F environment mice. Moreover, many of the genes under-expressed in the B relative to the F environment groups (e.g. Foxg1, Prss56, Dlx1) were also under-expressed in the C relative to the H genotype mice.

thumbnail
Table 5. Genes significantly differentially expressed (FDR-adjusted P-value < 0.007) and log2(fold change) > |4|) between mice in Blocked (B) and Free (F) activity environments.

https://doi.org/10.1371/journal.pone.0167095.t005

The correlation between the log2(fold changes) of the 123 genes both differentially expressed between activity genotypes (Table C in S1 File) and between activity environments (Table D in S1 File) was 0.98. Furthermore, among the 207 genes differentially expressed in either comparison (genotype or environment) the correlation in log2(fold change) ranged between 0.81 and 0.84. The similar gene expression profiles between the C—H and B—F contrasts indicate that the genotypes for activity reward compared in this study are a robust model to study the impact of reward availability in the environment on the cerebellum transcriptome, provided the genes do not exhibit significant interaction effects.

Terms enriched among differentially expressed genes between activity availability environments were similar to the terms identified for the genotype contrast and included voltage-gated cation channel activity (Enrichment Score = 5.43), neuroactive ligand-receptor interaction (Enrichment Score = 2.77), neuron projection morphogenesis (Enrichment Score = 2.11), GTPase regulatory activity (Enrichment Score = 1.79), regulation of cAMP metabolic process (Enrichment Score = 1.44), positive regulation of transcription (Enrichment Score = 1.11), and metal ion binding (Enrichment Score = 0.99). A notable finding is that neurotransmitter receptor activity did not reach enrichment level in the environment contrast unlike in the genotype contrast. The most enriched cluster is highlighted in Table 6.

thumbnail
Table 6. Enriched (enrichment score = 5.43) clusters of Gene Ontology (GO) biological process (BP) and molecular function (MF) Functional Annotation Tool (FAT) categories among the genes significantly differentially expressed (P-value < 0.001) between mice in Blocked (B) and Free (F) activity environments.

https://doi.org/10.1371/journal.pone.0167095.t006

Motivation towards high activity and addiction processes

Reward-motivated responses, such as those associated with addictive behaviors, are regulated by the dopaminergic mesocorticolimbic system in the brain [119121]. Substances of abuse, including opiates, cocaine, amphetamine, alcohol, and nicotine, are addictive pleasurable stimuli that impact the neurotransmission of dopamine in a manner similar to other more natural stimuli including specific foods or liquids, gambling, shopping, and exercise or running [15, 17, 122129]. While addiction to exercise is not considered an official addictive disorder [130], 3% to 5% of the US population experience this condition [131].

Our investigation of cerebellum transcriptomic differences between mice from high wheel running and control genotypes in environments enabling and blocking access to wheel running uncovered enrichment of dopamine and related signaling pathways, and neurotransmitter and neuropeptide-dependent processes. Dysregulation of genes in these processes have also been linked to reward-dependent behaviors. Our findings support the premise that the line selected for high running and motivated to run studied here can be an effective model to study the transcriptome associated with addictive disorders.

Conclusions

Comparison of the cerebellum transcriptome between H (High Runner) and C (Control) mouse lines in environments with F (free) or B (blocked) access to a running wheel on day 7 provided novel insights into environment dependent and independent differential gene expression and networks. Overall, 457 genes exhibited significant (FDR-adjusted P-value < 0.05) activity genotype-by-environment interaction effect. The number of genes differentially expressed in the orthogonal contrast between genotype-environment groups CB-HB, CB-CF, and CF-HF were 233, 225, and 248, respectively. These genes included Pak6, Adora2a, Drd2, and Arhgap8, neuropeptide prohormones: Adcyap1, Cck, Sst, Vgf, Npy, Nts, Penk, and Tac2 and related receptors: Gcgr, Vipr1, Vipr2, and Crhr2. These findings support the need to characterize transcriptome dysregulation between high voluntary running and control lines in an environment or context-dependent manner. This recommendation may be relevant to other reward-dependent behaviors.

The networks of the CB-HB and CF-HF contrasts exhibited extreme and opposite gene expression profiles and the differentially expressed genes were more connected in the CF-HF network. This suggest that expression changes in highly connected genes in the cerebellum of H genotype mice may be associated with the capacity of these mice to act upon their motivation to run and to execute high wheel running in the F environment.

The majority of the 183 differentially expressed genes between genotypes (including Drd1) were under-expressed in the cerebellum of C relative to H genotype mice. Also, the majority of the 148 differentially expressed genes were under-expressed in the cerebellum of mice in the B relative to F activity environment. Of these, 123 genes were differentially expressed in both comparisons.

Drug induced activity-dependent synaptic modifications in the cerebellum have been postulated to be key in the transition from a pattern of recreational drug taking to the compulsive behavioral phenotype [1]. This characterization is complemented by our study of the cerebellum that identified changes in the expression of genes associated with the motivation and reward-dependent behaviors in a mouse line selected for high running.

Our findings demonstrate that the cerebellum of a mouse line selected for high voluntary running share transcriptome patterns with reward-dependent behaviors, such as addiction to psychoactive substances. Significant changes in the expression of genes affiliated to dopamine-related and neuropeptide-dependent processes and their association with motivation and reward dependent behaviors and locomotor control were explored. The high voluntary activity selection line of mice evaluated in this study is a helpful model for understanding molecular mechanisms behind motivation, reward-dependent, and addictive behaviors.

Supporting Information

S1 File. Supplementary results.

Table A. Genes exhibiting significant (FDR-adjusted P-value < 0.05) activity genotype-by-environment interaction and Log2(Fold Change) by pairwise contrast encompassing genotype (C and H) and environment (B and F) groups. Table B. Enriched clusters of Gene Ontology (GO) biological process (BP), molecular function (MF) Functional Annotation Tool (FAT) categories, and KEGG pathways among the genes that exhibited significant genotype-by-environment interaction (FDR-adjusted P-value < 0.05) and differentially expressed (FDR-adjusted P-value < 0.05) between mice from genotype-by-environment groups in the 3 orthogonal contrasts (CB-HB, CF-HF, and HB-HF). Table C. Genes differentially expressed (FDR-adjusted P-value < 0.05) between mice from the Control and High activity genotypes. Table D. Genes differentially expressed (FDR-adjusted P-value < 0.05) between mice in Blocked and Free activity environments.

https://doi.org/10.1371/journal.pone.0167095.s001

(DOCX)

Author Contributions

  1. Conceptualization: SRZ JSR.
  2. Data curation: BRS.
  3. Formal analysis: SRZ KCA BRS.
  4. Funding acquisition: SRZ JSR TG.
  5. Investigation: TG SDP AGH.
  6. Methodology: SRZ JSR BRS.
  7. Project administration: SRZ JSR TG.
  8. Resources: JSR TG AGH.
  9. Software: SRZ KCA BRS.
  10. Supervision: SRZ JSR.
  11. Validation: BRS.
  12. Writing – original draft: SRZ KCA.
  13. Writing – review & editing: SRZ KCA JSR TG BRS SDP AGH.

References

  1. 1. Miquel M, Toledo R, García LI, Coria-Avila GA, Manzo J. Why should we keep the cerebellum in mind when thinking about addiction? Current Drug Abuse Reviews. 2009;2(1):26–40. pmid:19630735
  2. 2. Strick PL, Dum RP, Fiez JA. Cerebellum and nonmotor function. Annual Review of Neuroscience. 2009;32:413–34. pmid:19555291
  3. 3. Anderson CM, Maas LC, deB Frederick B, Bendor JT, Spencer TJ, Livni E, et al. Cerebellar vermis involvement in cocaine-related behaviors. Neuropsychopharmacology. 2006;31(6):1318–26. pmid:16237382
  4. 4. Yalachkov Y, Kaiser J, Naumer MJ. Sensory and motor aspects of addiction. Behavioural Brain Research. 2010;207(2):215–22. pmid:19751771
  5. 5. Carbo-Gas M, Vazquez-Sanroman D, Aguirre-Manzo L, Coria-Avila GA, Manzo J, Sanchis-Segura C, et al. Involving the cerebellum in cocaine-induced memory: pattern of cFos expression in mice trained to acquire conditioned preference for cocaine. Addiction Biology. 2014;19(1):61–76. pmid:23445190
  6. 6. Carbo-Gas M, Vazquez-Sanroman D, Gil-Miravet I, De las Heras-Chanes J, Coria-Avila GA, Manzo J, et al. Cerebellar hallmarks of conditioned preference for cocaine. Physiology & Behavior. 2014;132:24–35.
  7. 7. Tomasi D, Wang GJ, Wang R, Caparelli EC, Logan J, Volkow ND. Overlapping patterns of brain activation to food and cocaine cues in cocaine abusers. Human Brain Mapping. 2015;36(1):120–36. pmid:25142207
  8. 8. Stoodley CJ. The cerebellum and cognition: evidence from functional imaging studies. The Cerebellum. 2012;11(2):352–65. pmid:21373864
  9. 9. Taylor JA, Ivry RB. Cerebellar and prefrontal cortex contributions to adaptation, strategies, and reinforcement learning. Progress in Brain Research. 2014;210:217–53. pmid:24916295
  10. 10. Thoma P, Bellebaum C, Koch B, Schwarz M, Daum I. The cerebellum is involved in reward-based reversal learning. The Cerebellum. 2008;7(3):433–43. pmid:18592331
  11. 11. Schutter DJLG, Van Honk J. The cerebellum on the rise in human emotion. The Cerebellum. 2005;4(4):290–4. pmid:16321885
  12. 12. Caston J, Chianale C, Delhaye-Bouchaud N, Mariani J. Role of the cerebellum in exploration behavior. Brain Research. 1998;808(2):232–7. pmid:9767169
  13. 13. Dahhaoui M, Lannou J, Stelz T, Caston J, Guastavinot J. Role of the cerebellum in spatial orientation in the rat. Behavioral and Neural Biology. 1992;58(3):180–9. pmid:1456939
  14. 14. Garland T, Schutz H, Chappell MA, Keeney BK, Meek TH, Copes LE, et al. The biological control of voluntary exercise, spontaneous physical activity and daily energy expenditure in relation to obesity: human and rodent perspectives. Journal of Experimental Biology. 2010;214(2):206–29.
  15. 15. Kolb EM, Rezende EL, Holness L, Radtke A, Lee SK, Obenaus A, et al. Mice selectively bred for high voluntary wheel running have larger midbrains: support for the mosaic model of brain evolution. The Journal of Experimental Biology. 2013;216(Pt 3):515–23. pmid:23325861
  16. 16. Rhodes JS, Gammie SC, Garland T Jr. Neurobiology of mice selected for high voluntary wheel-running Activity. Integrative and Comparative Biology. 2005;45(3):438–55. pmid:21676789
  17. 17. Kolb EM, Kelly SA, Garland T Jr. Mice from lines selectively bred for high voluntary wheel running exhibit lower blood pressure during withdrawal from wheel access. Physiology & Behavior. 2013;112–113(0):49–55. http://dx.doi.org/10.1016/j.physbeh.2013.02.010.
  18. 18. Swallow JG, Carter PA, Garland TJ. Artificial selection for increased wheel-running behavior in house mice. Behavior Genetics. 1998;28(3):227–37. pmid:9670598
  19. 19. Uyttenhove C, Pilotte L, Theate I, Stroobant V, Colau D, Parmentier N, et al. Evidence for a tumoral immune resistance mechanism based on tryptophan degradation by indoleamine 2,3-dioxygenase. Nature Medicine. 2003;9(10):1269–74. pmid:14502282
  20. 20. Waters RP, Pringle RB, Forster GL, Renner KJ, Malisch JL, Garland T Jr, et al. Selection for increased voluntary wheel-running affects behavior and brain monoamines in mice. Brain Research. 2013;1508(0):9–22. http://dx.doi.org/10.1016/j.brainres.2013.01.033.
  21. 21. Rhodes JS, Garland T, Gammie SC. Patterns of brain activity associated with variation in voluntary wheel-running behavior. Behavior Neuroscience 2003;117(6):1243–56. pmid:14674844.
  22. 22. Brown JD, Green CL, Arthur IM, Booth FW, Miller DK. Cocaine-induced locomotor activity in rats selectively bred for low and high voluntary running behavior. Psychopharmacology. 2015;232(4):673–81. pmid:25106389
  23. 23. Mathes WF, Nehrenberg DL, Gordon R, Hua K, Garland T Jr., Pomp D. Dopaminergic dysregulation in mice selectively bred for excessive exercise or obesity. Behavioural Brain Research. 2010;210(2):155–63. pmid:20156488
  24. 24. Keeney BK, Meek TH, Middleton KM, Holness LF, Garland T Jr. Sex differences in cannabinoid receptor-1 (CB1) pharmacology in mice selectively bred for high voluntary wheel-running behavior. Pharmacology, Biochemistry, and Behavior. 2012;101(4):528–37. pmid:22405775
  25. 25. Majdak P, Bucko PJ, Holloway AL, Bhattacharya TK, DeYoung EK, Kilby CN, et al. Behavioral and pharmacological evaluation of a selectively bred mouse model of home cage hyperactivity. Behavior Genetics. 2014;55:516–34.
  26. 26. Malisch JL, Breuner CW, Kolb EM, Wada H, Hannon RM, Chappell MA, et al. Behavioral despair and home-cage activity in mice with chronically elevated baseline corticosterone concentrations. Behavior Genetics. 2009;39(2):192–201. pmid:19067154
  27. 27. Belke TW, Garland T. A brief opportunity to run does not function as a reinforcer for mice selected for high daily wheel-running rates. Journal of the Experimental Analysis of Behavior. 2007;88(2):199–213. pmid:17970415
  28. 28. Alexander GE, DeLong MR, Strick PL. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neuroscience. 1986;9:357–81. pmid:3085570
  29. 29. Balleine BW, Delgado MR, Hikosaka O. The role of the dorsal striatum in reward and decision-making. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 2007;27(31):8161–5. 27/31/8161.
  30. 30. Penney JB Jr., Young AB. Speculations on the functional anatomy of basal ganglia disorders. Annual Review of Neuroscience. 1983;6:73–94. pmid:6838141
  31. 31. Kohman RA, Rodriguez-Zas SL, Southey BR, Kelley KW, Dantzer R, Rhodes JS. Voluntary wheel running reverses age-induced changes in hippocampal gene expression. PLoS One. 2011;6(8):e22654. pmid:21857943
  32. 32. Claghorn GC, Fonseca IAT, Thompson Z, Barber C, Garland T. Serotonin-mediated central fatigue underlies increased endurance capacity in mice from lines selectively bred for high voluntary wheel running. Physiology & Behavior. 2016;161:145–54.
  33. 33. Gibbons TE, Pence BD, Petr G, Ossyra JM, Mach HC, Bhattacharya TK, et al. Voluntary wheel running, but not a diet containing (−)-Epigallocatechin-3-gallate and β-Alanine, improves learning, memory and hippocampal neurogenesis in aged mice. Behavioural Brain Research. 2014;272:131–40. pmid:25004447
  34. 34. Nixon SE, Gonzalez-Pena D, Lawson MA, McCusker RH, Hernandez AG, O'Connor JC, et al. Analytical workflow profiling gene expression in murine macrophages. Journal of Bioinformatics and Computational Biology. 2015;13(2):1550010. pmid:25708305
  35. 35. Caetano-Anollés K, Mishra S, Rodriguez-Zas SL. Synergistic and antagonistic interplay between Myostatin gene expression and physical activity levels on gene expression patterns in Triceps Brachii muscles of C57/BL6 mice. PloS One. 2015;10(2):e0116828. pmid:25710176
  36. 36. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature Protocols. 2012;7(3):562–78. pmid:22383036
  37. 37. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B (Methodological). 1995:289–300.
  38. 38. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols. 2009;4(1):44–57. pmid:19131956
  39. 39. Delfino KR, Rodriguez-Zas SL. Transcription factor-microRNA-target gene networks associated with ovarian cancer survival and recurrence. PloS One. 2013;8(3):e58608. pmid:23554906
  40. 40. Delfino KR, Serao NV, Southey BR, Rodriguez-Zas SL. Therapy-, gender- and race-specific microRNA markers, target genes and networks related to glioblastoma recurrence and survival. Cancer Genomics & Proteomics. 2011;8(4):173–83. 8/4/173.
  41. 41. Serao NV, Gonzalez-Pena D, Beever JE, Bollero GA, Southey BR, Faulkner DB, et al. Bivariate genome-wide association analysis of the growth and intake components of feed efficiency. PloS One. 2013;8(10):e78530. pmid:24205251
  42. 42. Serao NV, Gonzalez-Pena D, Beever JE, Faulkner DB, Southey BR, Rodriguez-Zas SL. Single nucleotide polymorphisms and haplotypes associated with feed efficiency in beef cattle. BMC Genetics. 2013;14:94-2156-14-94. pmid:24066663
  43. 43. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences. 2005;102(43):15545–50. pmid:16199517
  44. 44. Gonzalez-Pena D, Nixon SE, O'Connor JC, Southey BR, Lawson MA, McCusker RH, et al. Microglia transcriptome changes in a model of depressive behavior after immune challenge. PloS One. 2016;11(3):e0150858. pmid:26959683
  45. 45. Killcoyne S, Carter GW, Smith J, Boyle J. Cytoscape: a community-based framework for network modeling. Methods in Molecular Biology (Clifton, NJ). 2009;563:219–39. pmid:19597788
  46. 46. Martin A, Ochagavia M, Rabasa L, Miranda J, Fernandez-de-Cossio J, Bringas R. BisoGenet: a new tool for gene network building, visualization and analysis. BMC Bioinformatics. 2010;11(1):1–9. pmid:20163717
  47. 47. Gonzalez-Pena D, Nixon SE, Southey BR, Lawson MA, McCusker RH, Hernandez AG, et al. Differential transcriptome networks between IDO1-knockout and wild-type mice in brain microglia and macrophages. PloS One. 2016;11(6):e0157727. pmid:27314674
  48. 48. Serao NV, Delfino KR, Southey BR, Beever JE, Rodriguez-Zas SL. Cell cycle and aging, morphogenesis, and response to stimuli genes are individualized biomarkers of glioblastoma progression and survival. BMC Medical Genomics. 2011;4:49-8794-4-49. pmid:21649900
  49. 49. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for transcriptomics. Nature Reviews Genetics. 2009;10(1):57–63. pmid:19015660
  50. 50. Nekrasova T, Jobes ML, Ting JH, Wagner GC, Minden A. Targeted disruption of the Pak5 and Pak6 genes in mice leads to deficits in learning and locomotion. Developmental Biology. 2008;322(1):95–108. pmid:18675265
  51. 51. Brown RM, Short JL. Adenosine A(2A) receptors and their role in drug addiction. The Journal of Pharmacy and Pharmacology. 2008;60(11):1409–30. pmid:18957161
  52. 52. Short JL, Ledent C, Drago J, Lawrence AJ. Receptor crosstalk: characterization of mice deficient in Dopamine D1 and Adenosine A2A receptors. Neuropsychopharmacology: Official Publication of the American College of Neuropsychopharmacology. 2006;31(3):525–34.
  53. 53. Poleszak E, Malec D. Cocaine-induced hyperactivity is more influenced by adenosine receptor agonists than amphetamine-induced hyperactivity. Polish Journal of Pharmacology. 2002;54(4):359–66. pmid:12523489
  54. 54. Landoltl H. Editor Genetics, Adenosine and sleep-wake systems: how coffee keeps us awake. Fundamental & Clinical Pharmacology; 2012: Wiley-Blackwell CommerceP, MA USA.
  55. 55. Aoyama S, Kase H, Borrelli E. Rescue of locomotor impairment in dopamine D2 receptor-deficient mice by an Adenosine A2A receptor antagonist. The Journal of Neuroscience: the official journal of the Society for Neuroscience. 2000;20(15):5848–52. 20/15/5848.
  56. 56. Vučcković MG, Li Q, Fisher B, Nacca A, Leahy RM, Walsh JP, et al. Exercise elevates dopamine D2 receptor in a mouse model of Parkinson's disease: in vivo imaging with [18F] fallypride. Movement Disorders. 2010;25(16):2777–84. pmid:20960487
  57. 57. Noble EP. D2 dopamine receptor gene in psychiatric and neurologic disorders and its phenotypes. American journal of medical genetics Part B, Neuropsychiatric Genetics: the official publication of the International Society of Psychiatric Genetics. 2003;116B(1):103–25. pmid:12497624
  58. 58. Hoffmann HM, Nadal R, Vignes M, Ortiz J. Chronic cocaine self-administration modulates ERK1/2 and CREB responses to dopamine receptor agonists in striatal slices. Addiction Biology. 2012;17(3):565–75. pmid:21812869
  59. 59. Delis F, Mitsacos A, Giompres P. Pharmacological characterization and anatomical distribution of the dopamine transporter in the mouse cerebellum. The Cerebellum. 2008;7(3):242–51. pmid:18418665
  60. 60. Kim YS, Shin JH, Hall FS, Linden DJ. Dopamine signaling is required for depolarization-induced slow current in cerebellar Purkinje cells. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 2009;29(26):8530–8. pmid:19571144
  61. 61. Maldonado R, Saiardi A, Valverde O, Samad TA, Roques BP, Borrelli E. Absence of opiate rewarding effects in mice lacking dopamine D2 receptors. Nature. 1997;388(6642):586–9. pmid:9252189
  62. 62. Wang KS, Zuo L, Pan Y, Xie C, Luo X. Genetic variants in the CPNE5 gene are associated with alcohol dependence and obesity in Caucasian populations. Journal of Psychiatric Research. 2015;71:1–7. pmid:26522866
  63. 63. Hart AB, Engelhardt BE, Wardle MC, Sokoloff G, Stephens M, de Wit H, et al. Genome-wide association study of d-Amphetamine response in healthy volunteers identifies putative associations, including cadherin 13 (CDH13). PloS One. 2012;7(8):e42646. pmid:22952603
  64. 64. Salatino-Oliveira A, Genro JP, Polanczyk G, Zeni C, Schmitz M, Kieling C, et al. Cadherin-13 gene is associated with hyperactive/impulsive symptoms in attention/deficit hyperactivity disorder. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics. 2015;168(3):162–9.
  65. 65. Joshi S, Guleria RS, Pan J, Bayless KJ, Davis GE, Dipette D, et al. Ethanol impairs Rho GTPase signaling and differentiation of cerebellar granule neurons in a rodent model of fetal alcohol syndrome. Cellular and Molecular Life Sciences CMLS. 2006;63(23):2859–70. pmid:17109064
  66. 66. Rothenfluh A, Threlkeld RJ, Bainton RJ, Tsai LTY, Lasek AW, Heberlein U. Distinct behavioral responses to ethanol are regulated by alternate RhoGAP18B isoforms. Cell. 2006;127(1):199–211. pmid:17018286
  67. 67. Tang L, Tian ZJ, Xiong ZY, Zhang YQ. Study of the genes expression of SCD-2 and B-FABP in the mice brain of exercise-induced fatigue by genechip cDNA microarray. Chinese Journal of Applied Physiology. 2005;21(2):137–9. pmid:21171323
  68. 68. El Hajj Chehadeh S, Dreumont N, Willekens J, Canabady-Rochelle L, Jeannesson E, Alberto JM, et al. Early methyl donor deficiency alters cAMP signaling pathway and neurosteroidogenesis in the cerebellum of female rat pups. American Journal of Physiology, Endocrinology and Metabolism. 2014;307(11):E1009–19. pmid:25294213
  69. 69. Agger K, Christensen J, Cloos PAC, Helin K. The emerging functions of histone demethylases. Current Opinion in Genetics & Development. 2008;18(2):159–68.
  70. 70. Kim EH, Kim IH, Lee MJ, Thach Nguyen C, Ha JA, Lee SC, et al. Anti-oxidative stress effect of red ginseng in the brain is mediated by peptidyl arginine deiminase type IV (PADI4) repression via estrogen receptor (ER) beta up-regulation. Journal of Ethnopharmacology. 2013;148(2):474–85. pmid:23665163
  71. 71. Kovacic P. Role of oxidative metabolites of cocaine in toxicity and addiction: oxidative stress and electron transfer. Medical Hypotheses. 2005;64(2):350–6. S0306-9877(04)00439-6. pmid:15607570
  72. 72. Ito M, Nisimaru N. Cerebellar control of defense reactions under Orexin-mediated neuromodulation as a model of cerebellohypothalamic interaction. AIMS Neuroscience. 2014;1(1):89–95.
  73. 73. Valverde O, Maldonado R, Valjent E, Zimmer AM, Zimmer A. Cannabinoid withdrawal syndrome is reduced in pre-proenkephalin knock-out mice. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 2000;20(24):9284–9. 20/24/9284.
  74. 74. Neveu I, Remy S, Naveilhan P. The neuropeptide Y receptors, Y1 and Y2, are transiently and differentially expressed in the developing cerebellum. Neuroscience. 2002;113(4):767–77. pmid:12182884
  75. 75. Ruegsegger GN, Speichinger KR, Manier JB, Younger KM, Childs TE, Booth FW. Hypothalamic Npy mRNA is correlated with increased wheel running and decreased body fat in calorie-restricted rats. Neuroscience Letters. 2016;618:83–8. pmid:26921453
  76. 76. Zeyda T, Diehl N, Paylor R, Brennan MB, Hochgeschwender U. Impairment in motor learning of somatostatin null mutant mice. Brain Research. 2001;906(1):107–14.
  77. 77. Zhang H, Kranzler HR, Yang BZ, Luo X, Gelernter J. The OPRD1 and OPRK1 loci in alcohol or drug dependence: OPRD1 variation modulates substance dependence risk. Molecular Psychiatry. 2008;13(5):531–43. pmid:17622222
  78. 78. Negishi T, Takahashi M, Matsunaga Y, Hirano S, Tashiro T. Diphenylarsinic acid increased the synthesis and release of neuroactive and vasoactive peptides in rat cerebellar astrocytes. Journal of Neuropathology and Experimental Neurology. 2012;71(6):468–79. pmid:22588385
  79. 79. Sørensen G, Wörtwein G, Fink-Jensen A, Woldbye DPD. Neuropeptide Y Y5 receptor antagonism causes faster extinction and attenuates reinstatement in cocaine-induced place preference. Pharmacology Biochemistry and Behavior. 2013;105:151–6.
  80. 80. White CM, DeRidder A, Chadwick W, Martin B, Maudsley S. Chemical modification of class II G protein-coupled receptor ligands: frontiers in the development of peptide analogs as neuroendocrine pharmacological therapies. Pharmacology & Therapeutics. 2010;125(1):39–54.
  81. 81. Quignodon L, Grijota-Martinez C, Compe E, Guyot R, Allioli N, Laperrière D, et al. A combined approach identifies a limited number of new thyroid hormone target genes in post-natal mouse cerebellum. Journal of Molecular Endocrinology. 2007;39(1):17–28. pmid:17601882
  82. 82. Dale CS, Pagano RdL, Rioli V. Hemopressin: a novel bioactive peptide derived from the alpha1-chain of hemoglobin. Memórias do Instituto Oswaldo Cruz. 2005;100:105–6. pmid:15962106
  83. 83. Gelman JS, Fricker LD. Hemopressin and other bioactive peptides from cytosolic proteins: are these non-classical neuropeptides? The AAPS Journal. 2010;12(3):279–89. pmid:20383670
  84. 84. Rioli V, Gozzo FC, Heimann AS, Linardi A, Krieger JE, Shida CS, et al. Novel natural peptide substrates for endopeptidase 24.15, neurolysin, and angiotensin-converting enzyme. The Journal of biological chemistry. 2003;278(10):8547–55. pmid:12500972
  85. 85. Fernandez-Castillo N, Orejarena MJ, Ribases M, Blanco E, Casas M, Robledo P, et al. Active and passive MDMA ('ecstasy') intake induces differential transcriptional changes in the mouse brain. Genes, Brain, and Behavior. 2012;11(1):38–51. pmid:21951708
  86. 86. Singec I, Knoth R, Ditter M, Frotscher M, Volk B. Neurogranin expression by cerebellar neurons in rodents and non-human primates. Journal of Comparative Neurology. 2003;459(3):278–89. pmid:12655510
  87. 87. Watanabe D, Inokawa H, Hashimoto K, Suzuki N, Kano M, Shigemoto R, et al. Ablation of cerebellar Golgi cells disrupts synaptic integration involving GABA inhibition and NMDA receptor activation in motor coordination. Cell. 1998;95(1):17–27. pmid:9778244
  88. 88. Levran O, Peles E, Randesi M, Correa da Rosa J, Ott J, Rotrosen J, et al. Synaptic plasticity and signal transduction gene polymorphisms and vulnerability to drug addictions in populations of European or African ancestry. CNS Neuroscience & Therapeutics. 2015;21(11):898–904. pmid:26384852
  89. 89. Podder A, Jatana N, Latha N. Human Dopamine Receptors Interaction Network (DRIN): A systems biology perspective on topology, stability and functionality of the network. Journal of Theoretical Biology. 2014;357:169–83. pmid:24846730
  90. 90. Kim YP, Kim H, Shin MS, Chang HK, Jang MH, Shin MC, et al. Age-dependence of the effect of treadmill exercise on cell proliferation in the dentate gyrus of rats. Neuroscience Letters. 2004;355(1–2):152–4. S0304394003013089. pmid:14729257
  91. 91. Noonan MA, Bulin SE, Fuller DC, Eisch AJ. Reduction of adult hippocampal neurogenesis confers vulnerability in an animal model of cocaine addiction. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 2010;30(1):304–15. pmid:20053911
  92. 92. Gaiteri C, Sibille E. Differentially expressed genes in major depression reside on the periphery of resilient gene coexpression networks. Frontiers in Neuroscience. 2011;5:95. pmid:21922000
  93. 93. Wang J, Cui W, Wei J, Sun D, Gutala R, Gu J, et al. Genome-wide expression analysis reveals diverse effects of acute Nicotine exposure on neuronal function-related genes and pathways. Frontiers in Psychiatry. 2011;2:5. pmid:21556275
  94. 94. Garci-a-Fuster MJ, Ferrer-Alcan M, Miralles A, Garci-a-Sevilla Js. Modulation of Fas receptor proteins and dynamin during opiate addiction and induction of opiate withdrawal in rat brain. Naunyn-Schmiedeberg's Archives of Pharmacology. 2003;368(5):421–31. pmid:14530904
  95. 95. Konu Ö, Xu X, Ma JZ, Kane J, Wang J, Shi SJ, et al. Application of a customized pathway-focused microarray for gene expression profiling of cellular homeostasis upon exposure to nicotine in PC12 cells. Molecular Brain Research. 2004;121(1):102–13.
  96. 96. Liu J, Merkle FT, Gandhi AV, Gagnon JA, Woods IG, Chiu CN, et al. Evolutionarily conserved regulation of hypocretin neuron specification by Lhx9. Development (Cambridge, England). 2015;142(6):1113–24. pmid:25725064
  97. 97. Tsujino N, Sakurai T. Orexin/hypocretin: a neuropeptide at the interface of sleep, energy homeostasis, and reward system. Pharmacological Reviews. 2009;61(2):162–76. pmid:19549926
  98. 98. Bough KJ, Amur S, Lao G, Hemby SE, Tannu NS, Kampman KM, et al. Biomarkers for the development of new medications for cocaine dependence. Neuropsychopharmacology. 2014;39(1):202–19. pmid:23979119
  99. 99. Alliey-Rodriguez N, Zhang D, Badner JA, Lahey BB, Zhang X, Dinwiddie S, et al. Genome-wide association study of personality traits in bipolar patients. Psychiatric Genetics. 2011;21(4):190–4. pmid:21368711
  100. 100. Krezel W, Ghyselinck N, Samad TA, Dupe V, Kastner P, Borrelli E, et al. Impaired locomotion and dopamine signaling in retinoid receptor mutant mice. Science (New York, NY). 1998;279(5352):863–7.
  101. 101. Krzyzosiak A, Szyszka-Niagolov M, Wietrzych M, Gobaille S, Muramatsu S, Krezel W. Retinoid x receptor gamma control of affective behaviors involves dopaminergic signaling in mice. Neuron. 2010;66(6):908–20. pmid:20620876
  102. 102. Novak G, Fan T, O'Dowd BF, George SR. Striatal development involves a switch in gene expression networks, followed by a myelination event: implications for neuropsychiatric disease. Synapse. 2013;67(4):179–88. pmid:23184870
  103. 103. Samad TA, Krezel W, Chambon P, Borrelli E. Regulation of dopaminergic pathways by retinoids: activation of the D2 receptor promoter by members of the retinoic acid receptor-retinoid X receptor family. Proceedings of the National Academy of Sciences of the United States of America. 1997;94(26):14349–54. pmid:9405615
  104. 104. Dierssen M, Fedrizzi L, Gomez-Villafuertes R, de Lagran MM, Gutierrez-Adan A, Sahun I, et al. Reduced Mid1 expression and delayed neuromotor development in daDREAM transgenic mice. Frontiers in Molecular Neuroscience. 2012;5:58. pmid:22563308
  105. 105. Tobon KE, Catuzzi JE, Cote SR, Sonaike A, Kuzhikandathil EV. Post-transcriptional regulation of dopamine D1 receptor expression in caudate-putamen of Cocaine-sensitized mice. The European Journal of Neuroscience. 2015;42(2):1849–57. pmid:25900179
  106. 106. Couceyro P, Pollock KM, Drews K, Douglass J. Cocaine differentially regulates activator protein-1 mRNA levels and DNA-binding complexes in the rat striatum and cerebellum. Molecular Pharmacology. 1994;46(4):667–76. pmid:7969045
  107. 107. Beninger RJ, Hoffman DC, Mazurski EJ. Receptor subtype-specific dopaminergic agents and conditioned behavior. Neuroscience and Biobehavioral Reviews. 1989;13(2–3):113–22. S0149-7634(89)80019-3. pmid:2573022
  108. 108. Maldonado R, Robledo P, Chover AJ, Caine SB, Koob GF. D1 dopamine receptors in the nucleus accumbens modulate cocaine self-administration in the rat. Pharmacology, Biochemistry, and Behavior. 1993;45(1):239–42. 0091-3057(93)90112-7. pmid:8516365
  109. 109. Phillips GD, Robbins TW, Everitt BJ. Bilateral intra-accumbens self-administration of d-Amphetamine: antagonism with intra-accumbens SCH-23390 and sulpiride. Psychopharmacology. 1994;114(3):477–85. pmid:7855206
  110. 110. Self DW, Barnhart WJ, Lehman DA, Nestler EJ. Opposite modulation of cocaine-seeking behavior by D1- and D2-like dopamine receptor agonists. Science (New York, NY). 1996;271(5255):1586–9.
  111. 111. Franklin KB, Vaccarino FJ. Differential effects of amphetamine isomers on SN self-stimulation: evidence for DA neuron subtypes. Pharmacology, Biochemistry, and Behavior. 1983;18(5):747–51. pmid:6856648
  112. 112. Kornetsky C, Esposito RU. Reward and detection thresholds for brain stimulation: dissociative effects of cocaine. Brain Research. 1981;209(2):496–500. 0006-8993(81)90177-3. pmid:6971689
  113. 113. Scarr E, Sundram S, Deljo A, Cowie TF, Gibbons AS, Juzva S, et al. Muscarinic M1 receptor sequence: preliminary studies on its effects on cognition and expression. Schizophrenia Research. 2012;138(1):94–8. pmid:22391213
  114. 114. Nathanson NM. Molecular properties of the muscarinic acetylcholine receptor. Annual Review of Neuroscience. 1987;10:195–236. pmid:2436543
  115. 115. Wess J. Molecular biology of muscarinic acetylcholine receptors. Critical Reviews in Neurobiology. 1996;10(1):69–99. pmid:8853955
  116. 116. Chen J, Alberts I, Li X. Dysregulation of the IGF-I/PI3K/AKT/mTOR signaling pathway in autism spectrum disorders. International Journal of Developmental Neuroscience. 2014;35:35–41. pmid:24662006
  117. 117. Gong H, Dong W, Rostad SW, Marcovina SM, Albers JJ, Brunzell JD, et al. Lipoprotein lipase (LPL) is associated with neurite pathology and its levels are markedly reduced in the dentate gyrus of Alzheimer's disease brains. The Journal of Histochemistry and Cytochemistry: official journal of the Histochemistry Society. 2013;61(12):857–68. pmid:24004859
  118. 118. Heck A, Fastenrath M, Ackermann S, Auschra B, Bickel H, Coynel D, et al. Converging genetic and functional brain imaging evidence links neuronal excitability to working memory, psychiatric disease, and brain activity. Neuron. 2014;81(5):1203–13. pmid:24529980
  119. 119. Brown RR, Ozaki Y, Datta SP, Borden EC, Sondel PM, Malone DG. Implications of interferon-induced tryptophan catabolism in cancer, auto-immune diseases and AIDS. Advances in Experimental Medicine and Biology. 1991;294:425–35. pmid:1722946
  120. 120. Ivlieva N. Involvement of the mesocorticolimbic dopaminergic system in adaptive behavior. Neuroscience and Behavioral Physiology. 2011;41(7):715–29.
  121. 121. Koob GF. Drugs of abuse: anatomy, pharmacology and function of reward pathways. Trends in Pharmacological Sciences. 1992;13(5):177–84. pmid:1604710
  122. 122. Harber VJ, Sutton JR. Endorphins and exercise. Sports Medicine (Auckland, NZ). 1984;1(2):154–71.
  123. 123. Leuenberger A. Endorphins, exercise, and addictions: A review of exercise dependence. impulse: The Premier Journal for Undergraduate Publications in the Neurosciences. 2006.
  124. 124. Comings DE, Rosenthal RJ, Lesieur HR, Rugle LJ, Muhleman D, Chiu C, et al. A study of the dopamine D2 receptor gene in pathological gambling. Pharmacogenetics. 1996;6(3):223–34. pmid:8807661
  125. 125. Lim S, Ha J, Choi SW, Kang SG, Shin YC. Association study on pathological gambling and polymorphisms of dopamine D1, D2, D3, and D4 receptor genes in a Korean population. Journal of Gambling Studies / co-sponsored by the National Council on Problem Gambling and Institute for the Study of Gambling and Commercial Gaming. 2012;28(3):481–91. pmid:21853233
  126. 126. Bowman EM, Aigner TG, Richmond BJ. Neural signals in the monkey ventral striatum related to motivation for juice and cocaine rewards. Journal of Neurophysiology. 1996;75(3):1061–73. pmid:8867118
  127. 127. Carelli RM, King VC, Hampson RE, Deadwyler SA. Firing patterns of nucleus accumbens neurons during cocaine self-administration in rats. Brain Research. 1993;626(1–2):14–22. 0006-8993(93)90557-4. pmid:8281424
  128. 128. Chang JY, Sawyer SF, Lee RS, Woodward DJ. Electrophysiological and pharmacological evidence for the role of the nucleus accumbens in cocaine self-administration in freely moving rats. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 1994;14(3 Pt 1):1224–44.
  129. 129. Peoples LL, Gee F, Bibi R, West MO. Phasic firing time locked to cocaine self-infusion and locomotion: dissociable firing patterns of single nucleus accumbens neurons in the rat. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience. 1998;18(18):7588–98.
  130. 130. American Psychiatric A. Diagnostic and statistical manual of mental disorders. Arlington, VA: American Psychiatric Publishing; 2013.
  131. 131. Sussman S, Lisha N, Griffiths M. Prevalence of the addictions: a problem of the majority or the minority? Evaluation & the Health Professions. 2011;34(1):3–56. pmid:20876085