Identification of molecular signatures in the dopaminergic system associated with administration of various drugs of abuse in a mouse model

Andrew Jenkins1, Pierrette de Ron2, Sarah Dremier2, Ludovicus Staelens2, Charlotte Muzzi2, André da Costa2, Virgile Richard2. 1University of Bath, Bath, UK, 2UCB Pharma SA, Braine l'Alleud, Belgium

Preclinical abuse liability assessment of new CNS drugs is cumbersome and only performed in late phase development. It relies on classical behavioural pharmacology models which are notorious for conceptual and interpretational difficulties and whose utility for detecting abuse potential in novel centrally active drugs is unknown1. This study aimed to identify whether the administration of numerous psychotropes is associated with a common molecular signature in discrete regions of the mesocorticolimbic dopamine system, which is positively correlated with behavioural phenotypes indicative of a potential for abuse.

Male C57Bl/6J mice (20-26g) underwent conditioning in the conditioned place preference paradigm2 with either cocaine (20 mg/kg), modafinil (65 mg/kg), methylphenidate (MPH) (5 mg/kg), γ-hydroxybutyric acid (GHB) (150 mg/kg before stepping up to 175 mg/kg) or vehicle (all injections 10 ml/kg i.p.). Mice experienced four exposures to the drug, except GHB which was given in 7 exposures. All drugs were dissolved in 0.9% physiological saline, except modafinil which was suspended in 1% methylcellulose-0.1% Tween 80-0.1% silicone antifoam. After conditioning, the mice were sacrificed by cervical dislocation and the ventral tegmental area (VTA), nucleus accumbens and orbitofrontal cortex were dissected out, and analysis of both RNA and miRNA expression in these regions was performed using Affymetrix® microarrays.

An increase in preference for the drug-paired environment was seen in all of the drug-treated groups, which was significantly greater than in the vehicle-treated animals (Kruskall-Wallis p<0.0001; post-hoc Dunn’s multiple comparisons: cocaine p=0.0008, modafinil p=0.005, MPH p<0.0001, GHB p=0.02).

Transcriptomic analyses revealed the VTA to be the region of most interest, where 37 genes were found to be commonly differentially expressed in all four treatment groups at a threshold fold change value of 1.5 versus vehicle controls (p<0.05). Genes affected are known to be implicated in neurological, psychological and developmental disorders. Many are important for normal nervous system development. Together, the results suggest a resetting of the dopaminergic pathway to a state occupied during critical phases of development.

In summary, several potential novel molecular biomarkers were identified for use as mechanistic endpoints to support molecular studies in pre-clinical models of abuse liability. Our findings provide support at the molecular level for recent studies of drug-evoked synaptic plasticity3, and provide an insight into the lasting systemic re-wiring which occurs during acute drug administration and which is thought to be prerequisite to the neuroadaptations associated with the development of addiction4.

1. Moser P et al, J Pharmacol Toxicol 63: 160-167, 2011

2. Craige C & Unterwald E, Behav Brain Res 238: 206-210, 2013

3. Bellone C & Lüscher C, Front Mol Neurosci 5: 75, 2012

4. Lüscher C & Malenka RC, Neuron 69: 650-663, 2011