Rationale In classical conditioning, sign-tracking reflects behavior directed toward a conditioned stimulus (CS) in expectation of a reward (unconditioned stimulus, US); in contrast, goal-tracking describes behavior directed toward the location of delivery of a US. As cues previously paired with drugs of abuse promote drug-seeking and drug-taking behavior in both animals and humans and thus contribute to the severity of substance abuse, sign-tracking may represent a maladaptive cue-focused behavior that may increase addiction vulnerability as compared to goal-tracking. Recent studies do, in fact, support this possibility. Previous work in this area has focused primarily on paradigms using relatively limited exposure to drug rather than extended drug intake. Objectives Here, we used the DSM-IV–based 3-criteria (3-CRIT) model and examined whether a relationship exists between sign- or goal-tracking phenotypes and the prevalence of criteria associated with addiction-like behavior following extended cocaine self-administration as measured in this model. Methods Forty-six male Sprague Dawley rats underwent a Pavlovian conditioned approach (PCA) procedure and were characterized along a continuum as goal-trackers (GTs), intermediates (INTs), or sign-trackers (STs). The animals were subsequently trained to intravenous self-administer cocaine during 45 self-administration (SA) sessions and characterized for the 3 criteria outlined in the model: persistence of drug-seeking, motivation for cocaine-taking, and resistance to punishment. Results We performed correlational analyses on the traits measured, finding no relationships between PCA score and addiction-like characteristics measured using the 3-CRIT model of addiction. However, STs showed significantly greater resistance to punishment than GTs. Conclusions Phenotyping along a continuum of PCA scores may not be a valid predictor for identifying vulnerability to the addiction-like behaviors examined using the 3-CRIT model. However, PCA phenotype may predict a single feature of the 3-CRIT model, resistance to punishment, among those rats classified as either STs or GTs.

Ras/Raf/MEK/ERK (Ras-ERK) signaling has been demonstrated to play a role in the effects of drugs of abuse such as cocaine and alcohol, but has not been extensively examined in nicotine-related reward behaviors. We examined the role of Ras Guanine Nucleotide Releasing Factor 2 (RasGRF2), an upstream mediator of the Ras-ERK signaling pathway, on nicotine self-administration (SA) in RasGRF2 KO and WT mice. We first demonstrated that acute nicotine exposure (0.4 mg/kg) resulted in an increase in phosphorylated ERK1/2 (pERK1/2) in the striatum, consistent with previous reports. We also demonstrated that increases in pERK1/2 resulting from acute (0.4 mg/kg) and repeated (0.4 mg/kg, 10 daily injections) exposure to nicotine in WT mice were not present in RasGRF2 KO mice, confirming that RasGRF2 at least partly regulates the activity of the Ras-ERK signaling pathway following nicotine exposure. We then performed intravenous nicotine SA (0.03 mg/kg/infusion for 10 days) in RasGRF2 KO and WT mice. Consistent with a previous report using cocaine SA, RasGRF2 KO mice demonstrated an increase in nicotine SA relative to WT controls. These findings suggest a role for RasGRF2 in the reinforcing effects of nicotine, and implicate the Ras-ERK signaling pathway as a common mediator of the response to drugs of abuse.