In this study we sought to determine if repeated exposure to experimenter-delivered cocaine increases rats' tendency to seek out rewards when presented with a reward-paired cue and/or biases their tendency to acquire a habitual response selection strategy when pursuing a food reward. Our results support both hypotheses, demonstrating that repeated experimenter-delivered cocaine can 1) facilitate the expression of PIT for food reward, a relatively pure measure of cue-evoked incentive motivation, and 2) bias rats towards using a habitual control strategy when pursuing food reward. Furthermore, by using a within-subjects design to examine these behavioral phenomena, the current results indicate that the observed alterations in motivation and action selection can result from the same cocaine exposure regimen, and are therefore not likely to be particularly parameter dependent. While the additional time (5–7 d) between the PIT and outcome devaluation tests does not allow us to make a definitive statement about the dependence of these effects on the drug exposure-to-test interval, the current results do allow us to conclude that the cocaine exposure regimen used here results in an alteration of reward-seeking behavior that persists across a window of 15–22 d after the cessation of drug exposure.
Our findings are in line with previous reports that repeated peripheral administration of amphetamine facilitates expression of PIT 
, demonstrating that both psychostimulants have similar effects on cue-induced motivation despite differences in their mechanisms of action. This might be expected given demonstrations of behavioral cross-sensitization between these psychostimulants 
. Thus, despite their differential effects and modes of action on the dopamine system 
, these drugs appear to share the ability to produce a nonspecific enhancement in cue-evoked incentive motivation. A more recent study 
found that rats trained to self-administer intravenous cocaine exhibited an enhancement in PIT. Our results demonstrate that such an effect can also be produced with other modes of drug delivery, which supports the validity of using experimenter-delivered cocaine administration to model this consequence of cocaine taking. However, given recent findings that passive and self-regulated cocaine intake have different neurochemical effects 
and support distinct adaptations in the circuitry controlling dopamine signaling 
, further comparison of the effects of these treatments seems warranted.
In this experiment we have used the PIT task to assay cue-elicited incentive motivation. In this task, Pavlovian cues and instrumental actions are trained during separate phases of the experiment, preventing these events from becoming directly associated with one another. Because of the design of this task, any increase in responding occurring during the reward-paired cue at test must be due to its incentive motivational properties, acquired during the Pavlovian conditioning phase. Furthermore, because the reward-paired cue is noncontingently presented at test, its behavioral effects represent the elicitation or invigoration of reward seeking behavior, as opposed to a conditioned reinforcement effect. Using a PIT design much like that used here, we have recently shown that rats trained to self-administer cocaine increase their pursuit of cocaine when presented with a cocaine-paired cue 
. Although this finding demonstrates that Pavlovian incentive motivational processes contribute to instrumental drug motivated behavior, the role of cocaine sensitization
in this phenomenon is difficult to isolate from that drug's more direct role as a behavioral goal or reinforcer. Because the subject is repeatedly exposed to the drug during instrumental and Pavlovian conditioning sessions, it is not clear whether or to what degree the cue's effects on drug seeking have been altered by drug-induced sensitization. As discussed earlier, the nonspecific effects of repeated drug exposure on motivation are easier to evaluate when the target response is reinforced by a different reward, like palatable food, because in this case the drug exposure treatment is not conflated with its function as a reinforcer. The current findings add to a growing literature showing that repeated drug administration can enhance appetitive behaviors generated by non-drug rewards. 
. Such findings suggest that extended drug exposure produces nonspecific alterations in motivation, perhaps via adaptations in the dopamine system. Indeed, while it is firmly established that repeated drug exposure can sensitize the dopamine response to future drug challenges, there is also growing evidence that such treatment generates broad cross-sensitization of dopamine release to other drugs 
, as well as to chemical stimulation 
and natural reward stimuli and associated cues 
. Linking these changes in dopamine signaling to alterations in incentive motivation will require further research.
We found further evidence of the motivational impact of cocaine treatment during the sensitization phase of experiment 1, in that repeated cocaine treatments elicited an increase in magazine approach behavior. Importantly, this effect was not simply due to an increase in locomotor behavior; no such effect was observed in rats that were unfamiliar with the behavioral chamber (for whom the magazine approach response was most likely to be a simple exploratory behavior), even though these rats did show evidence of sensitization of their general locomotor activity. Thus, the tendency for repeated cocaine exposure to enhance magazine approach behavior depends on whether or not that activity is motivated by food reward. Although this behavioral measure does not provide the kind of pure assessment of cue-evoked incentive motivation that PIT provides, this finding does bolster the view that repeated cocaine exposure sensitizes motivated behavior.
Our results also provide evidence that repeated cocaine exposure promotes use of a habitual, rather than a goal-directed, response strategy, which follows with the suggestion of others that drug-induced sensitization facilitates acquisition of habitual control 
. This enhancement in S-R learning should leave drug-seeking behavior less sensitive to its various negative consequences or to the desire to abstain. It is possible that repeated exposure to cocaine and other drugs could impact the action selection strategies more generally. Our results suggest that cocaine can support such an effect, demonstrating that repeated cocaine administration enhances habitual control during the pursuit of natural rewards. Similar findings have been obtained by pretreating rats with amphetamine 
, indicating that these psychostimulants also have in common the ability to bias action selection towards habitual control.
It has been argued that habitual control, which is typically tested in extinction (without negative feedback), does not adequately model the compulsive nature of drug-seeking behavior, which tends to persist despite its negative consequences 
. However, in the current study, cocaine-treated rats also appeared to have some difficulty suppressing their instrumental performance even when given response-contingent negative feedback (i.e., the opportunity to consume the devalued reward). It should be noted, however, that we did not detect a significant group difference in the sensitivity of lever pressing to outcome devaluation during this rewarded test phase, and so it is not possible to draw strong conclusions about whether or not the cocaine group's performance should be considered abnormal. Other findings suggest extensive or chronic access to cocaine is required for the development of compulsive forms of cocaine seeking in rodents 
, including behavior that is relatively insensitive to negative feedback (response extinction, electric footshock or shock-paired cues) 
. We might therefore expect an even clearer disruption of outcome devaluation sensitivity during these tests (particularly the rewarded test) in rats given more frequent cocaine exposures or larger doses. However, since these other studies used response-contingent intravenous cocaine exposure, the mode of cocaine delivery may also be an important factor determining the long-term impact of cocaine on behavioral control. Future studies should investigate these possibilities.
It should be noted that, in the current study, rats were placed in the chamber used for behavioral training and testing after each daily cocaine (or saline) treatment. Although we did not explore this issue here, it seems likely that drug-induced context conditioning contributed to the behavioral effects that we observed in cocaine-treated rats. For instance, context-mediated learning has been shown to play an important role in other forms of behavioral sensitization 
. Furthermore, a recent study found that cocaine-paired contextual stimuli can provoke impulsive decision making in rats 
. We have also recently shown that contextual cues paired with alcohol intoxication produce a transient disruption in instrumental control, causing rats to shift from goal-directed to habitual performance 
. Thus, the impact of repeated cocaine exposure on PIT and outcome devaluation performance may be mediated, at least in part, by context-cocaine conditioning. Future studies should explore this possibility more directly.
Determining how aberrations in habitual control and incentive motivation work together to generate compulsive drug seeking is an important goal for future research. One interesting possibility is that these processes make distinct, stage-dependent contributions to the development of addiction. They may also affect different components of drug-related behavior. For instance, it has been argued that while exaggerated habits may contribute to drug-taking or consumption, it is the sensitization of incentive motivation that maintains compulsive drug-seeking and disrupts attempts to abstain 
. Conversely, studies have shown that compulsive drug-seeking behavior can be induced independently of any increased motivation for the drug, by producing habitual control of behavior 
, leading some to speculate that it is in fact increased incentive motivation for the drug that develops first, but it is the habitual, S-R action selection strategy that facilitates compulsive drug-seeking during the later stages of addiction 
. Interestingly, basic behavioral research has shown that reward-paired cues tend to facilitate performance by engaging habits 
, suggesting that these two processes work in tandem to control behavior. Our data show that, within a single set of rats, cocaine administration can sensitize both cue-evoked incentive motivation and habit formation, which is clearly compatible with this view. Finally, while this work suggests that drug-induced aberrations in motivation and behavioral control may contribute to addiction, the complex characteristics of this condition would seem to suggest that other cognitive and behavioral dysfunctions, such as alterations in prefrontal cortical areas and executive control, also play an important role.