Drug dependence—or addiction—is often described as a chronic mental disorder characterized by compulsive drug intake and loss of control over intake despite adverse consequences [1,2,3,4]. Moreover, the fact that vulnerability to relapse in addicts can persist after many years of abstinence implies that addiction is caused by long-lasting changes in brain function as a result of pharmacological insult (repeated remission and relapse episodes), genetic predisposition, and environmental associations made with drug use (learning). Thus, a primary problem for the treatment of drug abuse remains to be the understanding of the brain neuroadaptations following repeated drug use that might be responsible for the persistent vulnerability to drug craving and seeking after abstinence [5,6].

In humans, the average period of time to relapse to opiate use is 25 days, while the rate of relapse is 71% within the first 6 weeks and 95% within 3 months [7]. The high rate of relapse to opiate abuse in this time frame is closely related to prolonged vulnerability to craving, sleep disturbances, dysphoria and anhedonia, a cluster of symptoms called the protracted withdrawal syndrome [8,9]. General circadian activity, a correlate of arousal and sleep periods in humans, has been studied in animal models of opiate dependence. It has been shown that similar to humans [8,10], abrupt drug cessation can alter the circadian biphasic activity for several days during abstinence in animal models of opiate dependence [11,12,13].

A large body of evidence suggests that the consequences of opiate withdrawal and conditioned withdrawal support both maintenance of drug taking and relapse to opiate seeking [14,15,16,17,18]. On one hand, chronically morphine-pretreated abstinent rats express stronger and longer lasting preferences for morphine-associated environments than placebo-pretreated rats (at least 5 weeks, [19]), develop a progressive incubation of drug seeking behavior over a two-month withdrawal period [20], and a long-lasting hypersensitivity to the psychostimulant effect of an opiate injection [21]. On the other hand, while it has been very difficult to demonstrate the important role of drug withdrawal in the maintenance of, and relapse to drug use [17,22], recent work successfully reported that morphine-withdrawal conditioned cues significantly reinstate drug seeking [23,24]. Likewise, there is a persistence of withdrawal-induced negative motivational affects such as anxiety [25] and aversion up to 16 weeks in abstinent rats [26]. Altogether, these findings highlight how long lasting the vulnerability to opiates is in abstinent rats.

It is common for opiate addicts to experience multiple episodes of compulsive drug taking interspersed by periods of attempted abstinence. Similar to humans, opiate dependence in rodents is a progressive phenomenon that may begin with a single dosing [27], but its severity augments with the increasing number of exposures to the drug [27,28]. For instance, the role of a repeated withdrawal syndrome has been shown to be critical in the escalation of volitional drug intake, including nicotine intake [29]. In addition, morphine dependence provided by a constant infusion or implanted pellets does not activate the hypothalamic-pituitary-adrenal (HPA) axis [30,31], while repeated exposures with escalating doses of opiate alter the stress system responses, a key element in opiate addiction [32,33]. Thus, these data support the hypothesis that repeated cycles of addiction contribute to the development of neuroadaptations involved in the vulnerability to drug addiction.

Thus, our hypothesis was that a robust and repeated exposure to opiates would have more pronounced behavioral alterations during long-term abstinence. The aim of the present study was to show that the diverse drug regimens to induce opiate dependence (morphine pellets, morphine injections or heroin injections) might produce differential locomotor activity responses to a heroin challenge during abstinence. Second, we examined the effect of multiple cycles of heroin dependence on the alterations in circadian locomotor activity during both spontaneous withdrawal and protracted abstinence.

A critical problem in drug addiction is that a majority of addicts return to compulsive drug taking long after acute withdrawal and abstinence. This protracted abstinence state is associated with persistent behavioral and neurobiological abnormalities responsible for drug seeking reinstatement. In the present study, we first show that intermittent episodes of opiate injections rather than constant exposure with morphine pellets lead to longer lasting hyperactivity to heroin challenges (up to 18 weeks). Moreover, the number of dependence episodes is an important factor for the behavioral adaptations. Indeed, post-dependent rats with two episodes of heroin dependence have a disruption of the circadian activity for the first 7 days of spontaneous withdrawal and show abnormal circadian activity patterns up to 18 weeks of abstinence.

Disruption of the circadian activity is a hallmark of the early opiate withdrawal syndrome in opiate-dependent rats [11,12]. Here, we show that a similar pattern of disrupted circadian activity developed regardless of the opiate (morphine or heroin) and the route of administration (pellet vs. injections). It could be suggested that novelty-induced stress by the transfer of the rats into the activity cages may contribute to the cycle disruption. This is unlikely since the present study confirmed our previous work in which rats were habituated to the behavioral paradigm before the induction of opiate dependence [11,12].

Heroin is transformed into morphine once it crosses the blood-brain barrier and therefore produces the same central effects. However, it has been reported that heroin has a more rapid onset of action and is more potent than morphine in producing reinforcing effects in animals [34,35,36]. In the model of intracranial self-stimulation (ICSS), a predictive model for drug abuse liability in rats, it was demonstrated that heroin is 40 times more potent than morphine in lowering the reward threshold [35]. After a rapid deacetylation, heroin is also metabolized into 6-acetylmorphine (6-MAM), a metabolite equipotent to heroin in the ICSS model [35]. In line with these findings, our results indicated that, maybe due to its metabolization into 6-MAM as well as into morphine, heroin rather than morphine induces more persistent changes to subsequent opiate challenges during protracted abstinence. Indeed, hyperactivity to heroin challenges lasted for 6 and 18 weeks, respectively.

Interestingly, there has been substantial preclinical support for the fact that a single exposure to a drug of abuse could lead to long-lasting cellular neuroadaptations [37] as well as to long-lasting vulnerability to drug seeking behavior [38]. However, these “acute” addiction models do not completely mimic addiction in humans, which generally go through several relapses and remissions episodes [39]. Our results highlight the fact that numerous rather than a single episode of drug dependence augments the intensity of the long-lasting behavioral modifications [21]. The 1Dep group was subjected to heroin dependence and then to saline injections. It cannot be fully discarded that 1Dep animals treated with saline first and then heroin injections could have behaved such as 2Dep rats. However, the model of heroin dependence presented here is of importance for chronic treatment and presents an ideal model for longitudinal studies of the behavioral and neurobiological components associated with extended abstinence in rats.

Previous studies have shown that chronic opiate administration with continued drug access leads to increased wakefulness during the sleep phase of the sleep/wake cycle in rats [13,40] as well as changes in circadian food intake [40]. Interestingly, the extended drug access also correlates with an escalation in drug intake [40]. Thus, it has been suggested that a dysregulation of circadian behaviors could contribute to the transition to opiate addiction. In our study, we show for the first time that protracted abstinence to opiate may also induce a long-lasting dysregulation of the circadian activity in rats. Sleep disruption and diurnal variation of the craving state profoundly affect protracted abstinent users [8,41]. Several neurobiological substrates have been suggested to contribute to circadian alterations after drug cessation. Recently, light has been shed on orexin, which is important in arousal, sleep as well as drug addiction [42,43]. In addition, this neuropeptide shows molecular adaptations to morphine and morphine withdrawal, and orexin knockout mice develop attenuated opiate dependence [43]. Thus, a possible hypothesis could be that persistent alterations of the circadian activity in our model of protracted heroin abstinence might be a consequence of long-lasting dysfunction of the orexin system.

Finally, the last experiment showed that the disruption of the diurnal activity in the 2Dep animals was greater at 16 and 20 weeks, suggesting an incubation-like phenomenon in the intensity of activity alterations. The term incubation was first used in both addicts and rat models of addiction to describe the observations that time-dependent increases in drug-seeking occurred after withdrawal from the drug [44,45]. The present study shows a rat model of heroin exposure that can be used to trigger incubation of circadian alterations mimicking such observations seeing in abstinent heroin users [8].

In summary, while there are an increasing number of studies dedicated to the understanding of protracted abstinence, it remains unclear why it is difficult for individuals who have been previously dependent to maintain long-term abstinence [47]. Thus, the animal model of heroin dependence we have developed here provides an excellent research tool to study the underlying neurobiology of vulnerability to drug seeking behavior.