Eur J Neurosci. 2019 Apr 9. doi: 10.1111/ejn.14418.
Kawa AB1, Valenta AC2, Kennedy RT2, Robinson TE1.
Abstract
The temporal pattern of drug use (pharmacokinetics) has a profound effect on the ability of self-administered cocaine to produce addiction-like behavior in rodents, and to change the brain. To further address this issue, we compared the effects of Long Access (LgA) cocaine self-administration, which is widely used to model the transition to addiction, with Intermittent Access (IntA), which is thought to better reflect the pattern of drug use in humans, on the ability of a single, self-administered injection of cocaine to increase dopamine (DA) overflow in the core of the nucleus accumbens (using in vivo microdialysis), and to produce addiction-like behavior. IntA experience was more effective than LgA in producing addiction-like behavior – a drug experience-dependent increase in motivation for cocaine assessed using behavioral economic procedures, and cue-induced reinstatement – despite much less total drug consumption. There were no group differences in basal levels of DA in dialysate [DA], but a single self-administered IV injection of cocaine increased [DA] in the core of the nucleus accumbens to a greater extent in rats with prior IntA experience than those with LgA or Limited Access (LimA) experience, and the latter two groups did not differ. Furthermore, high motivation for cocaine was associated with a high [DA] response. Thus, IntA, but not LgA, produced both incentive and DA sensitization. This is consistent with the notion that a hyper-responsive dopaminergic system may contribute to the transition from casual patterns of drug use to the problematic patterns that define addiction. This article is protected by copyright. All rights reserved.
KEYWORDS: addiction; cocaine; dopamine; intermittent access; sensitization
PMID: 30968487
DOI: 10.1111/ejn.14418
DISCUSSION
The purpose of this study was to compare the ability of prolonged LgA and IntA cocaine self-administration experience to produce addiction-like behavior (relative to ShA), and how this influenced the ability of self-administered cocaine to change extracellular DA levels in the nucleus accumbens core in vivo. The main findings were: 1. As expected, LgA resulted in much greater total cocaine consumption than IntA. 2. Both IntA and LgA produced escalation of intake with increasing self-administration experience. 3. IntA (but not LgA) experience increased motivation for cocaine, as indicated by a decrease in α and increase in PMax. 4. IntA rats showed greater cue-induced reinstatement of cocaine-seeking than LgA rats. 5. LgA (but not IntA) experience increased the preferred level of cocaine intake when no effort was required (Q0). 6. There were no group differences in the basal levels of DA in dialysate, but a single self administered IV injection of cocaine increased DA in the core of the nucleus accumbens to a greater extent in rats with prior IntA experience than those with LgA or ShA experience, and the latter two groups did not differ. 7. Across all groups high motivation for cocaine was associated with a greater DA response. 8. There were no group differences in dialysate concentrations of glutamate, GABA, ACh, DOPAC or HVA, although cocaine increased 3-MT to a greater extent in IntA than ShA or LgA rats, consistent with the effects on DA.IntA experience was more effective in producing addiction-like behavior than LgA experience
Since its introduction in 1998 (Ahmed & Koob, 1998) the LgA procedure has been widely adopted to model the transition to cocaine addiction in rats, because it was thought to be especially effective in producing a number of addiction-like behaviors, relative to ShA (for reviews see, Ahmed, 2012; Edwards and Koob 2013). In their 1998 paper Ahmed and Koob reported that LgA, but not ShA, resulted in escalation of intake. Since that time it has also been reported that, relative to ShA, rats with LgA experience are more motivated to seek cocaine (Paterson & Markou, 2003; Wee et al., 2008), take more cocaine in the face of adverse consequences (Xue et al., 2012; Bentzley et al., 2014; also see Vanderschuren & Everitt, 2004), and show greater reinstatement of cocaine-seeking behavior following extinction (Mantsch et al., 2004, 2008; Ahmed & Cador, 2006; Kippin et al., 2006). As indicated by the excerpt cited in the Introduction from Ahmed (2012), it has been suggested that the critical factor necessary for the emergence of escalation and other addiction-like behavior is the amount of drug consumed. As put by Edwards and Koob (2013), “excessive drug exposure likely remains an indispensable element driving the development of addiction”. However, the findings presented here add to a growing literature indicating that this is not the case.
IntA self-administration results in much less total cocaine consumption than LgA. Yet as reported here, IntA also produced escalation of intake and was more effective than LgA in increasing motivation for cocaine and in producing cue-induced reinstatement of cocaine-seeking behavior. These findings are consistent with a number of recent studies that also report IntA produces escalation of intake, heightened motivation for cocaine, continued cocaine-seeking in the face of an adverse consequence, continued cocaine-seeking when it is not available, and greater cue-induced reinstatement (Zimmer et al., 2012; Kawa et al., 2016; Allain & Samaha, 2018; Allain et al., 2018; James et al., 2018; Kawa & Robinson, 2018; Singer et al., 2018). Collectively, these studies have established that the consumption of the large amount of cocaine associated with LgA is not necessary for the development of addiction-like behavior and other pharmacokinetic factors appear to be more important (Allain et al., 2015). The failure of LgA experience to increase motivation for cocaine in the present study is inconsistent with several previous studies using either the same behavioral economic indicators
IntA experience was more effective in producing addiction-like behavior than LgAexperience
Since its introduction in 1998 (Ahmed & Koob, 1998) the LgA procedure has been widely adopted to model the transition to cocaine addiction in rats, because it was thought to be especially effective in producing a number of addiction-like behaviors, relative to ShA (for reviews see, Ahmed, 2012; Edwards and Koob 2013). In their 1998 paper Ahmed and Koob reported that LgA, but not ShA, resulted in escalation of intake. Since that time it has also been reported that, relative to ShA, rats with LgA experience are more motivated to seek cocaine (Paterson & Markou, 2003; Wee et al., 2008), take more cocaine in the face of adverse consequences (Xue et al., 2012; Bentzley et al., 2014; also see Vanderschuren & Everitt, 2004), and show greater reinstatement of cocaine-seeking behavior following extinction (Mantsch et al., 2004, 2008; Ahmed & Cador, 2006; Kippin et al., 2006). As indicated by the excerpt cited in the Introduction from Ahmed (2012), it has been suggested that the critical factor necessary for the emergence of escalation and other addiction-like behavior is the amount of drug consumed. As put by Edwards and Koob (2013), “excessive drug exposure likely remains an indispensable element driving the development of addiction”. However, the findings presented here add to a growing literature indicating that this is not the case.
IntA self-administration results in much less total cocaine consumption than LgA. Yet as reported here, IntA also produced escalation of intake and was more effective than LgA in increasing motivation for cocaine and in producing cue-induced reinstatement of cocaine-seeking behavior. These findings are consistent with a number of recent studies that also report IntA produces escalation of intake, heightened motivation for cocaine, continued cocaine-seeking in the face of an adverse consequence, continued cocaine-seeking when it is not available, and greater cue-induced reinstatement (Zimmer et al., 2012; Kawa et al., 2016; Allain & Samaha, 2018; Allain et al., 2018; James et al., 2018; Kawa & Robinson, 2018; Singer et al., 2018).
Collectively, these studies have established that the consumption of the large amount of cocaine associated with LgA is not necessary for the development of addiction-like behavior and other pharmacokinetic factors appear to be more important (Allain et al., 2015). The failure of LgA experience to increase motivation for cocaine in the present study is inconsistent with several previous studies using either the same behavioral economic indicators intake may escalate with both LgA and IntA, but for very different reasons – because of tolerance to cocaine’s desired effect in the case of LgA and incentive-sensitization in the case of IntA (Kawa et al., 2016; Kawa & Robinson, 2018). Of course, the idea that consummatory and motivational aspects of behavior are psychologically (and neurobiologically) dissociable has used here (Zimmer et al., 2012; Bentzley et al., 2014) or Progressive Ratio (PR) tests (Paterson & Markou, 2003; Wee et al., 2008). However, the effects of LgA reported in these studies were often only assessed at a single time point and compared to ShA, and did not involve within subject comparisons. In the studies that measured how motivation changed with increasing LgA experience (Bentzley et al., 2014) the effects were modest relative to the changes that occur following IntA. Our findings are consistent with other reports that LgA experience does not increase motivation for cocaine, as assessed with either behavioral economic metrics (Oleson & Roberts, 2009) or PR tests (Liu et al., 2005; Quadros & Miczek, 2009; Willuhn et al., 2014 supplementary). In addition, it has been reported that changes in motivation produced by LgA experience are very transient, lasting for only a few days after the last self-administration session (Bentzley et al., 2014; James et al., 2018), whereas the increased motivation produced by IntA experience is long-lasting – still evident after 50 days of abstinence (James et al., 2018). In summary, evidence that LgA increases motivation for cocaine is somewhat mixed, whereas IntA has been consistently reported to do so.
When allowed to self-administer cocaine under low Fixed Ratio (FR) schedules of reinforcement, rats generally titrate their responding to achieve a preferred brain concentration of cocaine, which they defend within a wide range of doses (Gerber & Wise, 1989; Ahmed & Koob, 1999; Lynch & Carroll, 2001). This preferred level of consumption was quantified here by the metric Q0 – the preferred level of consumption when cost is nil. Q0 presumably represents the brain level of cocaine that produces some optimal desired effect, such that neither more nor less cocaine is better. Some have referred to Q0 as a “hedonic set-point” (Bentzley et al., 2013), although “settling-point” may be more appropriate (see Berridge, 2004). Of course, it is not possible to know if Q0 actually reflects subjective hedonic effects in rodents. Nevertheless, LgA experience does increase the preferred level of cocaine consumption, as indicated by escalation of intake (Ahmed & Koob, 1998), and by an increase in Q0, as reported here and by others (Oleson & Roberts, 2009; Bentzley et al., 2014; James et al., 2018). The present results suggest, therefore, that LgA experience produces tolerance to whatever desired effect of cocaine is defended as price increases, without any change in motivation for cocaine. In contrast, IntA increases motivation for cocaine without any concomitent change in cocaine’s desired effects. Although highly speculative, this may reflect a dissociation between cocaine “wanting” and “liking” (Robinson & Berridge, 1993; Berridge & Robinson, 2016). It also suggests that cocaine been suggested frequently (e.g., Nicola & Deadwyler, 2000; Sharpe & Samson, 2001; Oleson et al., 2011; Guillem et al., 2014).
Neither LgA nor IntA experience altered basal dopamine
A decrease in basal DA levels has been reported when testing occurred soon after the discontinuation of high-dose and/or high-intake cocaine self-administration procedures (Mateo et al., 2005; Ferris et al., 2011). However, in the present study neither LgA nor IntA experience had any effect on basal DA in dialysate. Also, we included 13C6 dopamine in the aCSF, which allowed us to calculate an extraction fraction for each sample, and thus more accurately estimate basal DA. There were no group differences in the extraction fraction, thus bolstering our conclusion that neither LgA nor IntA changed basal DA (relative to ShA). This result is consistent with other reports that LgA experience does not alter baseline DA concentrations in dialysate, relative to ShA rats (Ahmed et al., 2003) or drug-naïve rats (Calipari et al., 2014). In addition, baseline DA levels did not correlate with any of our measures of addiction-like behavior, consistent with other studies (Hurd et al., 1989; Ahmed et al., 2003).
IntA, but not LgA, sensitizes cocaine-evoked dopamine overflow
There have been very few studies on the neurobiological consequences of IntA experience, and those available have all involved ex vivo measures. Most relevant to the present study are reports by Calipari et al. (2013, 2015) that IntA experience sensitizes stimulated DA release from the nucleus accumbens core in tissue slices, relative to naïve rats or rats with a history of ShA, and also increases the ability of cocaine to inhibit DA uptake. The main purpose of the present experiment was to determine if a similar sensitization of DA neurotransmission is present in awake, behaving rats. Following prolonged IntA experience a single self-administered cocaine infusion, given in the absence of the cocaine cue, produced a greater increase in extracellular DA in the accumbens core than following either LgA or ShA experience, and these latter two groups did not differ. Furthermore, the magnitude of the DA response predicted motivation for cocaine, as assessed by a number of measures, including PMax, α, and cocaine-seeking on the microdialysis test day. In addition, the DA response to cocaine was greatest in rats that met the most criteria for addiction. These findings establish that IntA, a cocaine self administration procedure that is especially effective in producing incentive-sensitization and addiction-like behavior, also sensitizes the dopaminergic response to cocaine. Finally, IntA also has been reported to be especially effective in producing a number of other neurobiological effects related to the development of addiction-like behavior, including dysregulation of mGluR2/3 receptor function (Allain et al., 2017), elevated BDNF levels (Gueye et al., 2018), and increased activity in orexin/hypocretin neurons (James et al., 2018).
In contrast with the dopaminergic sensitization produced by IntA experience, there are a number of reports that LgA does the opposite – decreases DA function, relative to ShA. For example, following LgA, or other high dose cocaine procedures, the ability of cocaine to inhibit DA uptake, or for electrical stimulation to evoke DA release from the accumbens core, is decreased in tissue slices, as is cocaine-evoked DA overflow measured with microdialysis in vivo (Ferris et al., 2011; Calipari et al., 2013, 2014; Siciliano et al., 2016). It may seem surprising, therefore, that in the present study a single, self-administered IV injection of cocaine increased DA to the same extent in rats with LgA or ShA experience – that is, there was no evidence for tolerance. It is not clear what accounts for the discrepancy – e.g., ex vivo vs. in vivo measures, experimenter-administered IP cocaine challenge vs. self-administered IV injection, measurement technique, or other methodological differences. However, the present results are consistent with one other study on the effects of LgA experience on DA measured with microdialysis in vivo. Ahmed (2003) reported that, relative to ShA, LgA did not decrease the DA response in the nucleus accumbens to either experimenter-administered IV injections of cocaine, or cocaine selfa dministration. Thus, it seems that LgA experience does not consistently decrease DA activity. It should also be noted that effects may vary considerably as a function of how long after the discontinuation of self-administration rats are tested (e.g., Ferrario et al., 2005; Siciliano et al., 2016). In addition, Willuhn et al. (2014) reported that the magnitude of a phasic DA response seen following a nosepoke that delivered cocaine progressively decreased with increasing LgA experience, as measured with fast scan cyclic voltammetry. However, this phasic DA response peaked approximately 5 sec after a nosepoke, which is too soon to reflect the pharmacological effects of cocaine (Stuber et al., 2005; Aragona et al., 2008), and therefore, may not be relevant to the studies discussed above.
One hour after the cocaine injection the cue that had been associated with cocaine was presented and we expected to see a conditioned DA response. But the cocaine cue had no effect in any group, on any neurochemical measure. It is not clear why this was the case, because the cue certainly had motivational properties, as indicated by the cue reinstatement test. However, if there were only a very brief (seconds) and relatively small response it may not have been detectable over the 3 min sampling period used here, and other techniques may be required to study the effects of IntA on such conditioned responses.
It has been suggested that addiction is characterized by a hypodopaminergic, anhedonic state, and compulsive motivation to seek and take cocaine derives from a desire to overcome this DA deficiency (Dackis & Gold, 1985; Koob & Le Moal, 1997, 2001; Blum et al., 2015; Volkow et al., 2016). Reports that LgA cocaine self-administration experience reduces DA function have been interpreted as support for this view, especially given that LgA was thought to best model changes in brain and behavior that lead to a transition from casual patterns of drug use to the escalated use that characterizes addiction. However, as reviewed above, the evidence that LgA produces a hypodopaminergic state is equivocal, as is evidence it increases motivation for cocaine. In addition, studies using the more recently developed IntA self-administration procedure support a different theory. The IntA procedure was initially developed because it is thought to better model intermittent patterns of cocaine use in humans, especially during the transition to addiction (Zimmer et al., 2012; Allain et al., 2015). There is now considerable evidence that IntA produces incentive-sensitization, and is more effective than LgA in producing addiction-like behavior (Kawa et al., 2016; Allain et al., 2017, 2018; Allain & Samaha, 2018; James et al., 2018; Kawa & Robinson, 2018). Although the evidence is limited, and more work is required, the available evidence indicates that IntA experience also sensitizes DA function (Calipari et al., 2013, 2015), including the ability of cocaine to increase extracellular DA in vivo, as reported here.
In conclusion, studies utilizing the IntA procedure are more consistent with the view that the pathological motivation to seek and take cocaine in addiction is due, at least in part, to a hyper-responsive dopaminergic state, consistent with an incentive-sensitization view of addiction (Robinson & Berridge, 1993; Berridge & Robinson, 2016). Of course, a syndrome as complex as addiction is not going to be reducible to changes in a single neurotransmitter system, or even a single psychological process, and it remains to be seen what other neuropsychological functions are altered by IntA experience (e.g., Allain et al., 2017; Gueye et al., 2018; James et al., 2018). Nevertheless, the growing evidence concerning the importance of pharmacokinetic factors in promoting the development of addiction suggests these need to be given greater consideration in preclinical models of addiction (Allain et al., 2015).
;