Repeated dosing with cocaine produces strain-dependent effects on responding for conditioned reinforcement in Collaborative Cross mice

Abstract

Rationale.

Cocaine use disorder (CUD) is a highly heritable form of substance use disorder, with genetic variation accounting for a substantial proportion of the risk for transitioning from recreational use to a clinically impairing addiction. With repeated exposures to cocaine, psychomotor and incentive sensitization are observed in rodents. These phenomena are thought to model behavioral changes elicited by the drug that contribute to the progression into addiction, but little is known about how genetic variation may moderate these consequences.

Objectives.

Here, we describe the use of two Collaborative Cross (CC) recombinant inbred mouse strains that either exhibit high (CC018/UncJ) or no (CC027/GeniUncJ) psychomotor sensitization in response to cocaine to measure phenotypes related to incentive sensitization after repeated cocaine exposures; given the relationship of incentive motivation to nucleus accumbens core (NAc) dopamine release and reuptake, we also assessed these neurochemical mechanisms.

Methods.

Adult male and female CC018/UncJ and CC027/GeniUncJ mice underwent Pavlovian conditioning to associate a visual cue with presentation of a palatable food reward, then received five, every-other-day injections of cocaine or vehicle. Following Pavlovian re-training, they underwent testing acquisition of a new operant response for the visual cue, now serving as a conditioned reinforcer. Subsequently, electrically evoked dopamine release was assessed using fast scan cyclic voltammetry from acute brain slices containing the NAc.

Results.

While both strains acquired the Pavlovian association, only CC018/UncJ mice showed conditioned reinforcement and incentive sensitization in response to cocaine, while CC027/GeniUncJ mice did not. Voltammetry data revealed that CC018/UncJ, compared to CC027/GeniUnc, mice exhibited higher baseline dopamine release and uptake. Moreover, chronic cocaine exposure blunted tonic and phasic dopamine release in CC018/UncJ, but not CC027/GeniUncJ, mice.

Conclusions.

Genetic background is a moderator of cocaine-induced neuroadaptations in mesolimbic dopamine signaling, which may contribute to both psychomotor and incentive sensitization and indicate a shared biological mechanism of variation.

Only a subset of individuals that initiate use of cocaine will develop a cocaine use disorder (CUD; Fernandez-Castillo, 2021); for those that do, the transition from recreational use to a CUD is mediated by both environmental and genetic influences, as well as their interactions (Goldman, Oroszi, & Ducci, 2005; Kendler et al., 2003; Kreek et al., 2012; Pierce et al., 2018). Though environmental factors such as drug availability, household and/or peer drug use and early life adversity predict CUD development (Olivares et al., 2016; Pierce et al., 2018; Stone et al., 2012), the genetic component of CUD liability is consistently found to be one of the largest risk factors. Indeed, CUD heritability estimates range from ~65% to ~79%, with this magnitude possibly varying as a function of sex (Pierce, 2018). Considering the evident heritability of CUD, understanding the genetic risk factors and underlying biology of cocaine use is necessary to decrease risk and develop more effective interventions.

Just as there are heritable individual differences in cocaine use and CUD risk in humans, various responses to the drug are heritable in laboratory animals (Bagley et al., 2019; Bagley et al., 2022; Boyle & Gill, 2001; Cervantes, Laughlin, & Jentsch, 2013; Dickson et al., 2016; Gaines et al., 2022; Jones et al., 1999; Kantak et al., 2021; Philip et al., 2010; Phillips, Huson, & McKinnon, 1998; Wiltshire et al., 2015). Locomotor activation in response to acute psychostimulant exposure is commonly used to model initial drug sensitivity in rodents. Moreover, repeated cocaine dosing can elicit long-lasting, augmented locomotor behavior, referred to as behavioral or psychomotor “sensitization” (Kalivas, 1995; Steketee & Kalivas, 2011; Stewart & Badiani, 1993; Vanderschuren & Kalivas, 2000). The degree of psychomotor sensitization to repeated exposures to cocaine is believed to result from drug-induced neuroadaptations in mesocorticolimbic circuitry (Nestler, 1994; Nestler, 2001; Self & Nestler, 1995). Notably, just like CUD in humans, behavioral sensitization to cocaine in rodents is heritable (Eisener-Dorman, Grabowski-Boase, & Tarantino, 2011; Philip et al., 2010). Sensitization of locomotor responses has also been observed in human subjects after repeated amphetamine administrations (Boileau et al., 2006, Sax & Strakowski, 1998, Strakowski et al., 1996), a phenotype that emerged concurrent to enhanced dopamine release and was also predicted by degree of trait impulsivity (Boileau et al., 2006) and novelty seeking (Sax & Strakowski, 1998). Nevertheless, the validity of behavioral sensitization in rodents is better contextualized by recognizing its correlation to aspects of volitional drug-seeking behavior, including enhanced acquisition of self-administration and reinstatement following extinction (Koeltzow & Vezina, 2005; Piazza et al., 1989; Pierre and Vezina, 1997; Steketee & Kalivas, 2011).

The neuroadaptations elicited by repeated cocaine exposure alters incentive motivation systems and, consequently, the response to rewards and associated cues (Berridge & Robinson, 1998; Valjent et al., 2006). The resulting state of “incentive sensitization” is relevant to features of human addiction, particularly craving elicited by drug-predictive stimuli (Robinson & Berridge, 1993). Specifically, repeated exposure to potentially addictive drugs (including, but not limited to, cocaine) can increase the incentive salience attributed to rewards and associated cues, thereby eliciting extreme wanting/craving. In laboratory animals, incentive sensitization can be measured using a test of conditioned reinforcement (CR). CR testing typically involves measuring the acquisition of a new instrumental response that is reinforced by presentations of conditioned stimulus (e.g., an audiovisual cue with a history of association with a primary reward; Robbins et al., 1989; Taylor & Robbins, 1984). Repeated, intermittent cocaine administration augments responding for CR in rats, a possible reflection of incentive sensitization. Similar effects are observed in mice after repeated intermittent dosing with another psychomotor stimulant drug, amphetamine (Mead, Crombag, & Rocha, 2004). Individual differences in incentive sensitization are related to biological sex (Kawa & Robinson, 2019), but no studies have examined whether cocaine-induced incentive sensitization is influenced by genetic variation. Given that aspects of human addiction, particularly craving, may be modeled by cocaine-induced incentive sensitization in rodents, using this phenotype to identify associated, functionally relevant polymorphisms may reveal new biological insights into the disposition to develop CUD.

Neurogenetic analysis of behavioral traits can be greatly aided through the use of advanced genetic mouse populations. The Collaborative Cross (CC) is a recombinant inbred mouse panel that was generated from an 8-way cross of genetically heterogenous inbred founder strains (Threadgill et al., 2011). We measured cocaine-induced behavioral sensitization in the CC and have identified strains that show extreme differences in behavioral sensitization (Schoenrock et al., 2022). The identification of extreme responding strains provides not only an opportunity to identify the role of genetic variation in moderating cocaine-induced incentive sensitization, but also to examine whether heritable differences in cocaine-induced behavioral sensitization confer differences in incentive sensitization to the drug. Moreover, the identification of strain-level differences in incentive sensitization in a diverse genetic reference population would enable investigation of the associated cocaine-induced neuroadaptations that correlate with, and perhaps cause, behavioral trait variation: either uniformly in incentive and behavioral sensitization, or how such neuroadaptations uniquely contribute to the development of each dissociable phenotype.

In the present study, we selected two CC strains that exhibited substantial differences in behavioral sensitization after repeated exposures to cocaine (Schoenrock et al., 2022). CC018/UncJ (high sensitizing) and CC027/GeniUncJ (low sensitizing) mice show similar basal locomotor activity and locomotor response to a first dose of cocaine; however, cocaine-induced locomotor activity in CC018/UncJ mice increases upon repeated exposures to the drug, whereas CC027/GeniUncJ mice do not exhibit behavioral sensitization. To examine whether these strain differences in behavioral sensitization conferred incentive sensitization, males and females from each strain were first trained to associate a visual cue with delivery of a palatable reward in a Pavlovian conditioning procedure. They then received cocaine or saline injections using the same schedule of drug exposure described by Schoenrock et al. (2022). After a drug-free period, they were tested for CR. We hypothesized that high sensitizing CC018/UncJ mice would show greater incentive sensitization following cocaine exposure, such that they would lever press more for the CR than the low sensitizing CC027/GeniUncJ mice. As past studies have indicated that repeated intermittent cocaine administration can elicit neuroadaptations in mesocorticolimbic dopamine signaling (Anderson & Pierce, 2005; Orsini et al., 2018; Self, 2004; Thomas, Kalivas, & Shaham, 2009; Wiskerke, Schoffelmeer, & de Vries, 2016), as well as providing strong evidence linking dopamine to incentive motivation (Flagel et al., 2011; Salamone et al., 2007; Taylor & Robbins, 1986) we also examined aspects of dopamine release in the nucleus accumbens using fast scan cyclic voltammetry (FSCV) measurements in brain slices taken from these subjects.

Methods

Incentive Sensitization Studies

CC018/UncJ (RRID: IMSR_JAX:021890) and CC027/GeniUncJ (RRID: IMSR_JAX:025130) mice were bred at The Jackson Laboratory (Bar Harbor, ME) and subsequently shipped to Binghamton University for testing. All procedures were performed according to the “Guide for the Care and Use of Laboratory Animals” (National Research Council, 2011) in the AAALAC accredited programs at Binghamton University after Institutional Animal Care and Use Committee approval.

Subjects were shipped between 36–48 days of age in one of 6 cohorts, between August 2020 and July 2021. Mice were shipped based on availability with effort to balance cohorts for equal representation of strain and sex. Cohorts 1, 3, and 4 included representatives from both strains, cohorts 2 and 5 included only CC018/UncJ mice, and cohort 6 included only CC027/GeniUncJ mice. Details on the mice are provided in Table 1.

Table 1.

Conditioned reinforcement sample size by strain and sex.

Cohort	CC018 females	CC018 males	CC027 females	CC027 males	Total
1	3	6	8	8	25
2	4	2	0	0	6
3	2	1	1	8	12
4	4	3	4	0	11
5	5	7	0	0	12
6	0	0	6	5	11
Total	18	19	19	21	77

Upon arrival, mice were housed as shipped, in groups of 2–3 mice of the same strain and sex in a cage. The colony room was maintained on a 12-h/12-h light/dark cycle (lights on at 0615 h) and at an average temperature of 68°F. After arrival and during acclimation, food (Lab Diet 5001, ScottPharma Solutions) and water was available ad libitum. Cage enrichment included a red acrylic tube, a paper nestlet with Carfil nesting material, and a wooden chew stick. Mice were removed from the cage by picking them up with their tails using a gloved hand, as per our laboratory’s handling procedure when working with these strains. At 60 days, mice were individually housed in otherwise identical caging conditions.

Prior to the initiation of the operant conditioning protocols described below, mice were introduced to a schedule of limited access to chow. Mice were weighed while still receiving ad libitum access to chow, and that weight was recorded as their initial free feeding weight. Once limited access to food was introduced, each mouse was weighed daily, and its weight was divided by its free feeding weight to obtain a percentage change in body weight. During the period of limited access to food, mice were fed once a day in the early afternoon; chow quantity provided per day was titrated until mice reached 90 +/− 3% of their free feeding weights, at which point operant testing began. If, at any point during the testing period, a mouse dropped below 80% of its free feeding weight, the quantity of chow it was provided was increased. If increasing food availability did not lead to a recovery of body weight to > 80% within a day, the mouse was temporarily returned to ad libitum access to food until its weight had recovered.

Drug treatment

Cocaine hydrochloride was obtained from the National Institute on Drug Abuse Drug Supply Program and was prepared in a 0.9% saline solution at 1 mg/mL.

Phenotyping protocols

All operant testing utilized a highly palatable chocolate flavored Boost^™ solution (Nestle) as a food reward. Mice were introduced to Boost^™ for two days in their home cage. Boost^™ was dispensed in a petri dish and placed on top of the bedding; the solution was available for a 48-hour period and refreshed at the 24-h time point. Following Boost^™ exposure, mice were transitioned to the conditioning and testing procedures described below. Testing took place in 8.5” L × 7” W × 5” H operant modular chambers (Model ENV-307W, Med Associates Inc.) equipped with a stainless-steel grid floor (Model ENV-307W-GFW, Med Associates Inc.) two retractable levers and a photocell-equipped magazine on one side, and five nose poke holes capable of illumination on the other side. Each test chamber was housed within a sound attenuating cubicle with an overhead house light and fan. On each test day, mice were removed from their home cage by their tail and placed inside the testing box. Operant testing was conducted during the light part of the light/dark cycle between 10:00 and 15:00.

Mice were tested on the following phenotyping pipeline:

D1: Box Habituation

Mice were placed in the testing chamber with house lights and white noise active. There were no other programmed events. The habituation session lasted 30 min.

D2: Magazine Training

Mice were placed in the testing chamber with the house light and white noise active. 20–21 μl Boost was dispensed every 30-s following retrieval of the previous reward. The session ended after 1-h or after the mouse received 50 rewards, whichever came first.

D3–8: Pavlovian Conditioning

Pavlovian conditioning took place in six consecutive daily sessions. Mice were placed in the testing chamber, with the white noise active for the duration of the test. Paired presentations of a conditioned stimulus (CS; 6-s illumination of the magazine light) and an unconditioned stimulus (US; four second activation of the pump to dispense 20–21μl Boost) occurred every 60, 90, 120, 150 or 180 s and the magazine light co-terminated with the end of the reward delivery. The session ended after 1-h or after the mouse received 30 CS presentations, whichever came first. Duration of head entries during the intertrial interval (ITI), the CS period and the US period, as well as latency (in cs) to enter the magazine after CS onset, were recorded. In this procedure, subjects are learning various associations. For example, they learn that Boost is available intermittently in the magazine in the testing context; the acquisition of a context-reward association likely contributes to increased magazine checking at all times (ITI, CS, US). In addition, the acquisition of a cue-reward association may contribute additionally to the proclivity to rapidly approach the magazine after CS onset, leading to longer duration head entries during the CS and US periods, along with decreasing latency to initiate a head entry after CS onset. Overall, the strongest evidence that the CS has acquired motivational properties through the training is the subsequent expression of CR; that is, that the mouse will learn an instrumental response that is reinforced only by presentations of the CS.

D11–19: Cocaine and Vehicle Exposure

All mice received a two-day testing break prior to injections. Either drug or vehicle were administered every other day during the treatment period and on the final day of conditioned reinforcement testing.

During this treatment period, mice were weighed and injected every other day (D11, D13, D15, D17, D19) for five total exposures. Animals were assigned to drug condition (10 mg/kg cocaine or saline) counterbalanced across strains and sexes, with group assignment additionally assessed to ensure that CS and US magazine head entries in the final days of Pavlovian training were equivalent in the two treatment groups. Drug or saline were administered by intraperitoneal (IP) injections at a volume of 10 mL/kg body weight.

Mirroring the methods reported previously (Taylor and Horger 1988; Mead et al. 2004), mice were placed in the testing chambers immediately following injection. Background white noise was broadcast, but no other programmed events occurred. After 30 min, mice were removed from the chamber and returned to their home cage.

D22–23: Pavlovian Conditioning

All mice received a two-day break from testing prior to experiencing two reminder sessions of Pavlovian conditioning. After the two days, mice were placed in the operant chamber and exposed to a procedure identical to Pavlovian conditioning sessions described above.

D26–27: Conditioned Reinforcement Test

All mice received an additional two-day break prior to conditioned reinforcement (CR) testing. Following the two day break, mice were placed into the operant box with the white noise active and two ultrasensitive levers (left and right, to either side of the magazine) extended. One lever (left or right) was assigned as active and the other was assigned as inactive. Lever assignment was counterbalanced across strain, sex, and drug treatment. Active lever presses resulted in ten second CS presentation magazine light (this test is conducted in extinction, so no Boost was presented). After three active lever presses were made, a countdown timer of 1-h was initiated. Active (CS-reinforced) lever presses, inactive (no consequence) lever presses and the number of CSs earned were recorded.

Active lever presses are thought to measure the incentive motivational properties of the CS cue. Inactive lever presses served as a control for baseline activity or willingness to lever press without a reinforcing consequence.

Ex Vivo Fast Scan Cyclic Voltammetry

Ex vivo FSCV was used to characterize dopamine dynamics in the NAc core 3 to 4 days after the final day of behavioral testing, in a subset of mice (saline [n = 11] and cocaine [n = 10] exposed CC018/UncJ mice and saline [n = 9] and cocaine [n = 10] exposed CC027/GeniUncJ mice). Mice were euthanized by rapid decapitation under isoflurane anesthesia, and brains were rapidly removed and cooled in preoxygenated (95% O₂/5% CO₂), ice-cooled artificial cerebrospinal fluid (aCSF) consisting of 126 mM NaCl, 2.5 mM KCl, 1.2 mM NaH₂PO₄, 2.4 mM CaCl₂, 1.2 mM MgCl₂, 25 mM NaHCO₃, 11 mM glucose, and 0.4 mM l-ascorbic acid (pH adjusted to 7.4). A vibrating tissue slicer (Leica VT1200S, Leica Biosystems) was used to prepare 300 μm thick coronal brain sections while immersed in aCSF. For recording, slices containing the nucleus accumbens core (NAc core) were transferred to a submersion recording chamber perfused with oxygenated aCSF at the rate of 1 mL/min at 32 °C.

Endogenous dopamine release was evoked by single (tonic) stimulation or multiple electrical pulse (phasic) stimulation. Under phasic stimulation, both 2-pulse and 5-pulse stimulations were applied, each at 4 frequencies (5, 10, 20 and 100 Hz). Tonic and phasic stimulations were applied every 5 min. Initial stimulations were applied under tonic parameters until a stable baseline was established (<10% variation over last 3 baseline stimulations) followed by sequential application of phasic stimulations.

Extracellular dopamine concentration was recorded by applying a triangular waveform (−0.4 to +1.2 and back to −0.4 V, Ag vs AgCl) at a rate of 400 V/s. Extracellular concentrations of dopamine were assessed by comparing the current at the peak oxidation potential for dopamine with electrode calibrations of known concentrations of dopamine (3 μM).

Statistical analyses

Pavlovian approach training

Duration (cs) of head entries during the intertrial interval, duration (cs) of head entries during the CS presentation only, duration (cs) of head entries during the CS and US presentation, and latency (cs) per trial were collected for the twenty trials per day and then averaged within day for each of the six days. The averaged values were then assessed in separated mixed design ANOVAs, in which strain, sex, and assigned treatment group were between subject measures and day was within subjects. Latency data on Day 1 was omitted due to a technical error in collection, and only Days 2–6 were analyzed.

Pavlovian re-training

Averages for each variable were derived similarly to the previous procedure across two separate days. A mixed model ANOVA was conducted for intertrial interval head entry duration, CS head entry duration, US head entry duration, and latency, with strain, sex, and drug treatment as between-subjects factors and day as a within-subjects factor.

Conditioned reinforcement testing

Active lever presses, inactive lever presses, and CS-presentation head entry duration (cs) were collected and each variable was averaged within testing day. Lever presses were analyzed in a mixed model ANOVA with strain, sex, and drug treatment as between-subjects factors and lever (active or inactive) as the within-subjects factor.

Ex-Vivo Fast Scan Cyclic Voltammetry

All FSCV data were analyzed using Demon Voltammetry and Analysis software (Yorgason et. al., 2011), utilizing the Michaelis–Menten kinetics-based algorithm (Wightman et. al., 1988).

Results

Pavlovian approach training

Analyses of head entry duration data during the CS period revealed a strain by day interaction (F[5,330]=7.297, p<.001, η_p2=.303), along with main effects of strain (F[1,66]=4.215, p<.05, η_p2=.06) and day (F[5,330]=99.978, p<.001, η_p2=.834). Sex did not modulate behavior in this procedure (F[1,66]=.407, p=.526, η_p2=.006), nor did it interact with any other variables. There was additionally no effect of future exposure condition (saline vs. cocaine; F[1,66]=.345, p=.559, η_p2=.005), indicating that the two exposure groups had comparable head entry duration time prior to exposure group assignment. Overall, low sensitizing CC027/GeniUncJ mice made more head entries during the CS presentation as compared with high sensitizing CC018/UncJ mice, and both strains significantly increased head entry time across testing days (Figure 1A). The strain by day interaction was further investigated through day-wise contrasts that revealed that CC027/GeniUncJ mice had longer head entry durations on Days 1–3 (all p<.05), but not on Days 4–6 (all p>.09) of testing.

Figure 1. Strain differences during Pavlovian approach training.

A. Magazine head entry duration (cs) during the CS period in CC027/UncJ (n=38) and CC018 (n=36) mice. B. Magazine head entry duration (cs) during the US period. C. Magazine head entry duration (cs) during the inter-trial interval (ITI). D. Latency (cs) to enter magazine after CS onset. Data shown are the mean±S.E.M. Asterisks indicate significant strain differences (p<.05).

Magazine entry duration during the CS period was also normalized to ITI head entry duration (CS head entry duration/ITI head entry duration) to assess whether CS-paired head entry duration was disproportionately increasing across training, as would be expected if the CS is gaining motivational properties. A day by strain interaction (F[5,330]=6.517, p<.001, η_p2=.09) was observed, as was a main effect of day (F[5,330]=49.042, p<.001, η_p2=.426), but the main effect of strain was no longer significant after normalization (F[1,66]=.043, p=.837, η_p2=.001). Overall, these data suggest that mice from both strains are exhibiting progressively increasing magazine checking during the CS period, as would be expected if they are acquiring the CS-US association.

A strain by day interaction was similarly found for US head entries (F[5,330]=5.873, p<.001, η_p2=.257), along with main effects of strain (F[1,66]=26.128, p<.001, η_p2=.090) and day (F[5,330]=107.495, p<.001, η_p2=.820). CC027/GeniUncJ mice made more head entries during the US presentation, and both strains increased head entry time across testing days (Figure 1B). This interaction was probed for day, which revealed that CC027/GeniUncJ mice engaged in significantly longer head entry durations on Days 1–4 of testing (all p<.001) but not on Days 5–6 (all p>.115). There were no significant effects of sex (F[1,66]=.016, p=.900, η_p2=.001) or future exposure condition (F[1,66]=.093, p=.761, η_p2=.001), again indicating no group differences in US head entry duration prior to exposure group assignment.

As above, US period entry data was normalized to ITI head entry duration. This analysis yielded a strain by day interaction (F[5,330]=5.000, p<.001, η_p2=.070) and main effect of day (F[5,330]=30.390, p<.001, η_p2=.315), but no main effect of strain (F[1,66]=3.160, p=.08, η_p2=.046).

ITI head entry analysis revealed a strain by day interaction (F[5,330]=3.050, p<.01, η_p2=.044) and main effects of day (F[5,330]=18.295, p<.001, η_p2=.216) and strain (F[1,66]=15.284, p<.001, η_p2=.188). CC027/GeniUncJ mice had increased head entry durations during the ITI in comparison with CC018/UncJ mice, though both strains increased their head entry durations across days (Figure 1C). Day-wise contrasts revealed that CC027/GeniUncJ mice had significantly higher ITI head entry duration on Days 1–4 and Day 6 (p<.001) but did not differ from CC018/UncJ mice on Day 5 (p=.830). There were no significant effects of sex (F[1,66]=1.362, p=.247) or future exposure condition (F[1,66]=.489, p=.487), again indicating no group differences in ITI head entry duration prior to exposure group assignment.

Analysis of latency data revealed a three-way strain by sex by day interaction (F[4,264]=2.561, p<.05, η_p2=.196), as well as a strain by day interaction (F[4,264]=7.824, p<.001, η_p2=.253) and main effects of strain (F[1,66]=6.295, p<.05, η_p2=.087) and day (F[4,264]=27.097, p<.001, η_p2=.522; Figure 1D). Post-hoc analyses revealed that CC018/UncJ mice were slower to approach the magazine after CS onset on Days 2–3 of Pavlovian training (all p<.001), though no significant differences were found on Days 4–6 (all p>.224). Further post-hoc contrast analyses within strain and sex showed that this effect was found in both male (F[1,36]=12.234, p<.001) and female (F[1,30]=4.173, p<.05) subjects, though CC018/UncJ male mice showed significantly longer latencies to approach in comparison with CC027/GeniUncJ male mice (p<.05), until Day 6 of testing (p=.473). Again, there was no effect of future exposure condition (F[1,66]=.135, p=.714, η_p2=.001). These data indicate that the CS was acquiring increasing ability to elicit behavior across training days and that these effects were strain dependent.

Pavlovian re-training

Following repeated saline or cocaine exposure, data was collected on two consecutive days of Pavlovian re-training. A main effect of day was observed for both CS head entry duration (F[1,66]=26.496, p<.001, η_p2=.294; Figure 2A) and US head entry duration (F[1,66]=9.601, p<.01, η_p2=.127; Figure 2B), along with a significant main effect of strain for US head entry duration (F[1,66]=22.190, p<.001, η_p2=.252) but not CS head entry duration (F[1,66]=2.930, p=.092, η_p2=.043). This indicates that while both strains engaged in a longer duration of head entries during the CS and US presentation on the second day of re-testing, CC027/GeniUncJ mice had longer head entry duration during the US presentation compared to CC018/UncJ mice (p<.05). Strains did not differ on head entries during the CS presentation. There was no effect of exposure condition or sex on CS (F[1,66]=.109, p=.743, η_p2=.002; F[1,66]=.001, p=.977, η_p2=.001) or US (F[1,66]=.006, p=.938, η_p2=.001; F[1,66]=.100, p=.753, η_p2=.002) head entry duration. ITI head entry analysis found a main effect of strain that approached significance (F[1,66]=3.866, p=.053, η_p2=.055; Figure 2C), along with no significant effect of day (F[1,66]=.232, p=.632, η_p2=.003), sex (F[1,66]=2.652, p=.108, η_p2=.039) or drug exposure condition (F[1,66]=.448, p=.505, η_p2=.007). Consistent with pre-drug findings, CC027/GeniUncJ tended to engage in more head entries during the ITI interval compared to the CC018/UncJ mice, and this behavior in both strains is not influenced by cocaine or saline exposure.

Figure 2. Strain differences during Pavlovian re-training.

A. Magazine head entry duration (cs) during the CS period in CC027/UncJ (n=38) and CC018 (n=36) mice. B. Magazine head entry duration (cs) during the US period. C. Magazine head entry duration (cs) during the inter-trial interval (ITI). D. Latency (cs) to enter magazine after CS onset. Data shown are the mean±S.E.M. Asterisks indicate significant strain differences (p<.05).

Analyses of latency to approach the magazine revealed main effects of strain (F[1,66]=4.981, p<.05, η_p2=.070) and day (F[1,66]=11.067, p<.001, η_p2=.144), with findings consistent with pre-exposure testing: CC018/UncJ had slower latency after CS onset, though both strains reduced their latency across the two days (Figure 2D). There was no effect of drug exposure condition (F[1,66]=.280, p=.599, η_p2=.004) or sex (F[1,66]=.163, p=.688, η_p2=.002) on latency. These results confirm that CC027/GeniUncJ mice more readily approach the magazine, but latency decrease in both strains reflects successful CS-US learning.

Conditioned reinforcement testing

A mixed design ANOVA was conducted with strain, sex and exposure condition as between-subjects factors and lever (active or inactive) as a within-subject factor. These analyses revealed a trending strain by lever interaction (F[1,66]=3.614, p=.062, η_p2=.052; Figure 3). A contrast was conducted to isolate strain differences. Independent analysis of CC018/UncJ mice showed an exposure condition by lever interaction (F[1,32]=4.451, p<.05, η_p2=.122). Prior cocaine exposure increased active lever presses without altering inactive lever presses. An identical analysis was conducted on CC027/GeniUncJ mice and showed that chronic cocaine exposure did not alter responding on either lever (F[1,34]=.790, p=.380, η_p2=.023). CC018/UncJ mice demonstrated preference for the active lever that was enhanced by cocaine exposure, reflecting enhanced incentive salience towards the CS; this was not found in the CC027/GeniUncJ mice, that failed to show conditioned responding through preference of the active lever, as well as no potentiation of this response following cocaine exposure.

Figure 3. Strain and exposure effects in conditioned reinforcement.

Individual contrasts show a significant increase in active lever presses in the CC018/UncJ strain between saline and cocaine conditions, while no difference was observed in the CC027/GeniUncJ strain and exposure condition. Data shown are the mean±S.E.M. Asterisks indicate significant strain differences (p<.05).

Fast Scan Cyclic Voltammetry

Tonic Stimulation

Mice of both strains and treatment groups were tested for ex vivo stimulated dopamine release in the nucleus accumbens (NAc) core using FSCV. Dopamine release, following electrical stimulation, was assessed by between-subjects ANOVA. Analyses revealed a strain by exposure condition interaction (F[1,32]=4.2, p=0.048, η_p2=0.116) and a main effect of strain (F[1,32]=12.2, p=0.001, η_p2=0.276; Figure 4A). Post-hoc analyses within strain revealed that cocaine exposure suppressed dopamine release in CC018/UncJ (p=0.014) but had no effect on CC027/GeniUncJ (p=0.714). Furthermore, post-hoc analyses within exposure groups revealed an effect of strain (p<0.001) in the saline group but no effect of strain in the cocaine group (p=0.315); the former finding reveals that cocaine naïve CC018/UncJ mice exhibit greater dopamine release relative to CC027/GeniUncJ. No effects of sex were found (F[1,32]=2.2, p=0.147, η_p2=0.065).

Figure 4. Dopamine release and uptake under tonic stimulation.

A. CC018/UncJ demonstrated greater tonic dopamine release under drug-naive (saline group) conditions. Furthermore, CC018/UncJ, but not CC027/UncJ, demonstrated reduced dopamine release after chronic cocaine treatment. B. CC018/UncJ demonstrated greater dopamine uptake relative to CC027/UncJ. No significant effects of cocaine treatment were found.

The dopamine uptake rate (V_max, μM/sec) was assessed by ANOVA. The CC018/UncJ strain demonstrated greater uptake relative to CC027/GeniUncJ (main effect of strain F[1,32]=4.2, p=0.049, η_p2=0.116; Figure 4B). No sex (F[1,32]=0.6, p=0.452, η_p2=0.018) or exposure group (F[1,32]=0.7, p=0.397, η_p2=0.023) effects were observed.

Phasic Stimulation

Following tonic stimulation, dopamine release under phasic stimulation was assessed. A mixed ANOVA, with frequency of stimulation (Hz) as a within-subjects factor and exposure condition, strain and sex as between-subjects factors, was conducted for both 2-pulse (Figure 5A) and 5-pulse stimulations (Figure 5B). Under 5-pulse stimulation, analyses revealed a strain by frequency by exposure condition interaction (F[3,96]=2.8, p=0.043, η_p2=0.081), a strain by frequency interaction (F[3,96]=10.2, p<0.001, η_p2=0.242), an exposure condition by frequency interaction (F[3,96]=4.7, p=0.004, η_p2=0.128), a main effect of frequency (F[3,96]=41.8, p<0.001, η_p2=0.566) and a main effect of strain (F[1,32]=16.0, p<0.001, η_p2=0.334). Post-hoc analyses indicated that cocaine exposure suppressed dopamine release at all frequencies in the CC018/UncJ (p<0.05), but not in the CC027/GeniUncJ strain, effects that are similar to those observed under tonic stimulation. Furthermore, in the saline groups, CC018/UncJ exhibited greater release than CC027/GeniUncJ at all frequencies (p<0.05), similar to tonic stimulation. Under 2-pulse stimulation, a strain by frequency interaction (F[3,93]=6.1, p=0.001, η_p2=0.165), a main effect of strain (F[1,31]=14.9, p=0.001, η_p2=0.324) and a main effect of frequency (F[3,93]=10.2, p<0.001, η_p2=0.247) were found. The strain by frequency-exposure condition interaction approached significance (p=0.06), suggesting similar effects of cocaine to that observed in 5-pulse and tonic stimulation. No sex effects were observed (2-pulse, F[3,93]=1.4, p=0.243, η_p2=0.044; 5-pulse, F[3,93]=0.8, p=0.370, η_p2=0.025).

Figure 5. Dopamine release under 2-pulse or 5-pulse phasic stimulation.

A. Under 2-pulse stimulation, a main effect of strain was detected, indicating that CC018/UncJ had greater dopamine release at all frequencies tested. B. Under 5-pulse stimulation, cocaine exposed mice of the CC018/UncJ strain demonstrated greater dopamine release relative to saline exposed mice of the same strain, at all frequencies tested. Furthermore, CC018/UncJ demonstrated greater dopamine release relative to CC027/UncJ under drug-naive (saline group) conditions.

Discussion

The results of this study provide evidence that repeated cocaine administration augments incentive sensitization as measured by CR, an effect that may be associated with blunted accumbal core dopamine release; notably, both effects were moderated by genetic background. This study also marks one of the first efforts to relate cocaine-induced psychomotor sensitization to sensitized responding for an incentive stimulus, as a function of genetic variation. It also provides a neurobiological basis for strain-dependent emergence of cocaine-exposure related increases in responding for CR. Selection of the two candidate strains in this study was driven by our observation that they exhibit divergent psychomotor sensitization phenotypes after repeated, intermittent dosing with cocaine: CC018/UncJ mice showed heightened sensitization in comparison with CC027/GeniUncJ mice which did not show sensitization (Schoenrock et al. 2022). This pattern was recapitulated when we examine strain-dependent effects of cocaine on CR; specifically, chronic cocaine exposure increased active (CR) lever pressing only in CC018/UncJ mice. These behavioral findings are considered in light of the NAc dopaminergic firing pattern in each strain; cocaine exposure was sufficient to blunt tonic and phasic dopamine release in CC018/UncJ mice, while not affecting release in CC027/GeniUncJ mice. Overall, the results of this study are consistent with the hypothesis that genetic background moderates cocaine-induced neuroadaptations in mesolimbic dopamine signaling, which may contribute to both psychomotor and incentive sensitization and indicate a shared biological mechanism of variation.

Strain Differences in Pavlovian Responses and CR

CC018/UncJ and CC027/GeniUncJ strains exhibited a variety of significant behavioral differences. Both strains exhibited evidence of Pavlovian conditioning, increasing the duration of magazine approach behavior during the CS period as training progressed. During Pavlovian approach training, prior to cocaine/saline exposure, CC027/GeniUncJ mice initially engaged in more head entries during US, CS, and ITI periods compared to CC018/UncJ mice, though training tended to eliminate these differences. Given that the strain differences were greatest at the earliest stages of training, our data suggest that these differences may be attributable to baseline differences in investigatory or stimulus-evoked behavior; importantly, however, normalizing to ITI head entries still yielded CS period-specific head entry differences across strain, perhaps also suggesting strain-level differences in learning in response to cue-reward pairings. Regardless, this finding is unlikely to be attributable to baseline activity level, as previous data has indicated no locomotor differences between the CC018/UncJ and CC027/GeniUncJ strains (Schoenrock et al., 2022). Strain differences in head entries were again observed during Pavlovian re-testing, though only during the US period. CC027/GeniUncJ mice also exhibited shorter approach latencies until the final day of training. This strain effect was again observed in the re-testing period; this finding suggests that CC027/GeniUncJ are quicker than CC018/UncJ mice to approach the magazine following CS presentation irrespective of the level of Pavlovian conditioning.

Our findings demonstrate that Pavlovian association learning is not sufficient to stimulate CR (more responding on a lever that triggers the CS, as compared to responding on the inactive lever), and that the CC027/GeniUncJ strain’s failure to show CR cannot be attributed to an inability to acquire a conditional response to the CS (Pavlovian approach behavior). Both CC018/UncJ and CC027/GeniUncJ strains increased magazine approach behavior during the CS period across training sessions and during the Pavlovian re-training sessions, before and after cocaine or saline exposure, respectively. However, CC018/UncJ mice exhibited greater active than inactive lever pressing, while CC027/GeniUncJ mice failed to show an active lever preference, revealing strain differences in CR, with and without cocaine exposure. Coupled with training data, CC027/GeniUncJ mice appear to interact more readily with aspects of the testing chamber (magazine, levers), though this behavior does not translate to the enhanced acquisition of CR.

Chronic Drug Effects

While previous studies have established the ability of cocaine or amphetamine to enhance responding for CR (Taylor and Horger, 1999; Mead et al. 2004), this measure of incentive sensitization has not previously addressed the moderating influence of genetic background. Several studies have observed enhanced incentive sensitization and addiction-like behaviors through chronic intermittent access versus long access (Carr, Ferrario, & Robinson, 2020; Kawa, Bentzley, & Robinson, 2016), as well as through intravenous self-administration versus unsignaled delivery (LeBlanc, Maidment, & Ostlund, 2013). Regarding biological sex differences, it has been demonstrated that female rats show greater cocaine-induced psychomotor (Algallal et al., 2019) and incentive (Kawa & Robinson, 2019) sensitization over male rats, when cocaine is self-administered in an intermittent access procedure. This heightened sensitization may be attributed to activational effects of sex steroids, particularly estradiol levels during the follicular phase of the estrous cycle, which has been found to enhance self-administration of cocaine (Jackson, Robinson, & Becker, 2006) and cocaine-induced locomotor activity (Quinones-Jenab et al., 2020). Augmented psychomotor sensitization has been observed in the CC founder strains (Schoenrock et al., 2022), though the small sample sizes in that study did not make a formal analysis of sex differences possible. It therefore remains unknown how biological sex may moderate cocaine-induced incentive sensitization in CC mice.

Relationship to Dopamine Transmission

Genetically determined differences in incentive sensitization following cocaine may be mediated by dopamine transmission in the NAc. In the present study, we demonstrated that both dopamine transmission in drug-naïve subjects and in those exposed to chronic cocaine are dependent on genetic background. Furthermore, the strain (CC018/UncJ) that demonstrated elevated dopamine transmission under drug-naïve conditions, as well as dopamine blunting following cocaine exposure, also demonstrated incentive sensitization following cocaine. Following chronic cocaine exposure, reduced basal dopaminergic states and blunted dopamine responses to cocaine are often observed (Ferris et al., 2013; Homberg et al., 2003; Lack et al., 2008; McCutcheon et al., 2009). Reduced dopamine transmission following chronic cocaine may facilitate escalation of cocaine taking (Willuhn et al., 2014). Individual variation in the neuroadaptations affecting dopamine signaling following cocaine exposure may play a role in behavioral responses to cocaine, including self-administration (Ferris et al., 2013; Homberg et al., 2003). The present data suggest that similar, genetically determined dopamine neuroadaptations associate with differential responses to cocaine-induced incentive sensitization. In addition to neuroadaptations following cocaine exposure, variation in dopaminergic transmission that precede cocaine exposure may influence responses to cocaine and subsequent risk for addiction-like behaviors (Dalley et al., 2007; Hitzemann et al., 2003; Janowsky et al., 2001; Merritt & Bachtell, 2013; Nader & Czoty, 2005; Nelson et al., 2009; Shimamoto et al., 2015; Volkow et al., 1997, 1999), and this variation may be attributable to genetics, at least in part. For examples, heritable variation in dopamine transporter density is genetically correlated with locomotor responses to acute cocaine (Janowsky et al., 2001). The strain differences in dopamine responses we observed in cocaine-naïve CC018/UncJ and CC027/GeniUncJ mice, may mediate strain differences in behavior during the magazine training sessions, as well as the degree of CR responding observed in cocaine-naïve or -exposed mice.

Conclusions

Ultimately, this study marks one of the first efforts to connect psychomotor sensitization and incentive sensitization in two inbred mouse strains that exhibit widely divergent patterns of psychomotor sensitization. The emergence of the sensitization phenotype is furthermore associated with strain differences in cocaine-induced dopaminergic neuroadaptations in the NAc. Additionally, susceptibility to psychomotor and incentive sensitization in drug naïve subjects may be driven by heritable variation in NAc dopamine release, indicating a shared mechanism for individual sensitization variability.