Abstract
Behavioral sensitization to cocaine reflects neuroadaptive changes that intensify drug effects. However, repeated cocaine administration does not induce behavioral sensitization in all male Sprague-Dawley rats. Because cocaine inhibits the dopamine (DA) transporter (DAT), we investigated whether altered DAT function contributes to these individual differences. Freely moving rats had electrochemical microelectrode/microcannulae assemblies chronically implanted in the nucleus accumbens so that exogenous DA clearance signals were recorded simultaneous with behavior. The peak DA signal amplitude (Amax) and efficiency of clearance (k) were used as indices of in vivo DAT function. Low and high cocaine responders (LCRs and HCRs, respectively) were identified based on their locomotor responsiveness to an initial injection of cocaine (10 mg/kg i.p.). Consistent with DAT inhibition, cocaine elevated Amax and reducedk in HCRs, but not in LCRs. The same dose of cocaine was administered for six additional days and after a 7-day withdrawal. Baseline behavioral and dopamine clearance indices were unaltered by repeated cocaine or after withdrawal. Only LCRs expressed cocaine-induced sensitized locomotor activation, and this was accompanied by cocaine-induced elevations inAmax and reductions in k. These sensitized responses to cocaine persisted in LCRs after withdrawal. In contrast, neither locomotor nor electrochemical responses were altered by repeated saline administration or a saline challenge after repeated cocaine administration, suggesting that conditioning did not significantly contribute. Our results suggest that increased DAT inhibition by cocaine is associated with locomotor sensitization and that DAT serves as a common substrate for mediating both the initial and sensitized locomotor responsiveness to cocaine.
Repeated administration of psychomotor stimulants, such as cocaine and amphetamine, has been used as an experimental paradigm to model the progression of behavioral and neurochemical changes leading to compulsive drug use in addicts. Repeated stimulant administration often results in behavioral sensitization, a progressive increase in responsiveness to drug. Sensitization in rats is manifested as augmented locomotor activity and stereotypic behaviors, as well as self-administration and drug-seeking behaviors (Robinson and Berridge, 1993; Covington and Miczek, 2001; De Vries et al., 2001; Vezina et al., 2002). Thus, it is important to understand the neurobiological changes induced by repeated stimulant administration because they likely contribute to intensification of drug motivation and/or craving that promote relapse in addicts (Robinson and Berridge, 1993).
Cocaine potentiates dopamine (DA) signaling in mesocorticolimbic reward pathways by inhibiting uptake of DA by the DA transporter (DAT). With repeated administration, the responsiveness of DA systems to stimulants is enhanced in behaviorally sensitized rats (Vanderschuren and Kalivas, 2000). Transient increases in somatodendritic DA release and spontaneous firing activity of DA neurons in the ventral tegmental area have been associated with the induction of behavioral sensitization, whereas long-lasting alterations within DA neuronal terminal fields have been associated with its expression (Vanderschuren and Kalivas, 2000; Nestler, 2001; Everitt and Wolf, 2002). Initially, reduced sensitivity of inhibitory DA D2-autoreceptors may result in the potentiated cocaine-induced increases in extracellular DA concentrations in nucleus accumbens (NAc; Vanderschuren and Kalivas, 2000). However, the D2-autoreceptor changes seem to be transient, whereas the ability of a cocaine challenge to produce a greater augmentation in NAc DA levels persists. This suggests that other mechanisms must underlie the long-lasting changes in DA neuronal function.
Because cocaine blocks DAT, repeated cocaine may regulate DAT expression/activity. However, whether it does is controversial (Izenwasser and Cox, 1990; Ng et al., 1991; Cass et al., 1993a;Masserano et al., 1994; Meiergerd et al., 1994; Zahniser et al., 1995;Kuhar and Pilotte, 1996; Zahniser and Doolen, 2001; Chefer and Shippenberg, 2002). Discrepancies may result from differences in strains/individuals, drug administration/withdrawal paradigms, and/or assays used. We have combined high-speed chronoamperometry with local applications of DA to measure dynamic changes in extracellular DA concentrations after acute and repeated cocaine administration (Cass et al., 1993a; Zahniser et al., 1999; Sabeti et al., 2002b). Using this approach in striatum of drug-naı̈ve rats, we have demonstrated that the decay or clearance of locally applied DA primarily reflects in vivo DAT activity and that diffusion makes a more minor contribution to cessation of the DA signals (Cass et al., 1993b; Sabeti et al., 2002a,b). Furthermore, we observed cocaine-induced changes consistent with behavioral sensitization: greater cocaine-induced inhibition of exogenous DA clearance in NAc of anesthetized rats withdrawn from repeated cocaine administration, compared with rats given the same dose acutely (Cass et al., 1993a). This finding was consistent with reports of decreased DAT binding sites (but see Cass et al., 1993a; Pilotte et al., 1994; Wilson et al., 1994; Boulay et al., 1996; Letchworth et al., 1999) and increased cocaine potency (Lee et al., 1998) after repeated cocaine treatment. However, because the rats were anesthetized for the clearance measurements (Cass et al., 1993a), it was impossible to relate cocaine-induced regulation of DAT function with individual behavioral changes. This relationship was further complicated by the fact that not all male Sprague-Dawley rats exhibit behavioral sensitization with repeated cocaine administration (Cass et al., 1993a).
Greater locomotor activity in a novel environment predicts enhanced vulnerability to behavioral sensitization induced by repeated amphetamine (Bardo et al., 1996; Piazza and Le Moal, 1996; Cools and Gingras, 1998), although this relationship may not generalize to cocaine (Djano and Martin-Iverson, 2000; Sutton et al., 2000). High-novelty responding rats also exhibit higher DA release (Bardo et al., 1996; Piazza and Le Moal, 1996; Cools and Gingras, 1998) and firing rates of ventral tegmental area DA neurons (Marinelli and White, 2000). We have found that greater acute inhibition of DA clearance is correlated with high initial locomotor responsiveness to cocaine (Sabeti et al., 2002b). However, it is unclear whether this association extends to cocaine-induced locomotor sensitization.
To determine whether long-lasting adaptations in DAT function contribute to cocaine-induced locomotor sensitization, we simultaneously recorded behavior and DA clearance in NAc of freely moving rats over a 2-week period. We then 1) compared the time course of changes in behavior and DA clearance and 2) correlated individual variations in the magnitude of sensitization with initial cocaine-induced locomotor activation and with basal and cocaine-induced changes in DA clearance.
Materials and Methods
Animals.
Outbred male Sprague-Dawley rats were obtained from Charles Rivers Laboratory (Sasco, Omaha, NE). Before surgeries, rats were housed no more than six per cage with a 12-h light/dark cycle and unrestricted access to food and water. To habituate rats to handling and the insertion of injector tubing on experimental days, rats were handled for 1 to 2 days before surgery and before each recording session. After surgery, rats were housed individually. All animal care procedures followed the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at the University of Colorado Health Sciences Center.
Stereotaxic Implantation of Recording Microelectrode/Microcannulae Assemblies.
Procedures for the construction and implantation of recording microelectrode/microcannulae assemblies have been previously described in detail (Gerhardt et al., 1999; Sabeti et al., 2002a,b). Briefly, microelectrodes were fabricated from a single carbon fiber (fiber diameter, 30 μm; exposed length, 150–300 μm; Textron Systems, Wilmington, MA) that was coated with Nafion (5% solution; Aldrich Chemical Co., Milwaukee, WI). They were calibrated in vitro as described previously. Microelectrodes displayed a DA over ascorbic acid selectivity of ≥1000:1 and linear responses to DA (1–6 μM) in vitro. Each microelectrode was assembled onto two stainless steel guide microcannulae, which permitted the delivery of KCl or DA from injectors that were inserted before each recording session. Injectors were fabricated from fused silica tubing (40 μm i.d. × 150 μm o.d.; Polymicro Technologies, Phoenix, AZ). Insertion of the injector through the guide positioned the tip of the injector precisely within 250 to 350 μm from the exposed tip of the carbon fiber microelectrode.
For implantation of the microelectrode/microcannulae assembly, rats were deeply anesthetized with chloral hydrate, as described previously (Sabeti et al., 2002b). The assembly was then lowered stereotaxically into the core of the NAc (anterior-posterior, +1.2–1.5 mm from bregma; medial-lateral, 2.2 mm left from the midline; dorsal-ventral, 6.5–7.5 mm below surface; Sabeti et al., 2002b). A Ag/AgCl reference microelectrode (0.011-inch diameter; A-M Systems, Carlsborg, WA) was implanted into the posterior cerebral cortex. Leads from the recording and reference microelectrodes were soldered to a four-pin modular telephone connector and encased in heat shrink tubing (flexible polyolefin; 1/16 inch expanded; 1/32 inch recovered; JT&T Products, San Jose, CA). To ensure that the recording microelectrode was situated at a site densely innervated by DA terminals, as opposed to the anterior commissure, KCl (120 mM in 29 mM NaCl and 2.5 mM CaCl2, pH 7.4; 300–800 nl) was infused through the injector to stimulate DA release, which was measured using high-speed chronoamperametry (see below; Hebert and Gerhardt, 1998;Sabeti et al., 2002a,b). Once an appropriate recording microelectrode placement was verified, the entire assembly was cemented in place using dental cement, a thick layer of Quick Set Epoxy (Duro; Loctite Corporation, Rocky Hill, CT), and five small screws in the skull as anchors. Except during recording sessions, dummy injectors were inserted through the guide microcannulae to prevent obstruction.
Treatment Protocol and Timeline of Simultaneous Behavioral and Electrochemical Recordings.
Three to five days after implantation of the microelectrode/microcannulae assembly, each rat was transferred from its home cage to an open field activity apparatus (16 × 16 × 15 inches; San Diego Instruments, San Diego, CA; enclosed inside a 2 × 2 × 2-foot Faraday cage). On this day (day 0), initial behavioral and electrochemical responses in the activity apparatus and to a saline injection (1 ml/kg i.p.) were examined in all of the rats. Subsequently, rats were divided into two groups (Table1). On days 1 to 6, once daily injections of either saline (control group) or (−)-cocaine HCl (10 mg/kg i.p.; experimental group) were administered. On day 7, both groups were injected with cocaine. Rats in the control group continued to be injected with cocaine for five more days and then were injected with saline on the 7th day. After a 7-day withdrawal in the experimental group, cocaine was administered on day 15. All injections were given in the activity apparatus. The data on days 0 and 1 from rats in the experimental group have been previously presented (Sabeti et al., 2002b) and are included in the mean values reported here.
On recording days (days 0, 1, 3, 5, 7, and 15; Table 1), rats were acclimated to the activity apparatus for 1 h. During this period, rats were handled momentarily while new injectors were inserted through the guide microcannulae in preparation for the repeated application of DA into NAc. Next, “baseline” measurements of behavior and DA clearance signals were recorded for 30 min immediately before the i.p. injection of either saline or cocaine. Thus, rats had acclimated to the activity apparatus for a total of 1.5 h before the saline or cocaine injection. After injection, data were collected for an additional 60 min. Room lights were on throughout the experiment. Injectors were removed at the end of each recording session, dummy injectors were reinserted and rats were returned to their home cages. On all other treatment days, rats were immediately transferred to their home cages after injection.
Only rats with electrochemical assembly localizations in NAc were included in the results reported here. Placement was verified at the end of the experiment by visual inspection when removing the assembly from the brain of euthanized rats. In a limited number of rats, the location of the assembly was further confirmed by microscopic examination of coronal sections stained with cresyl violet (Sabeti et al., 2002a).
Behavioral Data Acquisition.
Automated recordings of locomotor activity were obtained in the open field apparatus using a single photo beam frame (eight beams per dimension) near the base of the apparatus. Consecutive beam interruptions were converted to distance traveled (centimeters) per 5-min period. Head/limb stereotypy was defined as repetitive head movements, including head bobs and side-to-side head sways, or back and forth repetitive forelimb movements. The frequency of stereotypy was determined by observation and expressed as the fraction of time during each 15-min interval in which the behavior was exhibited, as previously described in detail (Sabeti et al., 2002b). Rearing was the cumulative number of times within each 15-min interval during which both forepaws were lifted and then at least one forepaw was placed back onto the floor.
High-Speed Chronoamperometry in Freely Moving Rats.
Upon transfer to the activity apparatus, the rat was connected to a miniature potentiostat headstage/tether (RAT HAT; Quanteon, L.L.C., Lexington, KY) via a four-pin telephone connector mounted on the rat's head. This allowed the rat free movement within the activity apparatus (Gerhardt et al., 1999; Sabeti et al., 2002a,b). The headstage/tether was linked to an IVEC-10/FAST-12 electrochemical recording system (Quanteon, L.L.C.). Continuous 100-ms square-wave potential pulses (0.0 to +0.55 V versus Ag/AgCl reference) were applied at 5 Hz to the recording microelectrode. A stable background oxidation signal was established in the absence of exogenous DA and set to zero. Subsequently, exogenous DA (40–300 pmol in saline containing 100 μM ascorbic acid, pH 7.4 adjusted with sodium hydroxide) was applied at 5-min intervals into NAc, using a microprocessor-controlled syringe-pump (1.01 μl/s; Stoelting Co., Wood Dale, IL; Gerhardt et al., 1999; Sabeti et al., 2002a,b). During each recording session, the DA ejection volume was initially adjusted for each rat to evoke baseline peak DA signal amplitude (Amax) responses within a range of 0.3 to 1.2 μM. Once three reproducible (i.e., within 20%)Amax responses were elicited by the same ejection volume, DA was applied once every 5 min at this constant amount throughout the remainder of the recording session for that day.
High-frequency spike artifacts in the DA signals were digitally filtered (cutoff frequency >0.028 Hz), as previously described in detail (Sabeti et al., 2002a). Amaxresponses were determined from the peak of the DA signal amplitudes using in vitro electrode calibration data to convert oxidation currents, averaged over 1-s epochs, to micromolar concentrations above the background. The efficiency of DA clearance was determined by fitting the decay segment of each DA signal to a single monoexponential decay function [A(t) =Amax ·e−k(t−t0); where A is the amplitude of the DA signal (micromolar) at any time t (seconds) afterAmax;t0 is the time at which Ahad decayed to approximately 80% ofAmax; and k is the first-order decay rate constant (per seconds); Sabeti et al., 2002a].R2 values for the exponential curve fits to the smoothed data ranged from 0.8999 to 0.9966. At the low picomolar amounts of DA applied here, k reflects theVmax/Kmratio, or efficiency of DA clearance, according to the Michaelis-Menten kinetic model of uptake (Sabeti et al., 2002a).
Statistical Analysis.
Data are expressed as mean values + S.E.M. Statistical analyses, including one- and two-way analysis of variance (ANOVA) and Pearson correlation analysis, were performed using either SigmaStat (SPSS Science, Chicago, IL) or Prism (GraphPad Software, Inc., San Diego, CA) software. A level of p< 0.05 was considered statistically significant.
Chemicals and Drugs.
Dopamine and other chemicals were purchased from Sigma-Aldrich (St. Louis, MO). (−)-Cocaine HCl was obtained from the National Institute on Drug Abuse (Research Triangle Institute International, Research Triangle Park, NC).
Results
Individual Differences in Cocaine-Induced Behavioral Sensitization Are Associated with Initial Locomotor Responsiveness to Cocaine.
Previously we demonstrated that outbred male Sprague-Dawley rats can be effectively divided into two subgroups of cocaine responders, namely, low or high cocaine responders (LCRs or HCRs, respectively), using the median split of locomotor activity during the initial 30 min after an acute i.p. injection of 10 mg/kg cocaine (Sabeti et al., 2002b). In the follow-up longitudinal cocaine study reported here, a subset of rats from this previously characterized population, the group that had electrochemical microelectrode/microcannulae assemblies implanted chronically in NAc, continued to receive daily cocaine treatment. Thus, these rats received an injection of saline (1 ml/kg i.p.) on day 0 and daily injections of cocaine (10 mg/kg i.p.) on days 1 to 6 (experimental group; Table 1). Rats were subsequently challenged with an i.p. injection of the same dose of cocaine following a 24-h withdrawal (i.e., day 7) and a 7-day withdrawal (i.e., day 15) from the daily cocaine treatment.
We hypothesized that individual variability in initial responsiveness to cocaine might influence whether robust sensitized responses in behavioral activation and DA clearance inhibition were induced by the repeated cocaine treatment. To test this hypothesis, this subset of rats was reprofiled as either LCRs or HCRs. In this group, the median split of the distance traveled after the initial cocaine injection on day 1 was 7200 cm/30 min, resulting in 10 LCRs with a mean activity of 3800 ± 888 cm/30 min and 7 HCRs with a mean activity of 18100 ± 2910 (Fig. 1). As we previously observed in the larger population (Sabeti et al., 2002b), there was no correspondence between baseline activity in the 30 min preceding the cocaine injection and the cocaine-induced activity on day 1 (Fig. 1).
Before examining relationships between cocaine-induced alterations in behavior and DA clearance, first the time courses of the locomotor responses were compared within the two groups across all of the recording days (Fig. 2, A and B, left, LCRs and HCRs, respectively). All rats were fully acclimated to the activity apparatus before initiating the daily cocaine treatment. This was demonstrated on day 0 by the low levels of baseline activity in both LCRs and HCRs during the 30 min before and the 60 min after injection of saline. It should also be noted that these low levels of activity were observed in both groups despite the local applications of DA into NAc every 5 min. Furthermore, baseline locomotor activity did not change significantly over the 7 days of cocaine treatment or after the 7-day withdrawal in either LCRs or HCRs.
Cocaine-stimulated locomotor activity was analyzed over the 60 min after drug injection on days 1, 3, 5, 7, and 15 using a two-way ANOVA, with repeated measures on both factors (days1,3,5,7,15 × time0–60min). An overall significant effect of days was observed for cocaine-stimulated activity in LCRs (Fig. 2A, left; F4,402 = 7.102,p < 0.001), but not in HCRs (Fig. 2B, left). Further analysis showed that sensitized locomotor responses to cocaine (versus day 1) were expressed in LCRs only after day 3 of daily cocaine treatment and that they persisted in response to a cocaine challenge on day 15 after the 7-day withdrawal (Fig. 2A, left). The maximal effect of cocaine on locomotor activity was summarized for the two groups by averaging responses during the first 30 min after cocaine injection during the induction (i.e., days 1–3) and expression (i.e., days 5–7) of locomotor sensitization. These effects were compared with the saline-induced locomotor response over the same time interval on day 0 and to the cocaine-induced response on day 15 (Fig. 2, right). This analysis showed that during induction of sensitization in LCRs, cocaine-stimulated locomotor activity was not significantly different from saline-induced activity (Fig. 2A, right). However, by days 5 to 7 cocaine-stimulated activity was significantly potentiated by 400% above the cocaine-stimulated response during induction. After the 7-day withdrawal from the repeated treatment in LCRs, the cocaine-stimulated locomotor activity remained significantly augmented on day 15 versus days 1 to 3 and the saline-induced response. Interestingly, in HCRs cocaine-stimulated locomotor activity on days 1 to 3 and days 5 to 7 did not differ significantly (Fig. 2B, right) but was similar in magnitude to the sensitized responses in LCRs (Fig. 2A, right). The 7-day withdrawal from repeated cocaine produced no further augmentation in locomotor responses of HCRs to cocaine (Fig. 2B, right).
To determine whether cocaine-induced locomotor sensitization in individual rats was correlated with initial locomotor responsiveness to cocaine, the magnitudes of locomotor sensitization in LCRs on day 7 were plotted against their cocaine-induced locomotor activity on day 1 (Fig. 3). The magnitude of sensitization was defined as the ratio of each animal's day 7: day 1 cocaine-stimulated locomotor activity during the first 30 min after injection. Thus, a ratio in the range of 0 to 1 indicated a lack of cocaine-induced behavioral sensitization, whereas progressively higher values >1 were indicative of increasing levels of sensitization. In LCRs the magnitude of locomotor sensitization was robust, ranging from 2.1- to 21-fold increases over the initial cocaine-stimulated locomotor activity; and there was a significant inverse correlation between the magnitudes of initial locomotor responsiveness to cocaine and locomotor sensitization (Pearson r = −0.6832, p< 0.05; Fig. 3). Because locomotor sensitization was either absent or relatively modest in HCRs, with ratios varying from 0.3 to 1.7, only data from LCRs were included in the correlational analysis shown in Fig. 3. However, independent of the LCR/HCR classification, a significant inverse relationship was observed between the magnitudes of initial locomotor responsiveness to cocaine and locomotor sensitization in all of the rats studied (Pearson r = −0.5905, p < 0.05; n = 17).
We also examined whether cocaine-induced sensitization was manifested in other behaviors, in particular those which may have competed for the expression of locomotor sensitization. Therefore, the effects of the repeated cocaine treatment were examined on cocaine-stimulated head/limb stereotypy and rearing behaviors (Fig.4). In LCRs, cocaine-induced head/limb stereotypy seemed to increase progressively with repeated treatment and this persisted after withdrawal; however, a statistically significant effect of days was not found (Fig. 4A, left). Cocaine-induced rearing responses in LCRs were increased on all days tested; therefore, there was not a significant effect of days (Fig. 4A, right). In contrast, although head/limb stereotypy was displayed in HCRs to the same extent across all cocaine treatment days, cocaine-induced rearing responses were progressively and significantly augmented by the repeated cocaine treatments (days1–3,5–7,15:F2,72 = 13.7, p < 0.0001; Fig. 4B). Thus, HCRs did exhibit behavioral sensitization of rearing, but not locomotor, responses.
Individual Differences in Cocaine-Induced Locomotor Sensitization Are Associated with Differential Cocaine-Induced Modulations in DA Clearance in NAc.
Simultaneous with the behavioral measurements, changes in DA clearance in NAc of all the rats were continuously monitored by high-speed chronoamperometry on days 0, 1, 3, 5, and 7 (Fig. 5). Additionally, the effects of the 7-day withdrawal and challenge by cocaine were evaluated on DA clearance in a subset of these rats whose microelectrode/microcannulae assemblies remained patent on day 15 (n = 7, LCRs;n = 4, HCRs). A constant amount of DA (40–300 pmol; subsaturating relative to the maximal clearance rates measured in vivo;Sabeti et al., 2002a) was applied at the recording site every 5 min. These DA applications evoked reproducible oxidation signals that were measured over a 30-min recording interval to obtain the predrug baseline value on each recording day (Fig. 5). Changes over baseline in the Amax and k parameters of the DA signal were used as indices of altered DAT function (Fig. 5). On each recording day amounts of DA locally applied were initially adjusted for each rat so that similar baselineAmax responses were achieved. Thus, mean baseline Amax values did not differ between LCRs and HCRs; these averaged 0.8 ± 0.4 and 0.7 ± 0.1 μM, respectively, over the 2-week period in which rats were treated repeatedly and then withdrawn from cocaine. The amounts of DA applied relative to Amaxvalues at baseline were averaged across days 1 and 3 and across days 5 and 7 to correspond to the induction and expression phases of locomotor sensitization, respectively, in the LCRs (see above). Initially (i.e., days 1–3) HCRs required 3-fold significantly greater amounts of DA than LCRs to elicit similar Amaxresponses (Fig. 6A). This finding is in agreement with our previous observations on day 1 in the same LCR and HCR rats (Sabeti et al., 2002b). Furthermore, although DA applications in LCRs did not differ significantly across the repeated treatment days or after withdrawal, in HCRs significantly lower amounts were necessary on days 5 to 7 than on days 1 to 3 to achieve comparableAmax responses (Fig. 6A). Despite the differences in the amounts of DA applied, predrug baseline kvalues (0.016–0.023 s−1) were not significantly different between LCRs and HCRs across the repeated cocaine treatment days and after withdrawal (Fig. 6B). This finding is in agreement with our observation that k is independent of applied DA within the relatively narrow range of DA applied here, whereas higher amounts of exogenous DA modulate the k for DA clearance as expected by active uptake kinetics (Sabeti et al., 2002a). Furthermore, predrug baseline k values were not significantly correlated with initial locomotor responsiveness to cocaine in individual rats (p = 0.673).
In contrast to the modest changes in baseline DA clearance efficiency, robust alterations in cocaine-induced inhibition of DA clearance in NAc closely paralleled the time course of locomotor sensitization induced by the repeated cocaine administration. For example, in LCRs there was an overall significant effect of days on the cocaine-induced increases in Amax (Fig.7A, left;F4,314 = 4.101, p < 0.05), as revealed by a two-way ANOVA (days1,3,5,7,15 × time0–60min; time as the only repeated measure). This effect was not observed in HCRs (Fig. 7B, left), which also did not exhibit locomotor sensitization to cocaine (Fig. 2B). The effects of cocaine on Amax were summarized in LCRs and HCRs over the first 30 min after injection to correspond to the maximal effects of cocaine on locomotor activity (Fig. 2) and compared with the effect of saline on day 0 (Fig. 7, right). Specifically, on days 1 to 3 cocaine did not significantly alterAmax responses in LCRs, compared with saline (Fig. 7A, right). However, during the expression of locomotor sensitization on days 5 to 7 in LCRs, cocaine significantly potentiated the Amax response by 46 ± 12%, compared with both days 1 to 3 and saline.Amax responses were potentiated to 31 ± 11% by cocaine on day 15 after the 7-day withdrawal, although this effect was not significantly different compared with either days 1 to 3 or saline. As with locomotor activation, in HCRs there was no overall significant effect of days on cocaine-induced increases in Amax (Fig. 7B, right). In contrast to LCRs, cocaine administration on days 1 to 3 significantly potentiated Amax by 51 ± 18%, compared with saline. Although there was also a trend forAmax to be increased by cocaine on days 5 to 7 (33 ± 11%) and day 15 (23 ± 19%), these effects did not reach statistical significance versus the saline response.
Similar to cocaine-induced increases inAmax values in LCRs, an overall significant effect of days was observed on cocaine-induced reduction in the k for DA clearance in LCRs (Fig.8A, left;F4,304 = 3.396, p < 0.05). Specifically, on days 1 to 3 during the induction of locomotor sensitization, k was not significantly altered by cocaine, compared with saline on day 0 (Fig. 8A, right). However, during the expression of locomotor sensitization on days 5 to 7 in LCRs, cocaine significantly attenuated k by 24 ± 5%, versus days 1 to 3 and saline. This greater effect of cocaine on kpersisted on day 15 after the 7-day withdrawal from repeated cocaine treatment. Interestingly, in HCRs the cocaine-induced reductions ink on days 1 to 3 and days 5 to 7 did not differ significantly from each other but were of a similar magnitude as the sensitized responses in LCRs (Fig. 8B, right). The cocaine-induced reduction in k persisted, but did not change in magnitude, after the 7-day withdrawal from repeated cocaine (Fig. 8B, right). These results for cocaine-induced changes in the efficiency of DA clearance in LCRs versus HCRs are in agreement with both the cocaine-induced changes in locomotor activity and the DA clearance signal Amax responses.
Conditioning Does Not Contribute Significantly to Sensitized Behavioral and Electrochemical Responses to Cocaine Measured Here.
To control for the passage of time and any potential conditioned responses to the repeated injections of cocaine and/or local applications of DA, another group of rats with electrochemical assemblies chronically implanted in NAc received repeated saline injections for 7 days before an acute cocaine challenge injection (Table 1, control group). Furthermore, this same group subsequently received cocaine injections on five additional days, followed by a saline challenge injection on the 7th day (Table 1). Data collected in the experimental group on the day 7 cocaine challenge were averaged across the LCR and HCR responses and reanalyzed for significant differences from the control group. These comparisons are summarized in Fig. 9. Baseline locomotor activity during the 30 min immediately preceding either the cocaine or saline challenge injections did not differ significantly between the treatment groups, confirming that differences in locomotor responsiveness to the challenges were not due to differences in levels of baseline activity per se. The cocaine challenge on day 7 significantly increased locomotor activity above baseline in both the saline-pretreated control group and cocaine-pretreated experimental group. However, this increase was significantly greater (by 80%) in the cocaine-pretreated, compared with saline-pretreated, rats. This result confirmed that, independent of the LCR/HCR classification, the repeated cocaine regimen used here was effective in inducing locomotor sensitization. In contrast with the cocaine challenge after repeated saline treatment in the control group, the saline challenge after repeated cocaine treatment in these same rats did not increase locomotor activity above baseline. Thus, there was no apparent conditioned locomotor response to this particular repeated cocaine regimen.
Electrochemical responses in the control and experimental groups were analyzed in a similar manner to the behavioral responses (Fig.10). On averageAmax responses were increased by 42 ± 6% above baseline after the cocaine challenge in the cocaine-pretreated experimental group (Fig. 10A). This effect was significantly greater, but only by 25%, compared with the effect of the cocaine challenge in the saline-pretreated control group (Fig.10A). Importantly, the potentiation of Amaxresponses in cocaine-pretreated rats was expressed only after a cocaine, but not a saline, challenge injection (Fig. 10A). Also consistent with DAT inhibition, k was decreased by 24 ± 3% below baseline after the cocaine challenge in the cocaine-pretreated experimental group (Fig. 10B). This was a significant reduction compared with the control group, with respect to the effect of both the cocaine challenge after repeated saline administration and the saline challenge after repeated cocaine administration (Fig. 10B). Overall, there was good concordance between the challenge-induced changes in locomotor activity and DA clearance parameters.
Discussion
Whether DAT expression/activity is up- or down-regulated by repeated cocaine administration remains controversial (Zahniser et al., 1995; Kuhar and Pilotte, 1996; Zahniser and Doolen, 2001). Furthermore, the exact relationship between cocaine-induced adaptations in DAT function and changes in behavior has remained elusive. Previously, we found that DA clearance in NAc of anesthetized rats was more sensitive to cocaine inhibition after withdrawal from repeated cocaine administration (Cass et al., 1993a). Here, by monitoring the time course of cocaine-induced changes in behavior concomitantly with inhibition kinetics of DA clearance, we demonstrate the essential role of greater inhibition of DAT function to locomotor sensitization induced by repeated cocaine administration. Specifically, we found that the potential for expression of locomotor sensitization 1) can be predicted by the animal's initial locomotor responsiveness to an acute cocaine injection and 2) is confined to a subgroup of rats (LCRs) in which the efficiency to clear exogenous DA in NAc in the presence of cocaine diminishes with repeated administration.
Previously, we identified two distinct populations of cocaine responders using the median split of initial locomotor responses to acute low-dose cocaine in male Sprague-Dawley rats (Sabeti et al., 2002b). Subsequently, we have confirmed a trend for two components in this distribution (LCR component, mean 5,130 cm/30 min; S.D. = 2,140; proportion = 49%; HCR component, mean = 12,370; S.D. = 4,950; proportion = 51%; n = 32;p = 0.07; NOCOM program; Ott, 1979). Likewise, the distribution of Amax responses to acute cocaine in this larger population was bimodal (LCR component, mean 98% baseline; proportion = 58%; HCR component, mean 170% baseline; proportion = 42%; p < 0.01). Here, using a subset of these rats, we observed that differences in the initial locomotor responsiveness to cocaine accounted for nearly 47% of the variability in the magnitude of locomotor sensitization expressed by LCRs (Fig. 3). Furthermore, the LCR/HCR classification was predictive of the potential of individual rats to express cocaine-induced locomotor sensitization. LCRs exhibited minimal locomotor activation to the initial injection of cocaine but locomotor sensitization with repeated cocaine, whereas HCRs exhibited significant initial locomotor responsiveness to cocaine but failed to express locomotor sensitization with repeated administration. However, not all cocaine-induced behaviors were augmented in LCRs. Furthermore, HCRs exhibited sensitized cocaine-induced rearing responses. These observations, together with the finding that brain levels of cocaine are consistently increased in all male Sprague-Dawley rats receiving repeated i.p. injections of cocaine (Cass and Zahniser, 1993), suggest that pharmacokinetic differences can not satisfactorily explain the LCR versus HCR behavioral differences. It is also unlikely that increased rearing or head/limb stereotypy explained the lack of locomotor sensitization in HCRs because stereotypies were exhibited by LCRs to a similar extent on days 5 to 7 and did not preclude locomotor sensitization in these rats. However, we cannot rule out the possibility that a ceiling in locomotor activation precluded sensitization in HCRs because only the 10-mg/kg dose of cocaine was tested here. Dose-response studies would address whether HCRs are able to increase activity further and whether repeated cocaine administration shifts the dose-response relationship in LCRs, making cocaine a more potent or efficacious inhibitor of DAT. In any case, conditioned locomotor responses to the repeated injections were not apparent with our treatment paradigm, supporting earlier findings that the development of conditioned responses is not necessary for the expression of behavioral sensitization (Fraioli et al., 1999).
Greater activation in response to novelty, rather than to acute stimulant administration, has been used more often to predict enhanced vulnerability to psychostimulant-induced sensitization and self-administration (see Introduction). We observed the opposite relationship: HCRs exhibit higher spontaneous locomotor activity than LCRs during their initial exposure to the activity apparatus (i.e., day 0; Sabeti et al., 2002b) but did not express locomotor sensitization with repeated cocaine administration (this study). This lack of correspondence in expression of sensitization between our cocaine responders and the low- and high-novelty responders defined in the literature may reflect differences in experimental conditions. For instance, our rats underwent extensive handling and habituation to the testing environment before drug administration, factors previously documented to modulate drug responsiveness (Cools and Gingras, 1998;Fraioli et al., 1999; Tuinstra and Cools, 2000). Specifically, habituation can decrease the behavioral and NAc DA sensitivity to stimulants in high-novelty responders and increase sensitivity in low-novelty responders (Cools and Gingras, 1998; Tuinstra and Cools, 2000). These investigators have hypothesized that individual differences in the regulation of DA neurotransmission by neuroendocrine and/or noradrenergic systems underlie this reversal in sensitivity under various experimental conditions. On the other hand, the discrepancy may reflect differences underlying responsiveness to novelty and cocaine (Djano and Martin-Iverson, 2000; Sutton et al., 2000). The finding that unique provisional quantitative trait loci exist for novelty- versus cocaine-induced initial locomotor activity and sensitization (Phillips et al., 1998) further supports our hypothesis that HCRs may be phenotypically distinct from the high responders to novelty. Studies using self-administration and/or conditioned place preference paradigms are needed to define the relationship of the LCR/HCR phenotypes and cocaine reinforcement.
The temporal association between changes in behavior and DAT function in response to repeated cocaine (Figs. 2, 7, and 8) provides strong evidence that cocaine-induced regulation of DAT in NAc plays a critical role in the expression of locomotor sensitization to cocaine. This association reflected multiple recordings of exogenous DA clearance signals across the 7 days of repeated cocaine in the same individual rats. For example, initially on days 1 to 3, when locomotor sensitization was not yet expressed in LCRs,Amax and k parameters in NAc were not modulated by cocaine. The expression of cocaine-induced locomotor sensitization in LCRs on days 5 to 7 was, however, accompanied by potentiated Amax and reduced k, consistent with higher levels of extracellular DA and sensitized DA-related behaviors. Importantly, the sensitized effects of cocaine on behavior and DA clearance parameters in LCRs persisted after a 7-day withdrawal from repeated cocaine treatment, consistent with the long-lasting nature of sensitization. The absence of such regulation in HCRs and the control group ruled out a more general effect of the repeated DA applications on increased DAT sensitivity to cocaine. Overall, our findings are in agreement with reports of enhanced cocaine-induced inhibition of DA uptake after repeated cocaine administration (Izenwasser and Cox, 1990; Cass et al., 1993a; Lee et al., 1998; but see Ng et al., 1991; Masserano et al., 1994; Meiergerd et al., 1994; Chefer and Shippenberg, 2002).
The finding that cocaine-induced inhibition of DA clearance in LCRs was expressed only after 3 days of repeated cocaine administration strongly suggests that protein synthesis and/or recruitment of other systems was/were required for the long-term alterations in DAT sensitivity to cocaine in LCRs. On the other hand, LCR/HCR differences in the acute cocaine response may reflect intrinsic variations in more rapid, nongenomic mechanisms of DAT regulation in response to acute inhibition (i.e., cell surface trafficking; Daws et al., 2002; Little et al., 2002). Also, although the NAc is sufficient for mediating the initial locomotor response to acute cocaine (Delfs et al., 1990), a number of long-lasting neuroadaptations in NAc, as well as other DA projection sites, are necessary for the expression of stimulant-induced locomotor sensitization (Vanderschuren and Kalivas, 2000; Nestler, 2001; Everitt and Wolf, 2002). Whether and how glutamatergic and/or GABAergic systems are involved in the LCR/HCR differences in sensitization remains to be investigated. Furthermore, future studies will address whether sensitization in cocaine-induced rearing, as opposed to locomotor activity, in HCRs reflects alterations in DAT sensitivity to cocaine inhibition in the dorsal striatum, which plays an important role in stereotypic behaviors.
On day 1 greater amounts of exogenous DA were applied in NAc of HCRs than LCRs to achieve similar Amaxresponses (Sabeti et al., 2002b; present study). Here, we demonstrated that this group difference was surmounted after day 3, suggesting that the initial difference likely reflected higher baseline DA clearance capacity in NAc of HCRs than LCRs, rather than variability in injector locations. Although this is discordant with the observed higher initial sensitivity to cocaine inhibition, future kinetic experiments will address whether lower DAT affinity for DA might explain this apparent discrepancy. In contrast, in dorsal striatum equivalent amounts of DA were required on day 1 to achieve similarAmax responses in LCRs and HCRs (Sabeti et al., 2002b). This suggests that, by itself, differences in amounts of DA applied are unlikely to explain the absence of altered DAT sensitivity to cocaine inhibition over time, as observed here in HCRs. Two possible explanations for the decreased amounts of DA needed in NAc of HCRs over time are fewer functional uptake sites secondary to tissue damage and/or regulation resulting in reduced basal DAT function. Because the amount of DA needed in LCRs remained constant across time, the latter is the more likely explanation. Chefer and Shippenberg (2002) have reported such changes in behaviorally sensitized rats, namely a reduction in basal DAT function with no changes in cocaine-induced DAT inhibition. Interestingly, behavioral sensitization in this study was assessed by rating repetitive/stereotypic behaviors. Together, these findings underscore the importance of assessing sensitization by both locomotor activity and stereotypy to understand the relevance of DAT regulation to changes in behavioral responsiveness. Furthermore, our results, along with previous reports, support the hypothesis that cocaine-induced adaptations in DAT in the NAc are necessary for the expression of locomotor sensitization. Therefore, LCR/HCR rats may be useful models for further study of differential phenotypes for initial sensitivity and sensitization to cocaine.
Acknowledgments
We gratefully acknowledge Gaynor Larson for expert technical assistance in setting up the electrochemical recordings.
Footnotes
-
This work was supported by National Institutes of Health Grants DA04216 (to N.R.Z.), AG06434 and NS39787 (to G.A.G.), Training Grant GM07635 (to J.S.), and Research Scientist Awards DA15050 (to N.R.Z.) and MH01245 (to G.A.G.).
-
DOI: 10.1124/jpet.102.047258
- Abbreviations:
- DA
- dopamine
- DAT
- dopamine transporter
- NAc
- nucleus accumbens
- Amax
- peak signal amplitude
- ANOVA
- analysis of variance
- LCR
- low cocaine responder
- HCR
- high cocaine responder
- Received November 22, 2002.
- Accepted December 30, 2002.
- The American Society for Pharmacology and Experimental Therapeutics