|Home | About | Journals | Submit | Contact Us | Français|
Drug-induced malfunction of nucleus accumbens (NAc) neurons underlies a key pathophysiology of drug addiction. Drug-induced changes in intrinsic membrane excitability of NAc neurons are thought to be critical for producing behavioral alterations. Previous studies demonstrate that following short-term (2d) or long-term (21d) withdrawal from non-contingent cocaine injection, the intrinsic membrane excitability of NAc shell (NAcSh) neurons is decreased, and decreased membrane excitability of NAcSh neurons increases the acute locomotor response to cocaine. However, animals exhibit distinct cellular and behavioral alterations at different stages of cocaine exposure, suggesting that the decreased membrane excitability of NAc neurons may not be a persistent change. Here, we demonstrate that the membrane excitability of NAcSh neurons is differentially regulated depending on whether cocaine is administered contingently or non-contingently. Specifically, the membrane excitability of NAcSh MSNs was decreased at 2d after withdrawal from either 5-day intraperitoneal (i.p.) injections (15 mg/kg) or cocaine self-administration (SA). At 21d of withdrawal, the membrane excitability of NAcSh MSNs, which remained low in i.p.-pretreated rats, returned to a normal level in SA-pretreated rats. Furthermore, upon a re-exposure to cocaine after long-term withdrawal, the membrane excitability of NAcSh MSNs instantly returned to a normal level in i.p.-pretreated rats. On the other hand, in SA-pretreated rats, the re-exposure elevated the membrane excitability of NAcSh MSMs beyond the normal level. These results suggest that the dynamic alterations in membrane excitability of NAcSh MSNs, together with the dynamic changes in synaptic input, contribute differentially to the behavioral consequences of contingent and non-contingent cocaine administration.
Functional alterations of the nucleus accumbens (NAc) by exposure to drugs of abuse contribute to a variety of addiction-related behaviors (Hyman et al., 2006) The functional output of a neuron, by definition, is the action potential (Hille, 2001). For NAc medium spiny neurons (MSNs), two key electrophysiological parameters determining the rate of action potentials are the excitatory synaptic input, which depolarizes the membrane toward the threshold of action potential, and the intrinsic membrane excitability, which determines whether and how many action potentials are fired upon membrane depolarization. Thus, the functional output of NAc MSNs at a given moment is largely determined by the integration of synaptic inputs and membrane excitability (Wilson and Groves, 1981).
To understand the impact of cocaine exposure on the functional output of NAc MSNs, extensive efforts have been directed toward elucidating cocaine-induced alterations in excitatory synapses using both non-contingent and contingent administration procedures. For example, surface or synaptic alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) are increased in NAc MSNs after long-term (14–21 d) withdrawal from either repeated intraperitoneal (i.p.) injections or intravenous (i.v.) self administration (SA) of cocaine (Boudreau and Wolf, 2005; Kourrich et al., 2007; Conrad et al., 2008; Boudreau et al., 2009; Ferrario et al., 2009). Furthermore, the excitatory synaptic strength is decreased upon a re-exposure to cocaine after long-term withdrawal from repeated i.p. injections of cocaine (Boudreau et al., 2007; Kourrich et al., 2007). On the other hand, understanding of the impact of cocaine exposure on the membrane excitability of NAc MSNs is largely derived from procedures using i.p. injection of cocaine. A number of previous studies have shown that NAc shell (NAcSh) MSNs exhibit decreases in sodium and calcium conductances (Zhang et al., 1998; Zhang et al., 2002), an increase in potassium conductance (Hu et al., 2004) following short-term withdrawal from repeated injections, and a decrease in evoked action potential firing following either a short- or long-term withdrawal (Henry and White, 1992; Dong et al., 2006; Ishikawa et al., 2009; Kourrich and Thomas, 2009). These changes consistently suggest that withdrawal from i.p. injections of cocaine decreases the membrane excitability of NAcSh MSNs (White and Kalivas, 1998). However, it remains unclear how the intrinsic membrane excitability of NAcSh MSNs is regulated by contingent cocaine administration. The present study attempts to address this issue by comparing the evoked action potential firing of NAcSh MSNs between rats receiving two widely-used cocaine procedures, the i.p. injection of cocaine that triggers locomotor sensitization and self-administration (SA) of cocaine via i.v. infusion. Our results show that the membrane excitability of NAcSh MSNs was dynamically and differentially regulated by the two procedures and by different withdrawal times. These results, together with previous results describing cocaine-induced synaptic and membrane alterations, help assess the overall functional output of NAcSh MSNs following different cocaine procedures.
Male Sprague Dawley rats were used for all studies except the high positive affect line and low positive affect line rats (Burgdorf et al., 2005), which were bred from the Long Evans line (Fig. 1K–O). Rats with an age of 22–24 days were allowed to acclimate to their home-cages for 5–7 days and then received cocaine administration. Thus, rats started receiving either i.p. injection or SA of cocaine at ~one month old. We chose to use relatively young rats because younger animals appear to exhibit faster acquisition of cocaine-elicited behaviors (Perry et al., 2007) and higher magnitudes of cellular alterations (Huang et al., 2009; Ishikawa et al., 2009). In a subset of experiments (Fig. 1K–O), the high line and low line Long Evans rats were used to explore the potential behavioral correlates of cocaine-induced decrease in membrane excitability of NAcSh MSNs. High and low line rats exhibit high and low levels of spontaneous ultrasonic vocalization at ~55 kHz, respectively (Burgdorf et al., 2005). In rodents, vocalization at this specific frequency has been hypothesized as an expressive measure of the positive emotional/motivational state (Panksepp and Burgdorf, 2003). Our previous study demonstrates that the high line rats exhibited a higher magnitude of cocaine-induced locomotor sensitization than low line rats (Mu et al., 2009). Given the correlation between cocaine-induced locomotor response and the membrane excitability of NAcSh MSNs (Dong et al., 2006), we compared the membrane excitability of NAcSh MSNs between high and low lines of rats in a set of experiments (Fig. 1K–O).
We used a 5-day procedure of repeated cocaine administration, which was similar to earlier studies (Dong et al., 2005; Ishikawa et al., 2009). Briefly, rats received one intraperitoneal injection (i.p.) of either (−)cocaine HCl (15 mg/kg) or the same volume of saline per day for 5 days. Some rats were injected within the home-cage. Others were injected within the behavioral box (16 in × 16 in × 12 in., Accuscan Inc., Columbus, OH), and the locomotor responses were recorded subsequently for 15 min. For the treatment paired with novel environment (Fig. 1E), we made a special cage with the same size as the home cage, walled with white plastic plate, floored with white-diamond soft bedding, and filled with red balls, green plastic bones, a yellow wood basin, and a wood ladder. Immediately after the injection of saline or cocaine, the rat was placed in this cage for 15 min and then put back into the home cage. These i.p. cocaine procedures were followed by a withdrawal period of either 2 d or 21 d, during which rats were kept within the home-cages. On the last withdrawal day, some rats received a challenge injection of cocaine (12 mg/kg), followed by measurement of the locomotor response. Brain slices were prepared 1 hr later for electrophysiological recordings.
The rats at an age of 24–28 days were anesthetized with the cocktail of ketamine, xylazine and acepromazine (0.1ml/100 mg) before surgery. A silastic catheter was inserted into the right auricle through the external jugular vein, passed under the skin and fixed in the mid-scapular region. The rats recovered from surgery for >2 d prior to beginning self-administration training sessions. During this time, catheters were flushed every 24 h with sterile 0.9% saline. Experiments were conducted in operant-conditioning chambers enclosed within sound-attenuating cabinets (Med Associates, St. Albans, VT). Each chamber contained an active and inactive nose poke hole, a food dispenser, a light used as the conditioned stimulus (CS), and a housing light. Correct nose pokes resulted in cocaine infusion (0.75 mg/kg in 0.10 ml over 6 seconds) and illumination of a CS light inside the nose poke hole. The CS light remained on for 6 sec, while a house light was illuminated for 20 sec during which active nose pokes were counted but resulted in no further cocaine infusions. After this 20 sec period, the house light was extinguished, and the next active nose poke resulted in a cocaine infusion. Nose pokes in the inactive hole had no reinforcement effects but were recorded. No food and water were provided in the chambers during the 2-h sessions.
Rats were allowed to self-administer cocaine or saline 2 h/day for 5 days. This relatively short-term procedure with relatively short daily exposure time each day was used to match the treatment duration and daily dose of the i.p. injection procedure; the potential caveats related to the treatment duration and dosing could thus be minimized. The rats were then returned to the home-cages for a 2-d or 3-wk withdrawal. One day after the last withdrawal day, rats were brought back to the behavioral chambers for 2 h, during which time some rats were connected to the perfusion pumps for cocaine re-exposure by SA of cocaine (0.6 mg/kg in 0.10 ml over 6 seconds), whereas others were not connected to the perfusion pumps. Nose pokes were recorded for all rats.
Detailed procedures for obtaining NAc slices can be found in our previous publications (Dong et al., 2006; Huang et al., 2008; Lee et al., 2008; Ishikawa et al., 2009). Briefly, coronal NAc slices of 250–300 µm thickness were cut such that the preparation contained the signature anatomical landmarks that delineated the NAc subregions. Slices were submerged in a recording chamber and were continuously perfused with regular oxygenated aCSF (in mM: 126 NaCl, 1.6 KCl, 1.2 NaH2PO4, 1.2 MgCl2, 2.5 CaCl2, 18 NaHCO3, and 11 glucose, 295–305 mOsm, equilibrated at 31–34°C with 95% O2/5% CO2).
Electrophysiological recordings were preferentially made from the MSNs located in the ventral-medial subregion of the NAcSh (Fig. 1 A–C). MSNs in this subregion have been shown to be importantly implicated in a variety of addiction-related molecular, cellular, and behavioral alterations (Kelley, 2004; Dong et al., 2006; Huang et al., 2008). Although all neurons were recorded from the ventral-medial subregion of the NAcSh, the exact, or more accurate, location of each neuron within this subregion could not be definitively traced through the electrophysiological microscope and thus was not documented. Standard whole-cell recordings were made using a MultiClamp 700B amplifier (Molecular Device) through an electrode (2–6 mΩ) in all electrophysiological experiments (Ishikawa et al., 2009). Current-clamp recordings were used to measure evoked action potentials, in which the resting membrane potential was normalized to −80 mV in acute slices. For these experiments, a K+-based internal solution was used (in mM: 130 K-methansulfate, 10 KCl, 10 HEPES, 0.4 EGTA, 2.0 MgCl2, 2.5 MgATP, 0.25 Na3GTP, pH 7.2–7.4; 275–285 mOsm). Only electrophysiologically “stable” neurons were included for data analysis. Briefly, after forming the whole-cell current-clamp configuration, the recorded cells were given ~5 min for stabilization (oscillating < 5 mV) of their resting membrane potentials. Cells with non-stabilized resting membrane potential were discontinued for the following recording. The resting membrane potential was then adjusted to −80 mV by injecting a small positive or negative current (usually <30 pA). If the compensating current was >50 pA, the recording of this neuron was discontinued. A current step protocol (from −200 to +400 pA, with a 50 pA increment; inter-pulse interval, 15 sec) was then run for at least five runs (normally 20 runs). Right after the recording of a particular cell, the number of evoked action potentials was compared across all runs; cells with a run-up or run-down >15% were excluded from further analysis. Interneurons were occasionally confronted, which could readily be distinguished by their signature electrophysiological properties, such as relatively depolarized resting membrane potential (~−72 mV), higher frequency of action potential firing upon current injection, or presence of large voltage sag at hyperpolarized potentials. These interneurons were excluded from data analysis. Thus, only MSNs with stable resting membrane potential and stable evoked action potential firing were included in the final analysis. The afterhyperpolarization potential (AHP) was sampled following the first action potential spike, usually elicited by the rheobase current. The input resistance was separately measured at both depolarized and hyperpolarized potentials. The measurement was taken between the injection currents −100 pA and +100 pA to minimize the significant rectifying segments in the I–V curve. All chemicals were purchased from Sigma-Aldrich.
In electrophysiological recordings, two to five cells were obtained from one rat. Numbers of cells (n) and animals (m) are presented as “n/m” in the manuscript. “n” as the default sample size was used in all statistics. In experiments related to key conclusions, animal-based statistics were also performed. In these analyses, electrophysiological parameters of all recorded cells from a single rat were averaged and the mean was used to represent this parameter of this rat. Thus, in animal-based statistics, “n” stands for the number of rats. In behavioral tests, “n” stands for the number of rats. All results are shown as mean ± SEM. One or two-factor repeated-measures ANOVA were used in most analyses. In these analyses, Factor-A was assigned for the treatments (e.g., cocaine vs. saline) and Factor-B was assigned for current injections (9-levels, 0 – 400 pA with a 50-pA interval). The statistical results were primarily presented in the F- and P-values of the main effect of Factor-A, which was the primary research interest (the effects of different cocaine treatments on membrane excitability). Degrees of freedom of between (b) and within (w) treatments were presented as F(b, w). For results with Factor-A containing >2 levels, an LSD posttest was performed when significance was obtained.
The computational model of NAcSh MSN was developed in the NEURON simulation environment (Hines and Carnevale, 1997; Carnevale and Hines, 2005). Simulations were performed on a Mac computer and data analysis was performed using MATLAB (Mathworks Inc, Natick, MA). The NAc MSN model has been previously described in detail (Wolf et al., 2005). Briefly, the model MSN consisted of 189 compartments and included almost all the known intrinsic currents and synaptic inputs expressed in the MSN. Channels were distributed throughout the cell in accordance with published data when possible. Spines were not explicitly modeled, but were accounted for their contribution to membrane area (Segev and Burke, 1998). The internal calcium concentration in a thin shell just inside the cell membrane was tracked for each compartment. BK and SK currents were regulated by calcium influx via N-, P/Q-, and R-type calcium channels, while the remaining calcium currents contributed to a separate pool based on published experimental results (Vilchis et al., 2000). Explicit glutamatergic and GABAergic synapses were modeled using a modified two-state synapse. Glutamatergic synapses were placed throughout the dendrites, but more distally relative to previous versions of the model (Wolf et al., 2005) to better reflect the actual distribution of glutamatergic synapses (Wilson, 1992; Gracy et al., 1999). Synaptic inputs were modeled using a modified version of the NetStim object provided in the NEURON package. Each synapse (glutamatergic or GABAergic) received an independent spike train generated by the following algorithm: first, a constant interspike interval (ISI) train was generated at the desired frequency. Each spike was then pulled anew from a Gaussian distribution centered at the original spike time. The resulting train was then randomly shifted, and this process was repeated for each of the 168 total synapses. Input was generated by using a large shift (one ISI) and a large standard deviation (1/4 of the ISI). The ratio of glutamatergic inputs to GABA inputs was held constant at roughly 1:1 for all simulations (Blackwell et al., 2003).
We examined the relative effects of changes in SK channel and AMPAR conductance levels by simulating synaptic input at a frequency of 1300 Hz and uniformly altering the maximum conductance in all cell compartments. We examined the net effects of combined SK channel and AMPAR conductance levels at three different frequencies: 1200, 1300, and 1400 Hz. Again, we uniformly altered the maximum conductance of these channels in all compartments. Firing frequency was calculated over a 1-sec epoch following a 200 ms initialization period for each simulation. Modulation levels for the channels were determined empirically by matching cell output to physiological behavior.
Several prevailing animal models have been used to understand cocaine-induced cellular and behavioral alterations. Despite its clear caveats, experimenter-performed i.p. injection of cocaine remains the most widely used procedure and has helped identify a large number of cocaine-induced molecular and cellular changes. Meanwhile, SA of cocaine offers a higher clinical relevance, but its triggered neuronal changes are relatively less understood. Here, we compared the effects of i.p. injection vs. SA of cocaine on the intrinsic membrane excitability of MSNs located within the ventral-medial (vm) region of the NAcSh (Fig. 1A–C).
Previous studies show that the intrinsic membrane excitability of NAcSh MSNs is decreased after short-term (1–2 d) or long-term (21 d) withdrawal from repeated i.p. injections of cocaine (Dong et al., 2006; Ishikawa et al., 2009; Kourrich and Thomas, 2009), a process likely mediated by changes in a set of voltage-gated ion channels (Zhang et al., 1998; Zhang et al., 2002; Hu et al., 2004). However, environmental cues, which play a significant role in drug-induced behavioral alterations (Badiani et al., 1995; Badiani et al., 1999), were not paired in most of these studies. To examine the impact of contextual components, we compared the membrane excitability of NAcSh MSNs between rats receiving within-home cage injections and rats receiving novelty-associated injections (daily i.p. injection for 5 d; measured on d 6; no re-exposure). The membrane excitability was assessed as the number of evoked action potentials. Our results show that repeated cocaine injections (15 mg/kg/day × 5 days) within either the home-cage or novel cage decreased the membrane excitability of NAcSh MSNs to a similar degree (F(3,330) = 3.75, p = 0.016, two-factor ANOVA; p = 0.01, saline-novel vs. cocaine-novel; p = 0.835, cocaine-novel vs. cocaine-home, LSD posttest; Fig. 1D, E). To analyze whether this conclusion also holds in the animal-based statistics, we used the means of the numbers of action potentials from all recorded cells in a rat to generate a single set of data for this rat and made comparisons between animals (same strategy was used for all following animal-based statistics). This animal-based statistics confirmed the results from the cell-based statistics (F(3,84) = 5.37, two-factor ANOVA; p = 0.01, saline-novel vs. cocaine-novel; p = 0.62, cocaine-novel vs. cocaine-home, LSD posttest). Notably, saline injections (once/day × 5 days), when paired with a novel environment, produced a trend toward a decrease in the membrane excitability (p = 0.07, LSD posttest; Fig. 1E). Nonetheless, the above results suggest that the primary effect of cocaine alone is sufficient to decrease the membrane excitability of NAcSh MSNs.
A robust behavioral alteration triggered by i.p. administration of cocaine is locomotor sensitization (Post and Rose, 1976). We next examined the potential alterations in the membrane excitability of NAcSh MSNs in rats experiencing a locomotor sensitization procedure (i.p., 15 mg/kg/day × 5 d; followed by locomotor test within the behavioral boxes). This procedure produced a reliable cocaine-induced locomotor sensitization (day × treatment: day: F(4,52) = 4.90, p < 0.01; treatment: F(1,13) = 87.132, p < 0.01; two-factor ANOVA; Fig. 1F). The second day after withdrawal, cocaine- and saline-pretreated rats, when placed back into the behavioral boxes without injections, exhibited basal locomotor activity similar to each other (t(13) = −1.829, p = 0.10, t-test). We observed that, similar to the cocaine procedure without the contextual component (home-cage cocaine), 2-day withdrawal from the locomotor sensitization procedure also decreased the membrane excitability of NAcSh MSNs (F(1,108) = 9.31, p < 0.01, two-factor ANOVA; Fig. 1G,H; animal-based statistics: F(1,42) = 25.46, p < 0.01, two-factor ANOVA). It has been shown that up-regulation of SK type Ca2+-activated potassium channels is one of the key ionic bases for the cocaine-induced decrease in membrane excitability of NAcSh MSNs (Ishikawa et al., 2009). Similarly, we observed an increase in the medium (m) component of the afterhyperpolarization potential (AHP) (t(20) = −2.21, p = 0.03, t-test; Fig. 1I, J), but not the fast (f) AHP (t(20) = 0.63, p = 0.53, t-test; Fig. 1I, J) after short-term withdrawal from i.p. injections of cocaine, suggesting a selective up-regulation of the underlying SK channels. In addition to the AHP, the rheobase, the minimal current that elicits action potential, was increased in cocaine-pretreated rats, suggesting that other types of ion channels were also involved in the effect of cocaine on membrane excitability (Table 1). One such candidate is voltage-gated sodium current, which has been shown to be decreased by a similar cocaine procedure (Zhang et al., 1998). Notably, the decreased sodium current appears to be mediated by a right-shift of the voltage-dependent activation (Zhang et al., 1998), and thus should not affect the basal input resistance, a change of which was not detected in the current study (Table 1).
The above results suggest that the cocaine procedure that produces locomotor sensitization decreases the membrane excitability of NAcSh MSNs. On the other hand, it has been shown that experimentally decreasing the membrane excitability of NAcSh MSNs increases cocaine-induced acute locomotor responses (Dong et al., 2006). To further examine the correlation between the membrane excitability of NAcSh MSNs and cocaine-induced locomotor alteration, we included two lines of Long Evans rats, which, referred to as the high line and low line rats, exhibited a higher and lower cocaine-induced locomotor response, respectively (Mu et al., 2009). In the present study, the membrane excitability of NAcSh MSNs was substantially lower in the high line rats than in the low line rats (F(1,114) = 6.61, p = 0.02, two-factor ANOVA; Fig. 1K). Consistent with the previous study (Mu et al., 2009), the cocaine-induced, but not spontaneous (Fig. 1L), locomotor responses were significantly higher in the high line rats (treatment: F(1,10) = 4.98, p = 0.04, two-factor ANOVA; Fig. 1M). To further explore the behavioral correlates of membrane excitability of NAcSh MSNs, we tested the potential differential acquisition of cocaine SA in the high and low lines rats. During the 5 d training period and on the challenge day (d 7; 2 d withdrawal), high line rats self-administered more cocaine than low line rats (animals: F(1,10) = 5.41, p = 0.04, two-factor ANOVA; Fig. 1N). Accompanying this, high line rats made fewer inactive nose-pokes (F(1,11) = 4.91, p = 0.05, two-factor ANOVA; Fig. 1O). Taken together, our results with high/low line rats suggest a correlation between decreased membrane excitability of NAcSh MSNs and increased cocaine-induced behavioral responses.
We then measured the membrane excitability of NAcSh MSNs after long-term withdrawal from the sensitization procedure. Rats received i.p. cocaine injections for five days. At d 5 of the treatment, rats receiving cocaine injections exhibited a clear locomotor sensitization (day: F(4,24) = 3.13, p = 0.03; treatment: F(1,6) = 134.45, p < 0.01; Fig. 2A). Rats were then placed back into the home-cages for a 3-wk withdrawal. By the end of the withdrawal, the baseline locomotor activities of these rats were measured, which were not different from saline-pretreated rats (t(6) = −0.992, p = 0.36, t-test; Fig. 2A). The subsequent electrophysiological recordings showed that the membrane excitability of NAcSh MSNs remained low in cocaine-pretreated rats compared with saline-pretreated rats at this withdrawal time point (F(4,384) = 2.85, p = 0.031, two-factor ANOVA; p = 0.02, saline-w/d vs. cocaine-w/d, LSD posttest; Fig. 2B, E; animal-based statistics: F(4,102) = 3.75, p = 0.02, two-factor ANOVA; p = 0.01, saline-w/d vs. cocaine-w/d, LSD posttest). This result is consistent with the previously-observed effect of cocaine on membrane excitability of NAcSh MSNs using a procedure without a contextual component (Ishikawa et al., 2009). Note that the rats with the long-term withdrawal procedure were ~3 weeks older than the rats with short-term withdrawal, and the potential age-dependent regulation may contribute to the observed lower basal membrane excitability (Ishikawa et al., 2009).
We next tested whether the membrane excitability of NAcSh MSNs was affected by re-exposure to cocaine after long-term withdrawal. For rats with the 5-day sensitization procedure and 3-week withdrawal, a challenge injection of cocaine (12 mg/kg) was given on withdrawal d 21 to verify cocaine-induced locomotor sensitization (day: F(4,32) = 11.203, p < 0.01; treatment: F(1,8) = 97.37, p < 0.001, two-factor ANOVA; Fig. 2C). Then, approximately 1 h after the injection, electrophysiological recordings were performed. In rats re-exposed to cocaine, the membrane excitability of NAcSh MSNs returned to the baseline level (F(4,384) = 2.85, p = 0.031, two-factor ANOVA; p = 0.85, saline-w/d vs. saline-w/d + cocaine; p = 0.01, cocaine-w/d vs. cocaine-w/d + cocaine; LSD posttest; Fig. 2D; animal-based statistics: F(4,102) = 3.75, p = 0.02, two-factor ANOVA; p = 0.76, saline-w/d vs. saline-w/d + cocaine; p = 0.005, cocaine-w/d vs. cocaine-w/d + cocaine; LSD posttest). Notably, one injection of cocaine in the saline-pretreated rats produced a trend toward an increase in the membrane excitability of NAcSh MSNs (p = 0.07, saline-w/d vs. acute cocaine, LSD posttest; Fig. 2D, E; p = 0.14, saline-w/d vs. acute cocaine, LSD posttest, animal-based statistics), raising the possibility that the re-exposure-induced increase in membrane excitability may be an acute effect of cocaine. In addition, no difference in either the fAHP (t(25) = −0.19, p = 0.85, t-test) or mAHP (t(25) = −1.67, p = 0.11, t-test) in NAcSh MSNs was observed between saline- and cocaine-pretreated rats. In contrast to rats with short-term withdrawal, the rheobase was not significantly affected (Table 2), suggesting that a different set of ion channels is involved in the effect of long-term withdrawal from i.p. cocaine administration.
We used a SA procedure (0.75 mg/kg/0.1 ml/6 sec × 2 hr), in which rats established a reliable SA of cocaine over a 5 d training period with a total cocaine intake comparable to the daily dose from the i.p. procedure (day: F(4,32) = 7.84, p = 0.02; treatment: F(1,8) = 80.12, p < 0.001, two-factor ANOVA; Fig.3A). After 2 d withdrawal, rats were placed back into the SA boxes for a 2-hr session without any cocaine treatment (Fig. 3A). The membrane excitability of NAcSh MSNs was decreased in this group of rats (F(1,114) = 5.30, p < 0.05; two-factor ANOVA; Fig. 3B, C; animal-based statistics: F(1,42) = 11.21, p < 0.05; two-factor ANOVA). Similar to that produced by the i.p. cocaine procedure, the mAHP in NAcSh MSNs was selectively increased in rats after 2-day withdrawal from cocaine SA (mAHP: t(20) = −2.73, p = 0.01; fAHP: t(20) = −1.80, p = 0.08; t-test Fig. 3D, E). Moreover, when the rats with the 2 d withdrawal from the sensitization procedure and rats with the 2 d withdrawal from SA procedure were compared, no difference was observed in the membrane excitability of NAcSh MSNs (F(3,222) = 8.11, p < 0.01, two-factor ANOVA; p >0.05, i.p. vs. i.v. cocaine, LSD posttest).
To evaluate the potential environmental contribution, we ran another group of rats, which were not placed back into the behavioral boxes on d 2 of withdrawal following the 5 d SA procedure (without re-exposure to the SA box). We observed that the membrane excitability of NAcSh MSNs was also decreased in cocaine-pretreated rats (F(1,96) = 6.01, p < 0.05; two-factor ANOVA, Fig. 3F), and no difference in membrane excitability of NAcSh MSNs was observed between cocaine-pretreated rats with and without re-exposure to the SA boxes (F(3,222) = 5.67, p < 0.01, two-factor ANOVA; p = 0.68, cocaine vs. cocaine with SA-box pairing; p = 0.88, saline vs. saline with the last day pairing, LSD posttest).
We then examined the effect of long-term withdrawal from the SA procedure. After acquiring stable SA over a 5 d period of training (day: F(4,32) = 7.89, p <0.01; treatment: F(1,8) = 20.51, p < 0.01; two-factor ANOVA; Fig. 4A), rats were returned to the home-cages for a 3-week withdrawal. By the end of withdrawal, rats were placed back to the SA boxes for 2 h without being connected to the cocaine infusion pump. Immediately after this context exposure, rats were sacrificed for electrophysiological recordings. Our results showed that there was no significant difference in the membrane excitability of NAcSh MSNs between saline- and cocaine-pretreated rats (F(4,402) = 3.85, p = 0.01, two-factor ANOVA; p = 0.26, saline-w/d vs. cocaine-w/d, LSD posttest; Figs. 4B, E; animal-based statistics: F(4,114) = 3.25, p = 0.03, two-factor ANOVA; p = 0.26, saline-w/d vs. cocaine-w/d, LSD posttest). Thus, the membrane excitability of NAcSh MSNs becomes seemingly “normal” after the long-term withdrawal from cocaine SA.
To examine the effect of re-exposure to cocaine after long-term withdrawal, we trained rats with the 5 d SA procedure (day: F(4,40) = 4.429, p < 0.01; treatment: F(1,10) = 33.729, p < 0.001; two-factor ANOVA; Figs. 4C), followed by a 3-week withdrawal. On the last day of withdrawal, we placed the rats back into the SA boxes for a 2-hr session of cocaine SA. Thus, the re-exposure is “contingent”. Immediately after this behavioral procedure, brain slices were prepared for electrophysiological recordings. We observed that the membrane excitability of NAcSh MSNs became significantly higher than that in either saline- or cocaine-pretreated rats without re-exposure. Notably, in saline-pretreated rats with the last-day training of SA cocaine, the membrane excitability of NAcSh MSNs exhibited a trend toward an increase (F(4,402) = 3.85, p = 0.01, two-factor ANOVA; p = 0.05, saline-w/d vs. saline-w/d + cocaine; p = 0.04, cocaine-w/d vs. cocaine-w/d + cocaine; p = 0.11, saline-w/d vs. acute cocaine; LSD posttest; Fig. 4D,E; animal-based statistics: F(4,114) = 3.25, p = 0.03, two-factor ANOVA; p = 0.09, saline-w/d vs. saline-w/d + cocaine; p = 0.01, cocaine-w/d vs. cocaine-w/d + cocaine; p = 0.18, saline-w/d vs. acute cocaine). In contrast to the i.p. procedure, a re-exposure to cocaine selectively increased the fAHP (F(4,71) = 4.88, p < 0.01, one factor ANOVA; Fig. 4F, G) in NAcSh MSNs without affecting the mAHP (F(4,71) = 0.67, p =0.86, one factor ANOVA; Fig. 4F, H) or other active and passive membrane properties (Table 3). As demonstrated previously, an up-regulation of the BK type Ca2+-activated potassium channel, reflected in the current-clamp recording as the fAHP, may expedite the repolarization of the membrane during action potential and thus increase the firing rate (Ishikawa et al., 2009). Thus, this increase in the fAHP, or BK channels, may contribute to the re-exposure-induced increase in membrane excitability of NAcSh MSNs.
Using a computational model, we made an attempt to predict cocaine-induced functional alterations of NAcSh MSNs by integrating cocaine-induced membrane and synaptic adaptations. In this model, the functional output of NAcSh MSNs was defined as the number of action potentials (spikes) driven by excitatory synaptic inputs. For cocaine-induced membrane alterations, we focused on the SK channels and defined the computational range of changes in SK channel activity based on our present (Fig. 3, Fig. 4) and previous results (Ishikawa et al., 2009). For cocaine-induced alterations in excitatory synaptic input, we focused on AMPAR-based synaptic strength and defined the computational range of changes based on published results (Boudreau and Wolf, 2005; Boudreau et al., 2007; Kourrich et al., 2007; Conrad et al., 2008; Boudreau et al., 2009). With this model, we first examined the independent effects of cocaine-induced changes in membrane properties and excitatory synapses on the functional output (spikes) of NAcSh MSNs. The results show that if acting independently, cocaine-induced up-regulation of the SK channel activity decreased the spike output of NAcSh MSNs (Fig. 5A), whereas cocaine-induced up-regulation of excitatory synaptic strength produced the opposite effect (Fig. 5B). Furthermore, both membrane and synapse-mediated modulations exhibited a linear relationship (Fig. 5A, B) as suggested previously (Moyer et al., 2007). We next integrated cocaine-induced alterations in both the membrane property and excitatory synaptic input in this model to predict how the overall functional output (spikes) of NAcSh MSNs was affected by different cocaine procedures at different stages of exposure/withdrawal. The baseline excitatory synaptic activity was set at 1300 Hz, with a 100 Hz up- or down-regulation. This model predicted that following repeated i.p. injections of cocaine, the overall functional output (number of spikes driven by excitatory synapses) of NAcSh MSNs was increased at both short- and long-term withdrawal time points, but was decreased after re-exposure to cocaine (Fig. 5C). This model also predicted that cocaine-induced alterations in excitatory synaptic inputs may override the cocaine-induced membrane alteration and predominantly regulate the output of NAcSh MSNs. On the other hand, following cocaine SA, the functional output of NAcSh MSNs was decreased after short-term withdrawal (e.g., 1 d), increased after long-term withdrawal (e.g., 45 d), and increased further after re-exposure (Fig. 5D). Taken together, these modeling results suggest that the functional output of NAcSh MSNs is differentially regulated by different cocaine procedures, different withdrawal durations, and re-exposures to cocaine.
We have demonstrated that the intrinsic membrane excitability of NAcSh MSNs underwent dynamic changes upon different cocaine procedures and at different stages of cocaine withdrawal. These results, combined with the knowledge on cocaine-induced synaptic alterations, may help understand the dynamic functional output of NAcSh MSNs in cocaine-induced behaviors.
In anesthetized animals, NAc MSNs dwell at two alternating states, the functionally inactive downstate at which the membrane potential is relatively hyperpolarized and MSNs are often quiescent, and the functionally active upstate at which the membrane potential is relatively depolarized and MSNs are prone to firing action potentials (O'Donnell and Grace, 1995; Wilson and Kawaguchi, 1996). Both in vitro and in vivo studies show that the upstate is mediated by synchronous activation of excitatory synapses (O'Donnell and Grace, 1995; Plenz and Kitai, 1996; Plenz et al., 1998). In behaving animals, NAc MSNs likely spend most of their time in the upstate, presumably due to continuous excitatory inputs (Mahon et al., 2006). Nonetheless, it is excitatory synaptic input that depolarizes the membrane potential toward the action potential threshold, and it is the intrinsic membrane excitability that sets the action potential threshold and determines how often to fire action potentials once MSNs dwell at the threshold. Thus, the functional output of NAc MSNs critically relies on the integration of both excitatory synaptic input and membrane excitability.
Following i.p. cocaine injection-induced locomotor sensitization, the strength of excitatory synapses at NAcSh MSNs is increased after 7–21 d of withdrawal and decreased 24 h after re-exposure after withdrawal (Boudreau and Wolf, 2005; Boudreau et al., 2007; Kourrich et al., 2007; Ferrario et al., 2009). Our current results show that the membrane excitability of NAcSh MSNs was decreased at both short and long withdrawal time points and returned to the normal level upon approximately 1 h after cocaine re-exposure (Fig. 1). Together with previous results (Zhang et al., 1998; Zhang et al., 2002; Hu et al., 2004), our present findings suggest that cocaine-induced alterations in excitatory synapses and membrane excitability of NAcSh MSNs may partially cancel out each other’s effects at the functional output level. Homeostatic synapse-driven membrane plasticity (hSMP) was recently defined in NAcSh MSNs, through which the membrane excitability can be adjusted to functionally compensate for the changes in excitatory synapses (Ishikawa et al., 2009). This hSMP may be one of the mechanisms that coordinates excitatory synapses and membrane excitability of NAcSh MSNs following the i.p. injection procedure. Furthermore, although the role of NAc excitatory synapses in cocaine-induced locomotor sensitization has long been established (Wolf, 1998), our results suggest that it is more likely that the overall functional output, i.e., the integration of excitatory synaptic input and the membrane excitability as well as other components sets up the sensitized psychomotor state. The role of membrane excitability of NAcSh MSNs in cocaine-induced locomotor sensitization is emerging; a recent study shows that experimentally decreasing the membrane excitability of in vivo NAcSh MSNs by over-expressing potassium channels increases the acute effect of cocaine on locomotor responses (Dong et al., 2006). Consistent with this notion, our current study using a repeated cocaine procedure in high line and low line rats show that the decreased membrane excitability of NAcSh MSNs is correlative to cocaine-elicited locomotor sensitization (Fig. 1K–M). Furthermore, the differential acquisition patterns of cocaine SA in high line and low line rats (Fig. 1N, O) suggest that the membrane excitability of NAcSh MSNs is involved in not only the locomotor response but likely a large repertoire of cocaine-related behaviors.
Following the procedure of cocaine SA, the strength of excitatory synapses at NAcSh MSNs is increased after long-term withdrawal (Conrad et al., 2008). Whereas the change of excitatory synaptic strength upon re-exposure has not been comprehensively examined, one study suggests that the function of synaptic AMPARs may be up-regulated (Anderson et al., 2008). In the present study, the membrane excitability of NAcSh MSNs was decreased after short-term withdrawal but returned to normal levels after long-term withdrawal, and was increased upon re-exposure to cocaine (Fig. 3, Fig. 4). Thus, the net effect of cocaine on excitatory synapses and membrane excitability would be to increase the overall activation level of NAc MSNs after long-term withdrawal from cocaine SA (Fig. 5D).
Taken together, our results demonstrate that the intrinsic membrane excitability of NAcSh MSNs changes following different procedures of cocaine exposure and over different stages of withdrawal; these changes are temporally correlated with the synaptic alterations, which seem to be functionally opposite at certain withdrawal time points. Although some of the co-occurring, functionally counteracting changes across the subcellular domains appear to help maintain the stability of NAcSh MSNs on the overall functional level, it is likely that the functional output of NAcSh MSNs is still skewed during certain working range (Fig. 5C, D).
Focusing on two key components of the basic neuronal machinery, the above discussion together with the modeling results (Fig. 5) should provide a comprehensive understanding about how NAcSh MSNs respond to excitatory inputs after short- and long-term withdrawal following cocaine administration. Two key potential caveats should be noted in these modeling results. First, the membrane excitability and excitatory synaptic strength after short- and long-term withdrawal were approximated based on our current results and published results (Boudreau and Wolf, 2005; Boudreau et al., 2007; Kourrich et al., 2007; Conrad et al., 2008; Ferrario et al., 2009; Ishikawa et al., 2009). In these studies, the exact time points of withdrawal as well as the time points at which the biochemical and electrophysiological measurements of membrane excitability and excitatory synapses were performed were not always precisely matched. For example, at re-exposure after long-term withdrawal from cocaine SA (Fig. 5D), the data of membrane excitability included in the model were collected at 1 h after re-exposure (current results) whereas the data about excitatory synaptic strength were collected at ~30 min (Anderson et al., 2008). These imperfectly matched measurement time points may cause inaccurate estimations. Second, although a number of ionic conductances (e.g., Na+, K+, and Ca2+ currents) are likely involved in cocaine-induced membrane adaptation (Zhang et al., 1998; Zhang et al., 2002; Hu et al., 2004), the related data are not available at each key time point. We thus included only SK channel-related data (Ishikawa et al., 2009) in the modeling work. This simplification may in part account for the lower impact of membrane excitability on the overall functional output of NAcSh MSNs predicted by the model (Fig. 5A, B).
Moreover, to assess the real-time activity of these neurons in vivo, factors within the circuitry must also be considered. First, presynaptic inputs are the driving force to depolarize or hyperpolarize the membrane potential of NAcSh MSNs (Wilson, 1986; Plenz and Kitai, 1998). Presynaptic excitatory and inhibitory inputs, as well as their temporal integration, are the potential target of cocaine exposure (Shoji et al., 1997; Kozell and Meshul, 2004), and the resulting changes may substantially change the duration/frequency of the upstates of NAcSh MSNs. Second, the observations of increased dendritic spine density (Robinson and Kolb, 1997; Kolb et al., 2003) and potentially new synaptic connections (Huang et al., 2009) following cocaine exposure raise the possibility that new input pathways may be generated. If so, these potential new pathways may carry out qualitatively distinct information that cannot be quantitatively evaluated as an increase or decrease in synaptic strength. Third, a number of neuromodulators, such as dopamine, serotonin, norepinephrine, and neurotrophins, constitutively regulate both the membrane and synaptic properties of NAcSh MSNs (Koob and Nestler, 1997). Upon cocaine exposure, these neuromodulator/neurotrophin systems also undergo adaptive and dynamic changes (Koob and Nestler, 1997), which, in turn, may reshape the real-time functional output of these neurons. Nonetheless, with these interpretational caveats, predictions of the functional output of NAcSh MSNs based on the above analyses are largely consistent with some of the neuroimaging studies that directly measure the activation level of the NAc upon cocaine exposure. For example, in i.p. cocaine-pretreated rats, a re-exposure to cocaine slightly increases the activity level of the NAc, with an effect much less than acute cocaine exposure in naïve rats (Febo et al., 2005), whereas in human subjects who previously “self administer” cocaine, a re-exposure to cocaine substantially increases the activity level of the NAc (Breiter et al., 1997).
In summary, the current study provides essential information to assess the overall functional alterations of NAcSh MSNs upon cocaine exposure. Analyses based on the current results along with previous results (Boudreau and Wolf, 2005; Boudreau et al., 2007; Kourrich et al., 2007) support the hypothesis that NAc MSNs act dynamically during different stages of cocaine exposure and withdrawal. These results shed light on the path toward understanding the cellular and molecular bases underlying drug-induced emotional and motivational alterations, such as drug-craving and drug withdrawal-induced depression.
This research was supported in part by State of Washington Initiative Measure 171, National Institutes of Health (NIH) DA023206 (Y.D.), the Hope Foundation for Depression Research (Y.D., J.P.), the Humboldt Foundation (Y.D.), and European Commission Coordination Action ENINET LSHM-CT-2005-19063 (O.M.S.). Cocaine was supplied by the Drug Supply Program of National Institute on Drug Abuse. We thank Drs. Marina Wolf and Yanhua Huang for suggestions on the manuscript.