Nicotine-induced neuroplasticity in striatum is subregion-specific and reversed by motor training on the rotarod.

Abstract Nicotine is recognized as one of the most addictive drugs, which in part could be attributed to progressive neuroadaptations and rewiring of dorsal striatal circuits. Since motor‐skill learning produces neuroplasticity in the same circuits, we postulate that rotarod training could be sufficient to block nicotine‐induced rewiring and thereby prevent long‐lasting impairments of neuronal functioning. To test this hypothesis, Wistar rats were subjected to 15 days of treatment with either nicotine (0.36 mg/kg) or vehicle. After treatment, a subset of animals was trained on the rotarod. Ex vivo electrophysiology was performed 1 week after the nicotine treatment period and after up to 3 months of withdrawal to define neurophysiological transformations in circuits of the striatum and amygdala. Our data demonstrate that nicotine alters striatal neurotransmission in a distinct temporal and spatial sequence, where acute transformations are initiated in dorsomedial striatum (DMS) and nucleus accumbens (nAc) core. Following 3 months of withdrawal, synaptic plasticity in the form of endocannabinoid‐mediated long‐term depression (eCB‐LTD) is impaired in the dorsolateral striatum (DLS), and neurotransmission is altered in DLS, nAc shell, and the central nucleus of the amygdala (CeA). Training on the rotarod, performed after nicotine treatment, blocks neurophysiological transformations in striatal subregions, and prevents nicotine‐induced impairment of eCB‐LTD. These datasets suggest that nicotine‐induced rewiring of striatal circuits can be extinguished by other behaviors that induce neuroplasticity. It remains to be determined if motor‐skill training could be used to prevent escalating patterns of drug use in experienced users or facilitate the recovery from addiction.


| INTRODUCTION
Tobacco use is a leading preventable cause of death worldwide. Even though the use of conventional cigarettes has decreased during the last 20 years, smoking is still a major health problem. In addition, the introduction of electronic cigarettes (e-cigarettes) has once again increased nicotine use globally, and there is an association between initial e-cigarette use and cigarette smoking later in life. 1 In order to reduce smoking prevalence, there is a need for more basic research that defines the effects by nicotine on neuronal circuits, and studies outlining how nicotine-mediated maladaptation of neuronal function can be prevented.
Both clinical and preclinical studies have implicated the striatal nucleus in acute and long-lasting effect by nicotine, and striatal circuit dysfunction has repeatedly been shown in dependent smokers. [2][3][4][5][6][7] The rodent striatum can be subdivided into nucleus accumbens (nAc) shell and core (linked to drug reinforcement), dorsomedial striatum (DMS; linked to reward-guided behaviors), and dorsolateral striatum (DLS; linked to habitual responding). [8][9][10][11] Preclinical studies suggest that striatal circuits are progressively transformed during protracted drug taking, and that neurophysiological transformations are transferred and stored in the DLS, which may promote habitual/compulsive drug taking behavior (addiction). 2,5,12,13 Drug-induced rewiring of dorsal striatum has furthermore been suggested to be driven by functional shifts in neuronal circuits of the amygdala. 14 Even though a causal relationship between these neuronal transformations and the development of addiction has not been established, we speculate that interventions aimed at restoring neuronal function could be beneficial to improve smoking cessation.
Interestingly, during the acquisition and consolidation of a motor skill, striatal circuits are progressively rewired in a spatial and temporal manner that partially resembles the effects by nicotine. In fact, studies of motor-skill consolidation have led up to the postulate that neuroadaptations initially occurring in the DMS are transferred to the DLS where they may become permanent. 9

| Experimental outline
In this study, we test the hypothesis that motor-skill training on the rotarod is sufficient to block the temporal and spatial sequence of neuronal transformations that are elicited by nicotine. Group-housed male rats (200-220 g, n = 30 per round, with three rounds in total = 90 rats) were randomized to receive vehicle or nicotine injections over three consecutive weeks. Behavioral sensitization to the locomotor-stimulatory properties of nicotine was assessed in an open-field arena. Following behavioral sensitization, a subset of animals was trained for 5 days on the rotarod, resulting in a total of four groups of rats: vehicle control (n = 29), vehicle rotarod (n = 16), nicotine control (n = 29), and nicotine rotarod (n = 16).
One set of animals was used to monitor locomotor behavior, and the rest were examined in electrophysiological recordings to explore progressive neuroadaptations in neuronal circuits of the striatum and amygdala.

| Drugs and solutions
Nicotine tartrate was dissolved in 0.9% NaCl, adjusted to pH 7. 2-7.4 with NaHCO 3 and administered at 1.0 mg/kg sc (0.36 mg/kg nicotine

| Animals
Male Wistar rats (Janvier, France) were group housed and kept on a 12/12 hours light/dark cycle at 20°C with 50% humidity with free access to food and water. All experiments were approved by the Gothenburg Animal Research Ethics Committee and conducted during the light time cycle.

| Nicotine treatment and behavioral sensitization
The animals were randomly assigned to receive either nicotine or saline over a period of 3 weeks. Subcutaneous injections of nicotine were given five times a week (15 injections in total). To minimize the possible pain caused by acidic nicotine, pH was normalized using NaHCO 3 . 17 Even though not self-administered, this protocol has repeatedly been shown to induce robust behavioral sensitization to nicotine that is sustained for at least 7 months. 18 Locomotor behavior was monitored after the first and last nicotine injection, and in a subset of animals 1 week after rotarod training. Vertical and horizontal movements were registered in an open-field arena (40 × 40 cm, Med Assoc., Fairfax, VT, USA) placed in a sound attenuated, ventilated, and dim lit box. The open-field arena was equipped with a two-layer grid, consisting of rows of photocell beams, and consecutive beam breaks were tracked by a computer-based system (Activity Monitor 7, Med Assoc., St. Albans, VT, USA). Rats were allowed to habituate to the box for 30 minutes and were then injected with nicotine or vehicle, after which locomotor behavior was registered for another 30 minutes.

| Brain slice preparation for ex vivo electrophysiology
For brain slice preparation, rats were anesthetized with isoflurane (Forene, Baxter, Sweden) before decapitation, and brains rapidly removed and submerged in continuously oxygenated (95% O 2 , 5% CO 2 ) modified aCSF. Coronal brain slices (250 μm) were sectioned using a Leica VT 1200S Vibratome (Leica Microsystems AB, Bromma, Sweden) and were allowed to equilibrate at 34°C for 30 minutes in oxygenated normal aCSF before maintained in room temperature for the remainder of the day. The experimenters performing electrophysiological recordings were blind to drug treatment.

| Electrophysiological field potential recordings
Electrophysiological field potential recordings were performed as previously described. 19 In brief, one hemisphere of the slice was constantly perfused with prewarmed aCSF (30°C, 2 mL/min), and field population spikes (PSs) were evoked with a stimulating electrode  (Figures 1 and 3). Signals were amplified with a custom-made amplifier, filtered at 3 kHz, and digitized at 8 kHz. PSs were evoked with increasing stimulation strength (18-72 μA) to create an input/output function. When assessing the responsiveness to applied agonists/antagonists PS amplitude was set to half max response, and a stable baseline was recorded for 10 minutes before drugs were administered via bath perfusion.
Endocannabinoid-mediated long-term depression (eCB-LTD) was induced by four trains of 100 pulses delivered at 100 Hz with a 4second intertrain interval (HFS stimulation). This protocol has previously been shown to robustly induce eCB signaling.

| Statistics
All data were analyzed using Clampfit 10.2, Microsoft Excel, and GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). Gaussian distribution was tested with D'Agostino-Pearson omnibus normality test. For behavioral and electrophysiological data repeated measure, two-way analysis of variance (ANOVA) was used for comparisons over time, and input/output function, while paired or unpaired t tests and Kolmogorov-Smirnov test were used when applicable. All parameters are given as mean ± standard error of the mean (SEM), and the level of significance was set to P < .05.

| Temporal rewiring also involves other circuits of the striatal and amygdala nucleus
Striatal neuroplasticity has been reported following exposure also to other drugs of abuse, and it is possible that rewiring of striatal circuits could be a common denominator during the establishment of addiction. Importantly, studies evaluating neuroadaptations elicited by cocaine suggest that striatal rewiring is linked to parallel transformations in ventral striatum and neuronal circuits of the amygdala. 14 To assess changes in other brain subregions that could be involved in establishing nicotine-induced rewiring of dorsal striatum,

| Motor-skill training on the rotarod elicits neurophysiological transformations
Input/output function was not significantly altered in the DMS after rotarod training ( F 1,109 = 2.33, P = .13), but depressed in the DLS ( F 1,89 = 11.8, P < .001) ( Figure 4A,C). At the same time, whole-cell FIGURE 2 Nicotine elicits a region-specific modulation of GABAergic neurotransmission. A, Disinhibition induced by the GABA A receptor antagonist picrotoxin was significantly enhanced in the DMS of nicotine-exposed animals. B-E, Whole-cell recordings revealed no treatment effects on recorded sIPSCs in the DMS. Example traces show recorded sIPSCs in vehicle and nicotine-treated rats. Scale bar is 200 pA and 10 s. F, There was no effect by nicotine treatment on picrotoxin-induced disinhibition in the DLS. G-J, In the DLS, recordings performed in brain slices from nicotine-exposed animals showed an increase in sIPSC amplitude, while rise and decay time was depressed. K-N, Nicotine had no effect on the cumulative distributions of either event intervals or amplitudes in the DMS, while a significant effect on amplitude was present in the DLS.

| Rotarod training abolishes nicotine-induced rewiring of striatal circuits
Vehicle and nicotine-treated rats were trained for 5 days on the rotarod. In one set of rats, the ability for rotarod training to reverse

| Training on the rotarod rescues synaptic plasticity in the form of eCB-LTD
The eCB system plays a key role in mediating nicotine-induced neuroplasticity, and nicotine has previously been shown to produce maladaptive eCB signaling that may have relevance for maintenance of addiction-related behavior. 22 In the last set of experiments, acute and long-lasting effects by nicotine and rotarod training on eCB-LTD were assessed in the DMS and DLS. During acute withdrawal, there was no effect by nicotine on eCB-LTD

| DISCUSSION
The prevalence of e-cigarette use has risen rapidly since the introduction 2007, especially among youths. There is a belief that the harmful component from smoking is the combustion of tobacco, and that tobacco free nicotine, which is used in e-cigarettes, is safe. In the study presented here, the pure action of nicotine on neuronal circuits was assessed. It is important to note that nicotine was neither selfadministered nor smoked, and this experimental setup does thereby not resemble normal human smoking. At the same time, it is noteworthy that in despite of being unvoilational, pure nicotine exposure is sufficient to produce long-lasting behavioral transformations that are accompanied by brain region-specific transformations. These findings thereby support a potent effect by nicotine on biological systems and suggest that nicotine may produce prolonged reorganizations of neuronal circuits independent of route of administration or motivational aspects of drug taking.
Even though nicotinic acetylcholine receptors (nAChRs) are not expressed on striatal medium spiny projection neurons (MSNs), nicotine exposure exerts a complex modulatory effect on striatal neurotransmission. 21 Furthermore, the data presented here show that repeated exposure to nicotine produce neuronal transformations in a distinct temporal and spatial sequence. We also show that neuroadaptations in DLS, nAc shell, and CeA occur at the same time point, which might be fundamental for establishing maladaptive incentive habits. 8 If rewiring of amygdalo-striatal circuits is a neurobiological underpinning of substance abuse liability, restoring neuronal function in these regions might be sufficient to suppress escalated patterns of substance use and to reinstate the control over behavior.
Encouragingly, we found that a short period of training on the rotarod is sufficient to extinguish striatal neuroadaptations, both acutely and in a long-lasting manner. We furthermore show that rotarod training restores synaptic plasticity in the form of eCB-LTD. The eCB system has been reported to be impaired not only after nicotine exposure but also following exposure to other drugs of abuse, [22][23][24][25] and a maladaptive eCB system has been suggested to contribute to the FIGURE 5 Rotarod training prevents neurodaptations in the striatum. A,B, Training on the rotarod did not affect the sensitized response to the locomotor-stimulatory properties of nicotine but significantly reduced nicotine-induced rearing. C, Ex vivo electrophysiological recordings performed on rotarod-trained rats demonstrated no significant effects by nicotine treatment on input/output function in any brain region studied. For comparison, in brain regions where nicotine significantly modulated input/output function as compared with vehicle-exposed control, the input/output function in untrained rats exposed to nicotine is indicated by an additional curve in gray. D, Following 3 months withdrawal, input/ output function remained unaltered in striatal subregions, but an increase in evoked PSs was still present in CeA. Curves in gray represent input/ output function in untrained rats exposed to nicotine 3 months earlier. Data are mean values ± SEM; n = the number of recordings, based on at least five animals/treatment group. CeA, central nucleus amygdala; DLS, dorsolateral striatum; DMS, dorsomedial striatum; nAc, nucleus accumbens; PS, population spike; SEM, standard error of the mean; WD, withdrawal development and maintenance of addiction-related behavior. 26 The finding that rotarod training is sufficient to restore eCB-mediated plasticity thus further implicates a potential impact by behaviors that induce neuroplasticity on the treatment of addiction. If these theories hold true, motor-skill training combined with exercise could be a valuable addition to pharmacological and psychosocial interventions when reducing not only smoking but also substance use in experienced drug users.
The datasets presented here demonstrates a spatially and temporally specific suppression of evoked field potentials in animals treated with nicotine, where the DMS (reward driven behaviors) is initially modulated while DLS (habitual responding) is recruited at a later stage.
Even though inhibitory synaptic activity was not significantly modulated in the DMS of nicotine-treated rats, picrotoxin produced a greater disinhibition, indicating that tonic inhibition might be enhanced following nicotine administration, thereby putatively suppressing evoked field potentials. Changes in input/output function lasted for 1 month in DMS and nAc core, and after 2 months, there were no detectable effect on neurotransmission in any brain region analyzed. At this time point, it is possible that other structures, such as cortical or dopaminergic areas in the midbrain, are recruited as an intermediate state before neurophysiological alterations arise in the striatum and amygdala. 11 Following 3 months of abstinence from nicotine, robust transformations were established in the DLS, nAc shell, and CeA. These findings are especially interesting considering that previous studies, evaluating incubation of cocaine craving, reported a progressive increase in BDNF levels that peak in nAc and amygdala after 3 months of withdrawal. 27 It is thus possible that progressive shifts in amygdalo-striatal circuits not only contributes to the formation of incentive habits but also are linked to cue-induced craving and relapse after prolonged abstinence. 8,28 Furthermore, at this time point, there is an increase in exploratory/anxiolytic behavior, which also appears to be directly linked to neurophysiological transformations in nAc shell. 5 Previous studies have shown that dorsal striatum is recruited during consolidation of a motor skill, and that region-and pathway-specific plasticity sculpts striatal circuits as the motor skill become automatized. 15,16 Electrophysiological recordings, performed in vivo and ex vivo, show an increase in MSN activity in the DMS during the initial training period, with the DLS being affected first when the skill is fully consolidated. 16 In our hands, 5 days of rotarod training significantly increased sIPSC frequency in both the DMS and DLS and increased sIPSC amplitude in the DLS, indicating that presynaptic and postsynaptic transformations occur in a brain subregion-specific manner. In addition, input/output function was significantly depressed in the DLS and nAc shell, suggesting that other brain regions may be recruited during this process. It should be noted that these recordings were performed 3 to 7 days after the last training session, and we do not know if these transformations are transient or long lasting. Furthermore, these studies do not separate between MSNs expressing dopamine D1 and D2 receptors, which also might affect the experimental outcome. 29 Nevertheless, the data presented here implicate a selectivity with regard to rotarod-elicited neurophysiological transformations, which could be important for understanding how the formation of motor-related memories initially involves recruitment the DMS and how it is then transferred and stored in the DLS.
Progressive rewiring of dorsal striatum has been reported following exposure to several types of drugs of abuse and might be driven by neuroplasticity in other circuits of the striatal and amygdala nucleus. 14 In agreement with these theories, we found that nicotine acutely modulates neurotransmission in nAc core, with protracted and CeA. However, in contrast to what has been reported after cocaine self-administration, nicotine did not produce a significant effect on neurotransmission in BLA. 14 It is possible that parallel neuroplasticity in the nAc shell and DLS underlies the reduced control over incentive habits and might in this regard play a role for the transition from recreational drug use towards addiction. Importantly, 5 days of rotarod training during the initial week of withdrawal completely prevented not only the acute transformations in DMS and nAc core but also progressive neuroadaptations in DLS and nAc shell. Furthermore, we found that the sensitized response to nicotine-induced rearing, which previously has been linked to DMS, was blocked in animals trained on the rotarod. 2 These findings collectively suggest that motor-skill learning not only reverses nicotine-induced neuroadaptations in the acute phase but also prevents the progressive rewiring of striatal circuits that normally develop during withdrawal after nicotine use.

ACKNOWLEDGMENTS
We are thankful for the assistance provided by Amir Lotfi, Klara

CONFLICT OF INTEREST
The authors report no financial disclosure or potential conflicts of interest.

AUTHOR CONTRIBUTIONS
LA designed the study, performed field potential recordings, data analysis, and wrote the manuscript. VL performed whole-cell recordings, data analysis, and edited the manuscript. DE performed rotarod training, data analysis, and edited the manuscript. FB contributed during planning of the experiments, interpretation of the data, and preparation and writing of the manuscript. ME assisted during planning and execution of behavioral test and edited the manuscript.