Expression of sensitized β2 nAChR subunits in VTA neurons enhances intravenous nicotine self-administration in male rats

Abstract

Ventral tegmental area (VTA) nicotinic acetylcholine receptors (nAChRs) are important for nicotine reinforcement. To determine whether and to what extent these receptors are sufficient for nicotine reinforcement, we expressed β2Leu9′Ser (i.e. sensitized) nAChR subunits in the VTA of adult male rats and assessed the nicotine dose-response relationship in intravenous self-administration (SA). β2Leu9′Ser rats self-administered nicotine doses 50 to 100 fold lower than the lowest doses that control rats would respond for. Expression of WT β2 subunits confirmed that this enhanced sensitivity to nicotine was due to the Leu9′Ser mutation in β2. Higher unit doses were associated with strong escalation in β2Leu9′Ser rats over 17 fixed ratio sessions. Escalation was minimal or absent in control rats at the same unit doses. In progressive ratio SA, β2Leu9′Ser rats exhibited higher breakpoints than control rats when the nicotine unit dose was 1.5 μg/kg/inf or higher. In intermittent access SA, β2Leu9′Ser rats exhibited response patterns very similar to control rats. By adding nicotine dose-response data, progressive ratio assays, and intermittent access results that rule out stereotypy, these data significantly extend our previous finding that nicotine activation of the mesolimbic dopamine pathway is sufficient for nicotine reinforcement.

Introduction

Volitional nicotine intake in human smokers, whether via electronic cigarettes, combusted cigarettes, or other nicotine delivery systems, involves the interplay between multiple dissociable behavioral components or processes. Ongoing intake involves nicotine’s action in primary reinforcement and reinforcement enhancement, along with conditioned reinforcement and negative reinforcement [1]. Nicotine primary reinforcement, which is necessary but not sufficient for nicotine self-administration in rats [2], is dependent on β2-containing (β2*) nAChRs in the mesolimbic dopamine system [3, 4]. By gaining a better understanding of the mechanisms by which β2 nAChRs contribute to nicotine primary reinforcement, we may be able to develop improved smoking cessation therapeutics and/or strategies.

Nicotine primary reinforcement involves nAChRs expressed in neurons of the ventral tegmental area (VTA) and substantia nigra pars compacta (SNc). The key nAChR subtypes involved in nicotine primary reinforcement are α4β2*, α6β2*, and α7 nAChRs. Whereas the latter subtype is homopentameric, α4β2* and α6β2* nAChRs are heteropentamers comprising the following individual subtypes: α4β2, α4α6β2, α4α5β2, α4α6β2β3, α6β2, α6β2β3, and possibly others. α4β2* and α6β2* nAChRs are expressed by nearly all types of VTA neurons: dopaminergic, GABAergic, and glutamatergic [5–11]. These receptors are found in somatodendritic [11] and presynaptic [12–14] compartments of these cells. α7 nAChRs are expressed on glutamatergic afferents to VTA/SNc neurons [15, 16], and are also found in somatodendritic compartments in some of these cells [9].

To gain insight into the mechanisms by which β2* nAChRs contribute to nicotine primary reinforcement, we enhance their sensitivity such that low concentrations of nicotine can selectively activate only the sensitized receptor population (see [17]). We previously employed this approach to examine α6 subunits [10, 18–23], and others have used it for in vivo studies of α4 [24–29] and α7 subunits [9, 30]. β2* pharmacological sensitization is made possible by mutagenesis of a conserved leucine (the “Leu9′” residue) in the second transmembrane α-helix. β2 subunits harboring a Leu9′ to Ser (Leu9′Ser) mutation incorporate into pentameric receptors with native α4 or α6 subunits [31], conferring increased nicotine sensitivity to the receptor [32]. We recently used this approach to examine the role of VTA β2* nAChRs in nicotine reinforcement. We found that a low dose of nicotine (1.5 μg/kg/inf), which is too low to activate native WT β2* nAChRs, is sufficient for acquisition of nicotine intravenous self-administration in rats expressing β2Leu9′Ser subunits in VTA neurons [32]. We used TH-Cre rats to confirm that this effect is mediated by β2Leu9′Ser expression in DA neurons [32]. Electrochemistry experiments indicated that these behavioral changes are associated with modulated DA release kinetics following β2Leu9′Ser expression [32]. We also demonstrated that expression of β2Leu9′Ser subunits in VTA is sufficient to enhance acquisition of cocaine IVSA, implicating ongoing cholinergic transmission in self-administration of psychostimulants [33].

These results prompted us to ask a variety of additional questions. By only testing one low nicotine dose (1.5 μg/kg/inf), we did not identify additional nicotine doses that also support acquisition of nicotine SA in β2Leu9′Ser rats. Additionally, our initial study did not examine whether low nicotine doses in β2Leu9′Ser rats promote motivated responding in a progressive ratio paradigm. To address these gaps and others, the present study systematically examined nicotine SA in control and β2Leu9′Ser rats.

Materials and Methods

Materials –

Injectable heparin sodium (catalog #00409-2720-02) was obtained from Pfizer, Inc. (via Medline). Injectable meloxicam (catalog #07-893-7565) and isoflurane (catalog #07-894-8943) were obtained from Patterson Veterinary Supply. Nicotine hydrogen tartrate salt was obtained from Glentham Life Sciences (catalog # GL9693-5G).

Molecular Biology –

The AAV5.2-hSyn-DIO-Chrnb2Leu273Ser-P2A-GFP (β2Leu9′Ser) and AAV5.2-hSyn-DIO-Chrnb2-P2A-GFP (β2WT) vectors were prepared by Virovek, Inc. Briefly, a linear DNA open reading frame containing Leu273Ser or WT Chrnb2-P2A-GFP was ligated into a shuttle vector, followed by production of a recombinant baculovirus which was used to infect insect Sf9 cells for production and purification of AAV particles (2 × 10¹³ vg/mL). AAV5-hSyn-mCherry-Cre was obtained from the University of North Carolina Vector Core Facility.

Animals –

All experimental protocols involving vertebrates were reviewed and approved by the Wake Forest University School of Medicine Animal Care and Use Committee. Procedures also followed the guidelines for the care and use of animals provided by the National Institutes of Health Office of Laboratory Animal Welfare. All efforts were made to minimize animal distress and suffering during experimental procedures, including during the use of anesthesia. A total of n=159 male SD rats (Envigo) were used. SD rats were ~300 g (approximately eight weeks old) when they arrived at our facility. Rats were housed at 22°C on a reverse 12/12 h light/dark cycle (4 P.M. lights on, 4 A.M. lights off).

Operant Chambers –

Rats were trained in Med Associates operant chambers (interior dimensions, in inches: 11.9 × 9.4 × 11.3, catalog# MED-007-CT-B1) in sound-attenuating cubicles. The SA system was housed in a dedicated room within the same laboratory suite as the rat-housing room. A PC was used to control the SA system with Med-PC IV software. Each chamber had transparent plastic walls, a stainless-steel grid floor, and was equipped on the right-side wall with two nose pokes (2.4 inches from grid floor to nose poke center) flanking a pellet receptacle coupled to a pellet dispenser. A white stimulus light was located inside each nose poke with a house light located over the chamber’s left-side wall. In all sessions, nose pokes on the active nose poke activated either the pellet dispenser or an infusion pump, respectively. Nose pokes on the inactive nose poke had no consequence. For intravenous drug infusions, each rat’s catheter was connected to a liquid swivel via polyethylene tubing protected by a metal spring. The liquid swivel was connected to a 10-mL syringe loaded onto the syringe pump.

Operant Food Training –

Rats arrived pair housed and were only separated after surgery. A week after arrival (SD control rats) or stereotaxic surgery (β2Leu9′Ser & β2WT rats), rats were food-restricted to enhance operant responding. Rats were fed standard chow (LabDiet Prolab RMH 3000 5P00, catalog# 0001495), (20 g per rat, 40 g per cage) once per day at least 1 h after testing ended. Water was available ad libitum except during operant behavioral sessions. Food training sessions were 1 h in duration, and rats were trained to nose poke for food pellets (45 mg; Bio-Serv Dustless Precision Pellets, catalog# F0021) on the same nosepoke that would later be paired with nicotine or saline infusions. A fixed ratio 1 (FR1; no timeout) schedule was used for food training. No visual cues (stimulus, house light) were available during the session and rats could earn a maximum of 75 food pellets during a 1 h session. Once 50 pellets with a least a 2:1 preference for the active nose poke relative to the inactive nose poke was achieved for two sessions, no further food training was conducted. These criteria were typically met after 3–5 d.

Stereotaxic Surgery –

After arriving and acclimating to our facility for 6–7 d, rats were anesthetized with isoflurane (3% induction, 2–5% maintenance) and were introduced into a stereotaxic rat frame. Rats received a bilateral injection into the VTA using coordinates from bregma (AP: −5.40, ML: ±0.70, DV: −8.20) derived from “Brain Maps 4.0” (Larry Swanson; University of Southern California). Virus was infused with a 22-gauge Hamilton injection syringe at a rate of 100 nL/min. After injection, the injection needle was left in place for 5 min prior to retraction. Meloxicam (2 mg/kg) was administered immediately following surgery and for the next 2 d to reduce inflammation and pain. Rats were then given 6–7 d for recovery. Rats were singly housed following surgery and throughout the rest of their lifetimes.

Indwelling Jugular Catheter Surgery –

After acquiring food operant responding, rats were anesthetized with isoflurane (3% induction, 2–3% maintenance) and implanted with indwelling jugular catheters (Instech, catalog# C30PU-RJV1402). Meloxicam (2 mg/kg) was administered immediately following surgery and for the next 2 d to reduce inflammation and pain. Rats were then singly housed and were given 7 d for recovery from surgery with catheters being flushed daily with heparin sodium dissolved in sterile saline.

Intravenous Drug Self-Administration –

After recovery from catheter surgery, rats were allowed to self-administer nicotine in 2 h SA sessions, Monday through Friday (no SA sessions occurred on weekends). Rats were weighed biweekly and infusion duration varied (between 2.7–3.4 seconds depending on rat body weight). Nicotine hydrogen tartrate salt (Glentham Life Sciences) was dissolved in sterile saline and had pH adjusted between 7.4–7.6. The dose of nicotine was kept constant for each rat throughout each SA session but involved different cohorts on different doses (in μg/kg: 0.08, 0.21, 0.56, 1.5, 4.05, 11.3, 30). Infusions were delivered by an infusion pump and were paired with illumination of the stimulus light inside the active nose poke. Each infusion resulted in the house light turning off which signaled a 20 sec timeout period (TO 20s) during which responding was recorded but had no consequences. In all sessions, inactive nose pokes were recorded but had no consequence. Rats were removed from training chambers as soon as possible after the end of the 2 h session. Upon being returned to their home cage, rats were given 20 g of standard lab chow daily and approximately at the same time on weekends despite not having SA sessions.

Rats were allowed to self-administer different doses of nicotine for 17 d on an FR1TO20s schedule of reinforcement. Our inclusion/exclusion criteria for SD control and β2Leu9′Ser rats is described in detail in Results. Briefly, we used saline SA behavior in each group (SD control and β2Leu9′Ser) to establish a minimum number of infusions that a rat had to self-administer during sessions 4–10 of the FR1TO20s paradigm. If the mean number of nicotine self-infusions during sessions 4–10 fell below a certain value, the rat was deemed to have failed acquisition. In a few cases, rats were removed from experiments based on overt signs of a catheter failure. These signs include a dramatic drop in infusions, physical signs of pain/distress during daily flushes, and significant pooling of fluid around the catheter back plate following SA or flushing. After FR1TO20s, rats were switched to a progressive ratio (PR) paradigm for 5 final sessions of SA with sessions lasting for 2 h regardless of activity. The PR schedule used in this study was derived from the formula 5e^(0.2*infusion number)-5 [34]. Based on this formula, the response requirement for an infusion increased as follows: 1, 2, 4, 6, 9, 12, 15, 20, 25, 32, 40, 50, 62, 77, 95, 118, 145, 178, 219, 268, etc. Breakpoint was then defined as the last completed ratio prior to the end of the session. Of 159 rats that started the study, 35 failed to meet the above criteria, 2 failed due to catheter patency, and 1 outlier was excluded, resulting in a total of 38 rats being excluded and leaving a total of 121 rats in this study.

Immunohistochemistry and Fluorescence Microscopy –

Rats were deeply anesthetized with isoflurane and perfused transcardially with 50 mL of PBS followed by 250 mL of 4% paraformaldehyde (PFA) in PBS. Brains were removed and stored at 4°C overnight in PBS containing 4% PFA and 4% sucrose (to dehydrate tissue). Coronal sections (20 μm) were cut in triplicate on a microtome and collected into 12-well tissue culture plates filled with PBS and 0.1% sodium azide. Slices were washed in PBS with 0.3% Triton X-100, blocked in PBS with 0.1% Triton X-100 and 5% donkey serum for 1 h at 4°C, and incubated overnight at 4°C in PBS, 0.1% triton, and 5% donkey serum with primary antibody (rabbit-anti-GFP [A11122, Invitrogen] at 1:1000).

Slices were then washed three times for 10 minutes in a mix of PBS and 0.3% Triton X-100 before a 1-hr incubation in PBS, 0.1% triton, and 5% donkey serum with secondary antibody (chicken-anti rabbit Alexa Fluor 488 [A21441, Invitrogen] at 1:500), and three subsequent washes. Sections were immediately mounted and coverslipped with sodium bicarbonate then imaged on an Olympus IX83 inverted microscope equipped with a 20X UPLXAPO objective (0.8 NA) and a 16-bit (2048 × 2048 pixel frame size) Hamamatsu ORCA-Flash 4.0 camera. Tiled GFP images were stitched together in Olympus cellSens Dimension software.

Experimental Design and Statistical Analysis –

SA data files, produced by Med Associates MedPC IV software, were processed and analyzed with custom scripts written in MATLAB or Graphpad Prism 10. Scalable vector graphics were produced from MATLAB figures using functions written by Salva Ardid (https://github.com/kupiqu/fig2svg).

Results

To examine the effect of nicotine dose on operant self-administration in rats expressing gain-of-function β2* nAChRs, we microinjected male Sprague Dawley rats with AAVs expressing Cre and Cre-dependent β2Leu9′Ser nAChR subunits (Fig. 1A). After stereotaxic injections, rats recovered for 10–14 days and then underwent jugular catheter implantation surgery followed by an additional 7 days of recovery (Fig. 1B). Injection location in VTA was verified via anti-GFP immunostaining (Fig. 1C). The approximate injection site in the VTA was assessed for β2Leu9′Ser rats (n=21) (Fig. 1D). After recovery from surgery, rats were allowed to self-administer nicotine on a fixed ratio 1 schedule of reinforcement for 17 sessions. The dose of nicotine used for each group is specified below and is further discussed in Materials & Methods. After FR1 SA, rats were allowed to self-administer nicotine on a progressive ratio (PR) schedule for 5 sessions. In our recent study introducing the β2Leu9′Ser rat IVSA approach [32], we found that 1.5 μg/kg/inf was sufficient to establish robust IVSA in β2Leu9′Ser rats co-injected with Cre-expressing AAVs, but not in β2Leu9′Ser rats where Cre was omitted. To more directly show that self-administration of 1.5 μg/kg/inf nicotine is due to hypersensitive β2-containing nAChRs and not due to viral-mediated overexpression, two experiments were conducted. First, we recently used radioligand binding to study assembled, presynaptic β2* nAChRs in control and β2Leu9′Ser rats. We confirmed that there is no overexpression of assembled, mature nAChRs in this model [33]. Second, we designed another AAV for Cre-dependent expression of β2WT nAChR subunits. This vector is identical to our β2Leu9′Ser vector except that the codon for the Leu9′ residue was not changed to Ser as it was in the β2Leu9′Ser vector. We reasoned that if β2Leu9′Ser rats acquire nicotine SA at 1.5 μg/kg/inf because β2 is overexpressed and not because β2* nAChRs have gained sensitivity to nicotine, then AAV-mediated expression of β2WT subunits should also lead to nicotine SA acquisition at this dose. SD rats were microinjected in VTA with β2WT AAVs, catheterized, and allowed to self-administer 1.5 μg/kg/inf nicotine. Consistent with our contention that β2* nAChR sensitivity – not β2 subunit overexpression – is responsible for acquisition of nicotine IVSA at subthreshold doses, β2WT rats failed to acquire nicotine IVSA at 1.5 μg/kg/inf (Fig. 1E).

Figure 1.

SA acquisition in β2Leu9′Ser rats: dose response. A. Adeno associated virus for Cre-dependent expression of β2Leu9′Ser subunits. B. Nicotine SA paradigm. Stereotaxic surgery and jugular catheter surgery were conducted, then rats were allowed to self-administer nicotine (dosage variable) for 17 sessions at FR1 and then 5 additional sessions on a progressive ratio (PR) schedule on the same dose used for FR1. C. β2L9′S GFP reporter expression in VTA was assessed via anti-GFP immunostaining. D. Summary of VTA targeting for a subset (n=21) of β2Leu9′Ser rats. An opaque red circle was used to denote the approximate injection location, along with viral spread, in VTA. Overlapping circles (i.e. consistent VTA targeting) is indicated by darker red color. E. β2WT overexpression does not support acquisition of nicotine SA at 1.5 μg/kg/inf. SD rats were microinjected in VTA with AAVs for β2Leu9′Ser or β2WT and allowed to self-administer nicotine (FR1) for 17 sessions at 1.5 μg/kg/inf. Mean (±SEM) infusions are shown for both groups across FR1 sessions. F-G. Left-shifted nicotine SA dose response in β2Leu9′Ser rats. Nicotine infusions (F) and active/inactive responses (G) are shown for β2Leu9′Ser and SD control rats self-administering the indicated dose. Mean (±SEM) data for sessions 4–10 of FR1 acquisition are shown; a different group of rats were tested for each dose.

To rigorously determine whether/how β2Leu9′Ser rats are more sensitive to the reinforcing effects of nicotine, we conducted a dose-response study. To establish an acquisition criterion for inclusion or exclusion of control and β2Leu9′Ser IVSA data in the following experiments, we examined saline IVSA data from the lab. We allowed n=11 male β2Leu9′Ser rats to self-administer saline for 17 sessions. We analyzed self-infusions for sessions 4–10, yielding a mean (± 95% CI) of 6.8 (± 2.0) saline infusions. Control SD rats (n=25) self-administered a mean (± 95% CI) of 11.0 (± 2.6) saline infusions. Using a value of [mean # of infusions + 95% CI], we established that an animal positively acquired nicotine IVSA if their mean number of nicotine self-infusions for sessions 4–10 exceeded 13.6 (for SD control) or 8.8 (for β2Leu9′Ser). Separate groups of SD rats (control rats receiving no virus surgery or β2Leu9′Ser rats) were allowed to acquire SA on particular nicotine doses for 17 sessions followed by 5 sessions of PR on the same dose. For control SD rats, the following number of rats met the above inclusion criterion for each indicated dose group: 1 of 10 (1.5 μg/kg/inf), 6 of 13 (4.05 μg/kg/inf), 8 of 8 (11.3 μg/kg/inf), 6 of 8 (30 μg/kg/inf). With the exclusion of data from rats whose nicotine SA is not substantially greater than saline SA, we observed a monotonic nicotine dose response relation where the number of self-infusions (Fig. 1F) and active responses (Fig. 1G) was inversely related to nicotine dose. In β2Leu9′Ser rats, the following number of rats met the above inclusion criterion for each indicated dose group: 6 of 12 (0.08 μg/kg/inf), 7 of 14 (0.21 μg/kg/inf), 10 of 14 (0.56 μg/kg/inf), 14 of 14 (1.5 μg/kg/inf), 12 of 12 (4.05 μg/kg/inf), 9 of 10 (11.3 μg/kg/inf), 6 of 7 (30 μg/kg/inf). β2Leu9′Ser rats also exhibited an inverse relationship between nicotine self-infusions (Fig. 1F) and active responses (Fig. 1G) vs. nicotine dose, but the relation extended further to the left vs. control SD rats due to acquisition of SA at much lower doses of nicotine. These data substantially advance our previous understanding of nicotine reinforcement in β2Leu9′Ser rats, suggesting that these rats are 10–100 fold more sensitive to some or all of the reinforcing properties of nicotine.

In several nicotine dose groups, mean infusions in β2Leu9′Ser rats appeared to escalate over the 17 sessions of SA. For example, mean infusions in the β2Leu9′Ser 4.05 μg/kg/inf nicotine dose group steadily increased from an initial value of ~10 infusions on day 1 to ~80 infusions by day 17 (Fig. 2A). Only a very small fraction of SD control rats appeared to escalate their intake, as described further below. To objectively evaluate escalation of nicotine SA in individual rats among the various dose groups, we fitted a linear function to the self-infusion data across 17 sessions of FR1 nicotine SA in β2Leu9′Ser and SD control rats in each dose group. The linear regression r² value for each rat in this analysis quantitatively reflects our qualitative observation (Fig. 2A) of escalation vs. no escalation. We confirmed that no rats exhibited a steadily decreasing number of infusions, which would also generate a high r² value (but such rats would have failed the acquisition assessment described above). For all β2Leu9′Ser and SD control rats in each dose group, the escalation r² value was plotted. An arbitrary value of r² = 0.5 was used to segregate rats as escalated or non-escalated (Fig. 2B). Escalation appeared to be related to nicotine dose; few rats exhibited escalation at low nicotine doses of 0.08 μg/kg/inf and 0.21 μg/kg/inf, intermediate doses (0.56 and 1.5 μg/kg/inf) were associated with ~50% escalation, and rats self-administering higher doses (4.05, 11.3, and 30 μg/kg/inf) were also those that were very likely to escalate (Fig. 2B). SD control rats exhibited minimal escalation using the same analysis approach. Only n=1 of 6 SD control rats in the 4.05 μg/kg/inf group and n=2 of 8 SD control rats in the 11.3 μg/kg/inf group escalated their nicotine intake over 17 sessions of SA (Fig. 2B).

Figure 2.

Escalation of nicotine intake in β2Leu9′Ser rats. A. Daily mean (±SEM) nicotine infusions for β2Leu9′Ser and SD control rats self-administering 4.05 μg/kg/inf. B. Nicotine SA escalation in β2Leu9′Ser and SD control rats. Linear regression r² values are shown for individual β2Leu9′Ser and SD control rats in each dose group. A linear function was fitted to infusion data (sessions 1–17) for each rat and the resultant r² value for each fit is plotted.

As described in the experimental workflow (Fig. 1A), β2Leu9′Ser and SD control rats were transitioned to PR nicotine SA after 17 sessions of FR1 SA. Rats self-administered the same nicotine dose in PR that they previously acquired on in FR1 sessions. β2Leu9′Ser rats on doses of (in μg/kg/inf) 0.08, 0.21, and 0.56 self-administered ~5 infusions during PR, corresponding to a breakpoint of 9 nose pokes (Fig. 3A). Comparatively, β2Leu9′Ser rats on doses of 1.5 μg/kg/inf and above sharply increased their infusions in PR. The highest 4 nicotine dose groups (1.5, 4.05, 11.3, and 30 μg/kg/inf) averaged 11–14 infusions per day for 5 sessions, corresponding to mean breakpoints between 40–77 nose pokes (Fig. 3A). SD control rats self-administered 5–7 infusions per day for a breakpoint of 9 to 15 nose pokes (Fig. 3B). For β2Leu9′Ser rats, plotting mean infusions as a function of nicotine dose confirms that PR infusions are similar to SD controls at the 3 lowest nicotine doses (0.08, 0.21, and 0.56) but suddenly increase and reach a new stable value for nicotine doses of 1.5 and above (Fig. 3C). We compared the average cumulative response plot for β2Leu9′Ser and SD controls at the same nicotine dose (11.3 μg/kg/inf). The ratio of active to inactive responses appeared similar in both groups (Fig. 3D,E), but β2Leu9′Ser rats emitted 8–10 times the number of active responses as SD controls, consistent with the difference in infusions at this dose (Fig. 3C). These data, particularly the inactive response data, confirm that β2Leu9′Ser rats are not responding indiscriminately during PR sessions.

Figure 3.

Progressive ratio nicotine SA is augmented in β2Leu9′Ser rats. A-B. Daily mean nicotine infusions for β2Leu9′Ser (A) and SD control (B) rats during PR sessions at the indicated nicotine dose. C. PR nicotine SA dose response. Mean (±SEM) infusions across PR sessions 1–5 are plotted for each β2Leu9′Ser and SD control dose group. D-E. Cumulative response plots for active and inactive responses during PR sessions in the 11.3 μg/kg/inf dose group in β2Leu9′Ser and SD control rats. Mean (±SEM; shaded region) cumulative responses are plotted on the same scale to highlight the difference between β2Leu9′Ser and SD control.

Given that β2Leu9′Ser rats harbored a modification in the DA system, we asked whether stereotypy may be playing a role in the amount of responding we observed during PR. A drug-naïve group of β2Leu9′Ser rats were catheterized and allowed to acquire nicotine SA on a dose of 1.5 μg/kg/inf for 10 sessions on a FR1 schedule of reinforcement (2-h sessions). Following this, rats were transitioned to an intermittent access (IntA) paradigm [35, 36] (Fig. 4A). In this paradigm, there were regular periods where drug was available on a FR1 schedule and there were intervening periods where drug was not available; nicotine was available for 5 min, then unavailable for 15 min, then available for 5 min, etc. The house light signaled whether drug was available or not: the house light was on when drug was available and off when it was unavailable. Our use of nose pokes proved advantageous, as rats were able to make active responses even when the drug was unavailable. IntA sessions were 4-h with only 1-h of time with access to nicotine. We reasoned that if β2Leu9′Ser responding was due to stereotypy, they would continue aimlessly responding on the active nose poke even when drug was not available. β2Leu9′Ser rats acquired nicotine SA as expected on 1.5 μg/kg/inf during the first 10 sessions. They also successfully transitioned to IntA nicotine SA, exhibiting stable self-infusions (Fig. 4B) and nicotine intake (Fig. 4C) over the 7 sessions of IntA. Active (Fig. 4D) responses largely mirrored infusions, and inactive (Fig. 4E) responses were low. A group of β2Leu9′Ser rats self-administering saline for 10 sessions of acquisition and 7 sessions of IntA served as the primary negative control. During IntA, β2Leu9′Ser rats showed multiple signs that they learned the new contingencies and they did not show signs of stereotypy. Fig. 4F shows a raster plot of all rewarded (red ticks) and unrewarded (blue ticks) active responses for all β2Leu9′Ser rats on day 7 of IntA. Responding was usually confined to the nicotine-available period, but some rats responded for nicotine during one or more of the inter-trial intervals (ITI; Fig. 4F). When we examined active, unrewarded responding as a function of the ITI number during the 4-h IntA session, we noted that active unrewarded responding was 2–4X more likely during the first 3 ITIs compared to the final 9 ITIs (Fig. 4G). Aimless, stereotypic responding would likely have been more uniformly distributed across the 12 ITIs of the IntA session.

Figure 4 –

Intermittent access nicotine SA in β2Leu9′Ser rats. A. Intermittent access paradigm. Stereotaxic surgery and jugular catheter surgery were conducted, then β2Leu9′Ser rats were allowed to self-administer nicotine (1.5 μg/kg/inf) or saline for 10 sessions at FR1 and then 7 additional sessions on an intermittent access (IntA) schedule on the same drug and dose used for FR1. B-C. Nicotine infusions and intake during IntA. Mean (±SEM) nicotine infusions (B) and intake (C; in mg/kg) are shown for β2Leu9′Ser rats on nicotine or saline. D-E. Mean (±SEM) active (D) and inactive (E) responses are shown for infusion/intake data shown in B-C. F. Raster plot of responses during IntA day 7. Each rat’s final IntA session (IntA day 7) is shown as a raster plot, with vertical tick marks denoting a response on the active or inactive nose poke (red = active response, rewarded; blue = active response, not rewarded; black = inactive response, not rewarded). Each 5-min trial (drug-available period) is shown in gray with the intervening inter-trial interval shown in white. G-I. Probability analysis of β2Leu9′Ser nicotine IntA SA. Infusion probability was estimated across all 5-min drug-available trial periods (G) for day 7 of IntA in β2Leu9′Ser rats. The time of each infusion during the trial is denoted in the raster plot at the bottom of the plot. Kernel density estimation (KDE) was used to estimate the probability density of infusions (plotted as a continuous function) during the trial. For all 12 ITIs on day 7 of IntA, the relative number of active responses is shown as a histogram (H). Similar to (G), KDE was used to estimate the probability density function for active (unrewarded) responses occurring during the 15-min ITI (I).

We then analyzed the time when infusions or active responses occurred across all animals during IntA session 7. We pooled all 12 trials or ITIs to enable kernel density estimation. During the 5-min nicotine-available period, the probability of an infusion is relatively constant (Fig. 4H). Upon examination of the active response time distribution during the ITI, we noted that while such responses occur relatively stably over the first 12 min of the ITI, the probability of such responses spiked in the last 3 min (Fig. 4I). This anticipatory responding as the next drug-available period approaches would not be expected if β2Leu9′Ser rats are responding for nicotine in a stereotypical manner. These data suggest that β2Leu9′Ser rats respond for nicotine in a manner shaped by the contingencies, that they learn to ignore the operandum when the reinforcer is not available, and that they actively anticipate and seek the drug reinforcer when its availability is imminent.

Discussion

In this paper, we confirmed and extended our earlier finding [32] that β2* nAChR activation in VTA neurons is sufficient for nicotine SA. By systematically examining a range of nicotine doses, we show that the dose-response relation for nicotine SA is substantially left-shifted in β2Leu9′Ser rats compared to controls. β2Leu9′Ser rats also exhibit dose-dependent escalation in responding during the 17 day FR1 paradigm. The lowest doses tested in β2Leu9′Ser rats do not support strong motivated responding in PR sessions, but doses at or above 1.5 μg/kg/inf are associated with much higher PR responding. Intermittent access FR1 SA experiments were conducted to rule out stereotypy as an explanation for nicotine SA in β2Leu9′Ser rats. These results add to our finding that VTA β2* nAChR activity is sufficient to explain nicotine SA acquisition, maintenance, and motivated responding in rats.

Enhanced Acquisition and Escalation of Nicotine SA:

nAChR responsiveness to specific agonist concentrations can be enhanced via several mechanisms, including increased receptor numbers. This is evident in the medial habenula, whose neurons respond strongly to nicotine or ACh due largely to exquisitely high levels of low-sensitivity β4-containing receptors [37, 38]. Because our previous studies with β2Leu9′Ser expression did not address the possibility that the enhancement in nicotine SA could be due simply to overexpression of β2 subunits rather than enhanced sensitivity, we directly compared AAV-mediated expression of β2Leu9′Ser to β2WT subunits. All other aspects of the experiment (AAV microinjection, β2 subunit expression, assembly with other nAChR subunits, etc.) being equal, a single Leu to Ser amino acid change and the resultant increase in nicotine sensitivity is sufficient to explain our result that rats acquire nicotine SA at 1.5 μg/kg/inf (Fig. 1E).

SD rats emit operant responses for cue delivery alone [39], which can be examined via saline SA sessions using cue conditions identical to those used for nicotine SA sessions. As described in Results, we used saline SA results in SD control and β2Leu9′Ser rats to establish a threshold for inclusion/exclusion of rats self-administering nicotine. This was important especially for low nicotine doses in β2Leu9′Ser rats. Since there is no prior data in the literature on low nicotine doses in β2Leu9′Ser rats, there would be no objective rationale for selecting an arbitrary (i.e. 10 infusions in 1 h) infusion cutoff. When we used saline SA to identify rats that acquired nicotine SA, we found that infusions and active responses were inversely correlated with nicotine dose. At the first dose above some threshold level of nicotine, infusions and active responses were at or near their highest level (Fig. 1F,G). Each nicotine dose above this threshold level was associated with progressively fewer infusions and active responses. This was true for both SD control and β2Leu9′Ser rats but at very different threshold doses; β2Leu9′Ser infusions peaked at ~0.08 μg/kg whereas infusions peaked at 4.05 μg/kg in SD control rats – a 50-fold difference. Fewer animals acquired at or near the threshold dose compared to higher doses such as 11.3 or 30 μg/kg/inf.

Our finding of an abrupt rise in self-infusions at the lowest doses of nicotine is consistent with a study by Neisewander and colleagues [40], who found that SA was maximal in male Sprague Dawley rats at 15 μg/kg/inf and declined as the dose increased. This study shared two key methodological details with ours that are likely important in explaining the similarity in findings. First, both studies employed a careful approach to determine whether individual animals had truly acquired nicotine SA. Second, both studies used an FR1 schedule of reinforcement. Donny and colleagues also reported an inverse relationship between nicotine dose and self-injections using an FR5 schedule of reinforcement [41]. The shape of our nicotine dose response curve contrasts with some studies that report an “inverted U” dose response relationship. For example, Kenny and colleagues have reported an inverted U nicotine dose response in rats [42, 43]. Methodological differences exist, however. We employed a fixed ratio 1 (FR1) schedule of reinforcement whereas the studies noted above use an FR5 schedule. Further, the SA acquisition criteria are unclear in these studies and do not appear to use saline SA to assess drug SA acquisition.

Our nicotine dose response relation data are plausible when one considers several rat SA studies examining cocaine, another psychostimulant. Nestler et al. reported a declining cocaine SA dose-response curve without an ascending limb [44]. Ahmed et al. examined cocaine SA acquisition and showed that dose-response behavior for individual rats is discontinuous at a threshold cocaine dose [45]; below each rat’s threshold dose there is no cocaine SA whereas at this threshold, SA spikes to its highest level and declines gradually at higher doses [45]. A key insight from this study is the hypothesis that the threshold cocaine dose is a random variable, potentially with a normal distribution in a population or a sample of rats [45]. Martin et al. reported a similar result: low doses of cocaine were either self-administered at high rates or not at all [46]. Without accounting for the variability in the SA acquisition threshold dose, it is easy to generate an inverted U dose response curve with an “ascending limb”. Norman and colleagues [47] posited that the ascending limb of the cocaine SA dose response curve is indeed an artifact generated by averaging high responding with minimal responding.

β2Leu9′Ser rats showed prominent, dose-dependent escalation of nicotine SA across 17 sessions of FR1 sessions (Fig. 2B). Less than 50% of rats exhibited escalation at or below 0.56 μg/kg/inf, whereas most or even all rats showed escalation at 1.5 μg/kg/inf or above (Fig. 2B). Interestingly, this increase in escalation at or above 1.5 μg/kg/inf is similar to the dose-response pattern we observed in PR. Although it is tempting to speculate about the cause of this escalation behavior in β2Leu9′Ser rats, there are many possible, non-exclusive causes. Escalation could be due to one or more of the following: enhanced motivation over time, improved learning of the cue-reward contingencies [48, 49], tolerance to the locomotor suppressive effects of nicotine [26], sensitization of nicotine-mediated dopamine release [50], sensitized nicotine-mediated reinforcement enhancement of the visual cues associated with reward delivery [2, 51–53], and sensitized responding for the visual cues alone [54]. Sensitization of nicotine-stimulated dopamine signaling is the most likely cause; escalation behavior in β2Leu9′Ser rats appears to be related to the dose of nicotine, which is likely to be related to the magnitude or duration of nicotine-stimulated dopamine release. Our viral approach non-selectively infects VTA neurons, which includes DA and intermingled GABA neurons. We expect nicotine to activate β2Leu9′Ser nAChRs in both cell types, resulting in DA neuron stimulation as well as inhibition of DA neurons by VTA GABA neurons [11, 55–59].

Motivated Responding and Intermittent Access in β2Leu9′Ser Rats:

Progressive ratio SA is useful for exploring drug reinforcement, as it often reveals aspects of drug taking behavior that would not be predicted by fixed ratio paradigms. For example, cocaine PR breakpoint is maximized at high unit doses of cocaine that are associated with relatively low rates of responding in a fixed ratio paradigm [34, 60]. Donny et al. also reported an inverse relationship between nicotine unit dose in FR and PR paradigms [41]. Our SD control PR dose response data are consistent with Donny’s original PR study. For example, 30 μg/kg/inf is associated with fewer infusions in FR1 sessions (Fig. 1F) but elicits the greatest number of infusions in PR sessions (Fig. 3C). In PR sessions in β2Leu9′Ser rats, the 3 lowest unit doses (in μg/kg/inf: 0.08, 0.21, 0.54) elicited similar responding to SD controls on their respective 3 doses (in μg/kg/inf: 4.05, 11.3, 30) (Fig. 3C). Notably, PR infusions abruptly increased at/above 1.5 μg/kg/inf in β2Leu9′Ser rats (Fig. 3C). Mean infusions did not further increase with higher unit doses (Fig. 3C), possibly suggesting a ceiling effect in β2Leu9′Ser rats. PR responding in β2Leu9′Ser and SD control rats was strongly directed toward the active nose poke, which argues against non-specific or stereotypical responding.

Intermittent access nicotine SA in β2Leu9′Ser rats was used to further address the possibility of stereotypy. This approach takes advantage of 1) multiple periods of access/no access during a single session, and 2) nose poke operanda which remain available for rats to respond to even when drug availability has been temporarily revoked. During IntA sessions, β2Leu9′Ser rats responded to 1.5 μg/kg/inf similarly to SD controls responded to 30 μg/kg/inf in our prior study [61]. As in that prior study, β2Leu9′Ser rats had little interest in the inactive nose poke (Fig. 4E) and quickly learned to disregard the active nose poke when drug became unavailable (Fig. 4H). The use of nose pokes allowed us to visualize overall responding on the active nose poke during the period when drug was unavailable. This analysis revealed an anticipatory increase in responding when the next drug-available period was imminent (Fig. 4I). Stereotypical responding is expected to look qualitatively different from these results. In particular, stereotypy is expected to manifest as operant responding without regard to contingencies such as cues, drug delivery, etc. [62]. If the β2Leu9′Ser mutation were inducing dopamine release and stimulating operant responding without regard to reinforcement, we would not expect β2Leu9′Ser rats to have learned to disregard the active nose poke during the inter-trial interval (Fig. 4H). We would also not expect them to emit responses in anticipation of the next period when nicotine is available (Fig. 4I).

Limitations and Future Studies:

Our exclusive use of male rats is a limitation of our study. Compared to males, female rats often display greater sensitivity to various neuropharmacological effects of nicotine [63]. Although female rats would be expected to respond to β2Leu9′Ser expression similarly to the way males responded, they may have responded to doses lower than 0.08 μg/kg/inf. Additionally, our PR dose response results are limited by the fact that each group first acquired nicotine SA on a different dose in FR1. Each group’s cumulative nicotine exposure and number of cue-reward pairings was different when PR sessions began, which may have had an impact on dose-dependent responding during PR. PR sessions were 2 h in duration, which may have caused an underestimation of breakpoints in some animals. Overall, our results provide useful mechanistic information into nicotine reinforcement mediated by the mesolimbic dopamine pathway. Gain-of-function β2* nAChR activity in this pathway is sufficient to convert nicotine into a more potent reinforcer in both fixed ratio and progressive ratio paradigms. Future studies that could provide more insight into these results include 1) examination of the relative contribution of β2* nAChRs on VTA neuron somata vs. presynaptic terminals, 2) PR dose-response studies in rats that acquired nicotine SA under the same conditions, 3) examination of dopamine release in β2Leu9′Ser rats during self-administration sessions, and 4) exploration of positive and negative allosteric modulation of β2* nAChRs in nicotine self-administration.

HIGHLIGHTS.

• Male rats were made with hypersensitive β2 nicotinic acetylcholine receptors in ventral tegmental area neurons and their projections.

• Rats expressing ectopic β2WT subunits do not acquire intravenous nicotine self-administration faster at low doses.

• β2Leu9′Ser rats acquire nicotine self-administration at doses 10- to 50-fold lower than control rats.

• β2Leu9′Ser expression promotes escalation of nicotine self-administration.

• In progressive ration sessions, β2Leu9′Ser rats work harder for lower doses of nicotine than control rats.