Nicotine vapor inhalation escalates nicotine self-administration
Abstract
Humans escalate their cigarette smoking over time, and a major obstacle in the field of pre-clinical nicotine addiction research has been the inability to produce escalated nicotine self-administration in rats. In experiment 1, male Wistar rats were trained to respond for nicotine in 2-hour operant sessions, then exposed to chronic intermittent (12 hours/day) nicotine vapor and repeatedly tested for nicotine self-administration at 8–12 hours of withdrawal. Rats were tested intermittently on days 1, 3 and 5 of the vapor exposure procedure, then tested with nicotine vapor exposure on 6–15 consecutive days. Rats exhibited transient increases in operant nicotine responding during intermittent testing, regardless of vapor condition, and this responding returned to baseline levels upon resumption of consecutive-days testing (i.e. nicotine deprivation effect). Nicotine vapor-exposed rats then escalated nicotine self-administration relative to both their own baseline (∼200% increase) and non-dependent controls (∼3× higher). In experiment 2, rats were exposed or not exposed to chronic intermittent nicotine vapor, then tested for spontaneous and precipitated somatic signs of nicotine withdrawal. Eight hours following removal from nicotine vapor, rats exhibited robust mecamylamine-precipitated somatic signs of withdrawal. There was a strong correlation between nicotine flow rate and air–nicotine concentration, and the air–nicotine concentrations used in experiments 1 and 2 resemble concentrations experienced by human smokers. Collectively, these results suggest that chronic intermittent nicotine vapor inhalation produces somatic and motivational signs of nicotine dependence, the latter of which is evidenced by escalation of nicotine self-administration.
Introduction
Nicotine addiction is responsible, either directly or indirectly, for millions of deaths worldwide each year [National Institutes of Health (NIH), World Health Organization]. The financial cost of nicotine-related problems to US society alone was recently estimated at $250 billion annually (NIH), a majority of which is due to negative biological outcomes (e.g. cancers and cardiovascular disease). Although nicotine itself is not always the compound responsible for health problems and mortality associated with smoking, nicotine is the psychoactive ingredient that produces addiction to smoking and understanding the neurobiology of this addictive behavior is critical for developing smoking cessation treatments.
Rats have long been used as an animal model of nicotine self-administration that mimics acquisition, maintenance and relapse-like behaviors in human smokers. Rats exhibit reliable intravenous (i.v.) self-administration of nicotine during limited-access operant sessions in which they are allowed to press a lever for nicotine infusions, and will adjust their lever-pressing behavior to account for increasing or decreasing unit doses of nicotine (O'Dell & Koob 2007). Rats allowed long periods of access (up to 23 hours/day) to nicotine self-administration do not typically exhibit escalation, often defined as higher intakes during the first hour of access relative to short-access rats (Paterson & Markou 2004). That said, increases in long-access self-administration over time are facilitated by (1) intermittence of testing and (2) co-administration of a monoamine oxidase inhibitor (MAOI), which is another ingredient in cigarettes (Cohen, Koob & George 2012). Indeed, rats exhibit a nicotine deprivation effect, defined as a transient increase in nicotine responding following a period without nicotine access, under some (i.e. long-access) conditions (George et al. 2007; O'Dell & Koob 2007). In sum, achieving escalated voluntary nicotine self-administration in rats has been an obstacle for animal models that aim to mimic patterns of escalated smoking observed in humans (e.g. Kim, Fleming & Catalano 2009). Therefore, the aim of the current investigation was to utilize a route (inhalation) of chronic nicotine administration with face validity for the human condition, without co-administration of other drugs, to produce escalation of nicotine self-administration in rats with short access to nicotine self-administration.
Machines have existed for some time to expose rodents to cigarette smoke in quantities representative of first-hand or second-hand smoke, most often to assess the biological consequences of exposure to the carcinogens contained in cigarettes (Griffith & Standafer 1985). More recently, a nicotine vapor model has been described that exposes rats to pure nicotine in breathing air on a pattern that can be controlled by the experimenter (George et al. 2010), similar to what has been described for alcohol vapor inhalation procedures (Gilpin et al. 2008). This nicotine vapor model is sufficient to produce somatic signs of withdrawal (WD) following systemic injection of the non-specific nicotinic acetylcholine receptor (nAchR) antagonist, mecamylamine. In these experiments, we hypothesized that chronic intermittent (12 hours/day) nicotine vapor inhalation would increase nicotine self-administration during vapor WD, and also produce somatic signs of dependence.
Methods
Subjects
Twelve adult male Wistar rats were used in experiment 1 and 11 male Wistar rats were used in experiment 2, all obtained from Charles River (Kingston, NY, USA). Animals were group housed in standard plastic cages with wood chip bedding under a 12-hour light/12-hour dark cycle (lights off at 8:00 am). Animals were given ad libitum access to food and water throughout except during experimental test sessions. All procedures were conducted in the dark cycle and met the guidelines of the NIH Guide for the Care and Use of Laboratory Animals.
Experiment 1
Male Wistar rats (352.76 ± 6.61 g body weight at start of self-administration) were handled for 4–5 days then implanted with indwelling jugular catheters. Rats were anesthetized with isoflurane and implanted with a silastic catheter (0.3 mm i.d. × 0.64 mm o.d., Dow Corning Co. Midland, MI, USA) into the right external jugular vein under sterile conditions. The distal end of the catheter was subcutaneously threaded to the back of the rat where it exited via metal guide cannula (22 ga, Plastics One, Roanoke, VA, USA) anchored to the back of the rat. After surgery, rats were injected once with an analgesic (Flunixin, 2.5 mg/kg s.c., Sigma-Aldrich, St. Louis, MO, USA). Starting at 4 days post-surgery, catheters were flushed once daily with cefazolin (20 mg i.v., Sigma-Aldrich) in heparinized saline (30 U/ml, 0.1 ml total volume). On self-administration days, catheters were flushed once immediately prior to operant sessions with heparinized saline (0.1 ml) and once after sessions with cefazolin in heparinized saline (0.1 ml).
Following 5 days of recovery, rats were allowed 12 consecutive days of 2-hour daily sessions in which they pressed an active lever for i.v. nicotine infusions (0.03 mg/kg/100 μl/1 second, free base, FR-1, timeout 20 seconds) or an inactive lever that had no consequence. Nicotine solution was prepared twice per week (to account for changing body weights) by dissolving nicotine hydrogen tartrate salt in saline. No food or water was available during 2-hour sessions, but rats were never otherwise food deprived during training. Across the last 3 days of baseline rats exhibited 7.26 ± 1.90 responses on the active lever per 2-hour session (no rats were excluded from the study due to baseline responding), at which point rats were divided into two groups (n = 6/group): those that would be exposed to chronic intermittent nicotine vapor (nicotine-dependent group) and those that would be exposed to ambient air (non-dependent group). Rats were tested for nicotine self-administration in 2-hour sessions (as described above), 8–12 hours into WD from vapor on days 1, 3 and 5–15 of vapor exposure.
To induce dependence on nicotine, animals were housed in Plexiglass chambers in a vapor delivery system (La Jolla Alcohol Research, Inc., La Jolla, CA, USA) and exposed daily to intermittent (12 hours on/12 hours off) nicotine vapor (George et al. 2010). Nicotine vapor was produced by bubbling air at a flow rate of 5 l/minute (LPM) through a gas-washing bottle containing a solution of pure nicotine (free base, Sigma-Aldrich). The highly concentrated nicotine vapor was then passed through a drop-catch bottle and further diluted by the addition of 60 LPM of clean air in a 2000 ml Erlenmeyer vacuum flask at room temperature. The final nicotine–air mixture was homogeneously distributed between chambers at a flow rate of 15 LPM. Nicotine vapor concentrations were adjusted by varying the flow rate at which nicotine is bubbled. Air controls were treated in a similar manner except that air entering the cages did not contain nicotine.
Experiment 2
Rats were exposed to chronic intermittent nicotine vapor (as described above; n = 6) or ambient air (n = 5), and tested for behavioral signs of physical dependence on nicotine (Malin et al. 1992). At 8 hours WD from nicotine vapor on day 13 of vapor exposure, rats were injected with saline (3 ml/kg, s.c.) and observed (10 minutes) for spontaneous behavioral signs of nicotine dependence. On day 14 of vapor exposure, rats were injected with mecamylamine (1.5 mg/3 ml/kg, s.c.), a non-specific antagonist of nAchRs, and observed (10 minutes) for precipitated behavioral signs of nicotine dependence. During each test, each rat was observed by a treatment-blind experimenter for 10 minutes, during which time the number of blinks, gasps, writhes, head shakes, ptosis (drooping eyelids), teeth chattering and yawns was recorded. Multiple successive counts of any sign required a distinct pause between episodes. Total occurrences of all somatic signs were summed for a single overall WD score for each rat.
Measurement of air–nicotine concentrations
Measurement of air–nicotine levels was performed using the method developed by the National Institute for Occupational Safety and Health (NIOSH). Briefly, Nicotine air was sampled on sorbent tubes (XAD-2, 80/40 mg) at 1 LPM for 60 minutes. Samples were analyzed by the Hartford Laboratory using gas chromatography coupled with a nitrogen phosphorous detector (method 2544; NIOSH 1977a,b).
Statistical analysis
Data from FR-1 self-administration tests were analyzed using two-way repeated-measures analyses of variance (RM ANOVAs) where day was the within-subjects factor, and nicotine vapor history was the between-subjects factor. Data from somatic WD tests were analyzed using RM ANOVAs where mecamylamine dose (0 or 1.5 mg/kg) was the within-subjects factor, and nicotine vapor history was the between-subjects factor. Post hoc comparisons were conducted using the Student Newman–Keuls test. Statistical significance was set at P < 0.05.
Results
Self-administration on days 1, 3 and 5 of vapor exposure was analyzed relative to baseline to assess the effects of intermittent testing (i.e. deprivation effect) and also the early effects of vapor on nicotine self-administration. Two-way RM ANOVA indicated a marginally significant main effect of day on nicotine self-administration, F(3,30) = 2.80, P = 0.057, suggestive of a nicotine deprivation effect across all rats (Fig. 1a). There was no effect of vapor on nicotine self-administration at this early stage of testing (P > 0.05), nor was there a day × dependence interaction effect. Rats were then tested on consecutive days to assess the effect of chronic nicotine vapor on nicotine self-administration (Fig. 1a). A two-way RM ANOVA for data from vapor days 6–15 indicated a significant day × dependence interaction effect, F(10,100) = 2.12, P = 0.029, on nicotine responding. There was a tendency toward a main effect of nicotine vapor (P = 0.07) and no main effect of day (P > 0.05) on operant nicotine responding. Post hoc analyses revealed that nicotine-dependent rats responded more for nicotine than non-dependent controls on days 13, 14 and 15 (P < 0.02 in all cases). When active lever presses were expressed as a function of baseline (Fig. 1b), a two-way RM ANOVA for data from vapor days 6–15 indicated a significant day × dependence interaction effect, F(9,90) = 2.54, P = 0.012. Post hoc analyses revealed that, relative to controls, change in active lever presses from baseline was greater in nicotine-dependent rats on days 14 and 15 (P < 0.05 in both cases).
(a) Mean ± standard error of the mean nicotine infusions (FR-1 schedule, 20-second timeout) by nicotine vapor-exposed (black circles) and air-exposed (white circles) rats during 2-hour operant sessions prior to (baseline = 3-day mean) and during chronic intermittent (12 hours/day) nicotine vapor exposure. Rats were tested intermittently during the first 5 days of vapor exposure (72 hours between end of baseline and day 1 test, 48 hours between day 1 and 3 tests, and 48 hours between day 3 and 5 tests). Rats were then tested daily during days 6–15 of nicotine vapor exposure. All tests occurred at 8–12 hours following removal from nicotine vapor (i.e. withdrawal). Data suggest a tendency toward a nicotine deprivation effect on days 1–5 regardless of nicotine vapor condition, #P = 0.057 main effect of day. Also, nicotine vapor-exposed rats exhibited significant elevations in nicotine responding on days 13, 14 and 15 of vapor exposure, *P < 0.02 relative to non-dependent controls. Also shown are (b) active and (c) inactive levers by nicotine vapor-exposed (black circles) and air-exposed (white circles) rats over days, expressed as percent of baseline (3-day mean) prior to the start of vapor exposure. Change from baseline active lever responding by individual rats confirms raw data and statistical analyses. Nicotine vapor-exposed rats exhibited no change from baseline inactive lever responding, nor did non-dependent controls with the exception of a slight increase on days 6 and 7 of the protocol, confirming that the difference between groups in inactive lever responding was due to baseline differences after rats were split into groups matched for active lever responding. *P < 0.05 relative to non-dependent controls
Nicotine vapor-exposed rats responded more than controls on the inactive lever over days during both intermittent, F(1,10) = 12.38, P = 0.006, and consecutive days, F(1,10) = 11.23, P = 0.007, of testing (data not shown). However, relative to baseline, inactive lever responding did not change over time for either group (Fig. 1c), confirming that rats exposed to nicotine vapor had higher baseline inactive lever pressing than air-exposed controls, Mann–Whitney U = 2.50, n1 = n2 = 6, P = 0.009.
As shown in Fig. 2, a two-way RM ANOVA for data collected in experiment 2 indicated a significant dose × dependence interaction effect on somatic WD signs, F(1,9) = 7.55, P = 0.023. Main effects showed that nicotine-dependent rats exhibited higher somatic WD scores, F(1,9) = 6.49, P = 0.031, and somatic WD scores were higher after mecamylamine injections relative to saline injections, F(1,9) = 13.82, P = 0.005. Post hoc analyses revealed that nicotine-dependent rats injected with mecamylamine exhibited significantly higher somatic WD scores relative to mecamylamine-injected controls (P = 0.002) and also relative to their own scores following vehicle injections (P = 0.001).
Mean ± standard error of the mean somatic withdrawal scores by nicotine vapor-exposed (black bars) and air-exposed (white bars) rats injected with saline (vapor day 13) or mecamylamine (1.5 mg/kg, vapor day 14) at ∼8 hours of withdrawal from nicotine vapor. Scores from the 10-minute observation period represent a summation of counts for behavioral signs that include blinks, gasps, writhes, head shakes, ptosis, teeth chattering and yawns (Malin et al. 1992). Nicotine vapor-exposed rats exhibited higher scores overall than controls (P < 0.05), and nicotine vapor-exposed rats injected with mecamylamine exhibited robust withdrawal relative to both saline injections and nicotine-naïve rats. *P < 0.01 relative to non-dependent controls, #P < 0.01 relative to saline injection
Finally, as shown in Fig. 3, measurement of nicotine in the air demonstrated a robust positive correlation between nicotine flow rate and the concentration of nicotine in the air (r2 = 0.99, P < 0.001).
Air–nicotine concentrations expressed as parts per million (ppm; left y-axis) and mg/m3 (right y-axis). Air–nicotine concentrations are shown as a function of nicotine flow rate, expressed in liter per minute (LPM). There was a strong positive correlation between nicotine flow rate and air–nicotine concentrations (R2 = 0.99). Rats in experiments 1 and 2 were exposed to a nicotine flow rate of 5 LPM, which produces air–nicotine concentrations similar to those experienced by human smokers
Discussion
A major challenge in the field of pre-clinical nicotine addiction research has been the effort to produce escalated nicotine self-administration in rats. In this study, we achieve escalation of nicotine self-administration during daily limited-access (2-hour) operant sessions by exposing rats to chronic intermittent (12 hours/day) nicotine vapor and testing rats for nicotine self-administration during WD (8–12 hours) from vapor. Early in the vapor exposure procedure (days 1–5), rats were tested for nicotine self-administration on non-consecutive days (48–72 hours between sessions), and all rats exhibited a tendency toward a nicotine deprivation effect, regardless of nicotine vapor condition. Following 12 days of chronic intermittent nicotine vapor inhalation, rats exhibited increases in operant nicotine responding relative to non-dependent controls, and also relative to their own baseline both in terms of mean responding and percent change from baseline.
Baseline nicotine self-administration rates during short-access operant sessions vary greatly based on whether rats were previously trained to perform an operant response for food followed by substitution of an i.v. nicotine infusion for the food reinforcer. In our study, rats were never food deprived and were never trained to perform an operant response for food. The baseline infusion rates observed in our study were slightly lower (per hour) but comparable to infusion rates previously observed in male Wistar rats trained in this way (e.g. Cohen et al. 2012).
Extended access to many drugs of abuse (e.g. stimulants, opiates) produces escalation of self-administration in rats (Koob & Kreek 2007). However, simply allowing rats longer periods of access to operant nicotine, e.g. 6 hours (Paterson & Markou 2004) or 23 hours (O'Dell et al. 2007) per day, is not sufficient to produce escalation of nicotine self-administration. A recent study showed that long-access self-administration levels are increased over time by (1) intermittence of testing and (2) co-administration of a MAOI, although 1-hour intakes were not compared between long-access and short-access groups in that study (Cohen et al. 2012). In this study, we show that chronic intermittent nicotine vapor produces robust escalation of operant nicotine self-administration during short-access sessions, without co-administration of other drugs, and without the use of food substitution training. In the present study, all operant sessions occurred at 8–12 hours of WD from levels of nicotine vapor that did not produce any observable signs of spontaneous somatic WD, but did facilitate mecamylamine-precipitated somatic WD.
There does not appear to be a one-to-one relationship between somatic signs of nicotine WD and escalation of nicotine self-administration. Long-access operant nicotine sessions reliably produce mecamylamine-precipitated (but not spontaneous) somatic WD signs (Paterson & Markou 2004; O'Dell et al. 2007). Furthermore, severity of mecamylamine-precipitated nicotine WD is positively correlated with mean total nicotine intake (but not escalation per se; O'Dell et al. 2007). However, in Paterson & Markou's (2004) study, rats exhibited spontaneous and precipitated nicotine WD in the absence of escalation. In another study, long-access rats with both daily and intermittent access to nicotine exhibited spontaneous and precipitated WD, but only rats with intermittent access exhibited escalation of intake (Cohen et al. 2012). In our study, nicotine vapor-exposed rats exhibited mecamylamine-precipitated, but not spontaneous, WD (similar to George et al. 2010) and also exhibited escalation of short-access nicotine self-administration. The sum of these results plus the fact that rats are not injected with mecamylamine prior to self-administration sessions suggests that physical WD may contribute to, but is not solely responsible for, escalated nicotine self-administration, and that there is likely a neural dissociation between physical and motivational signs of nicotine dependence (Koob 2008).
In this study, rats were exposed to a constant air–nicotine concentration of ∼7.5 mg/m3 over a 12-hour period. The average dose of nicotine per puff by a heavy smoker is ∼75–200 μg (Xie et al. 2006), and the average daily nicotine intake in smokers is ∼42 mg/day (Djordjevic, Stellman & Zang 2000). Considering that the total volume of air entering the lungs per minute in a healthy adult is 5–8 LPM at rest, the average human smoker is exposed to a range of air–nicotine concentrations between 4 and 12 mg/m3, similar to air–nicotine concentrations observed in the present study. This averaged daily concentration in humans does not take into consideration the smoking pattern and associated spike in nicotine level after each puff and cigarette, although puff-associated spikes in brain–nicotine concentrations appear to be dampened in dependent smokers due to slower release from the lungs (Rose et al. 2010). These results suggest that chronic exposure to ∼7.5 mg/m3 nicotine in rats may mimic the human condition.
Rats with long access (21–23 hours/day) to operant nicotine exhibit a nicotine deprivation effect following 3 days without nicotine access (George et al. 2007; O'Dell & Koob 2007; Cohen et al. 2012). Conversely, prior studies report that rats with short access to operant nicotine in 1-hour sessions do not show any increase in nicotine self-administration, transient or otherwise, when operant sessions are spaced by 48 or 72 hours (George et al. 2007; Cohen et al. 2012). In the present study, we show that rats with short access to operant nicotine in 2-hour sessions exhibit increases in nicotine self-administration when operant sessions are spaced by 48 or 72 hours, but this increase in responding is transient and fades upon resumption of consecutive-days testing. It is not clear what produced the different patterns of results in this experiment versus previous studies, as both studies used the same rat strain and gender, the same unit dose and the same deprivation durations. Two possible causes of these differential effects are (1) the length of operant nicotine self-administration sessions (2 hours in this study versus 1 hour in previous studies) or (2) slightly lower baseline levels of responding (per hour) in this study versus previous studies.
In conclusion, the present investigation shows that chronic intermittent nicotine vapor inhalation produces escalation of nicotine self-administration, as well as physical dependence on rats. Furthermore, we report that rats exhibit a nicotine deprivation effect when 2-hour operant nicotine self-administration sessions are spaced by 48–72 hours. It will be important for future studies to explore both the negative affective components of nicotine WD as well as dose–response curves of i.v. nicotine self-administration in rats made dependent on nicotine via vapor inhalation. Overall, these data suggest that nicotine vapor can be used to induce escalation of nicotine self-administration, a long-standing obstacle in the field of nicotine addiction research.
Acknowledgements
The authors thank Dr. Ami Cohen and Ilham Polis for their assistance with nicotine vapor and nicotine self-administration methodologies. This work was funded by LSUHSC SOM Faculty Start-Up Funds (N.W.G.) as well as the following NIH/NIAAA grants: AA018400 (N.W.G.), LSUHSC T35 training grant (AA021097) and LSUHSC T32 training grant (AA007577).
Authors Contribution
AMW, BB, AYA, MTW conducted Experiments 1 and 2. OG conducted the air–nicotine concentration studies. NWG was responsible for the study concept and design, and also drafted the manuscript. AMW and OG contributed to the writing and editing of the manuscript. All authors approved the final version of the manuscript for publication.
