β-Nicotyrine and e-cigarette abuse liability II: Behavioral economic demand and intracranial self-stimulation in rats

Abstract

Background:

Given the prominence of electronic nicotine delivery systems (ENDS), understanding the product characteristics that underlie ENDS abuse liabilty is a public health research priority. β-Nicotyrine (β-Nic) is a uniquely prevalent constituent (up to 25 % of nicotine [Nic] levels) in ENDS aerosols that may contribute to ENDS abuse liability. Previously, we found clinically relevant concentration of β-Nic doubled the elimination half-life and prolong the interoceptive (discriminative stimulus) properties of Nic in rats, but that it had little or no psychoactive effects itself. The present study used i.v. self-administration (23 hrs/day) and intracranial self-stimulation (ICSS) paradigms to further evaluate if β-Nic alone has abuse liability or if it enhances the abuse liability of Nic.

Methods:

We examined if β-Nic (> 25 % of Nic levels) substitutes for Nic (0.03 mg/kg/inf) during drug self-administration, and whether combinations of Nic and β-Nic alter the acquisition and behavioral economic demand of i.v. Nic. Reinforcement-enhancing and aversive effects of β-Nic alone or in combination with Nic were also evaluated using ICSS.

Results:

Moderately high β-Nic doses (0.03 & 0.10 mg/kg/inf) increased demand for Nic (0.03 mg/kg/inf). β-Nic alone did not maintain self-administration, affect ICSS, or influence the effects of Nic on ICSS.

Conclusion:

β-Nic itself does not have abuse potential in either i.v. self-administration or ICSS models, but it does increase the reinforcing efficacy of self-administered Nic. As such, β-Nic may play a role in the abuse liability of ENDS, which has important treatment and regulatory implications.

1. Introduction

The Monitoring our Future Survey found that high school seniors report higher past 30-day use of electronic nicotine delivery systems (ENDS) compared to cigarettes (15 % vs. 2.5 %, respectively; Meich et al., 2022). This is in stark contrast to 2011, when cigarette use (15.8 %) was much more prevalent than ENDS use (1.5 %; Arrazola et al., 2015). Thus in just over a decade, ENDS have replaced cigarettes as the primary means through which tobacco use disorder (TUD) is established in adolescents. Consequently, understanding the product characteristics that underlie ENDS abuse liabilty has emerged as a public health research priority (Kaondera-Shava et al., 2024; National Academies of Science, 2018).

Nicotine (Nic) is the primary constituent responsible for establishing TUD (Benowitz, 2010; Benowitz et al., 2016). However, there are other non-Nic constituents that mimic or enhance the central nervous system (CNS) effects of Nic that could also contribute to ENDS dependence (Brennan et al., 2014; Hall et al., 2014; Hoffman and Evans, 2012; LeSage and Smethells, 2018). A prime example is β-nicotyrine (β-Nic), a minor alkaloid that can reach substantially higher concentrations in ENDS aerosol than in tobacco smoke (e.g., 50x than in smoke or 25 % of Nic levels in ENDS aerosols; Jacob et al., 1999; Martinez et al., 2015; Son et al., 2018). β-Nic is formed in ENDS through dehydrogenation during ENDS aerosolization (Clayton et al., 2009) and oxidation as ENDS liquids age (Martinez et al., 2015; Wada et al., 1959). Studies from our lab (Harris et al., 2019) and others (Martinez et al., 2015; Son et al., 2018a) have found that the aerosol produced by recently opened e-liquids can contain β-Nic concentrations up to ~13 % of Nic levels, with the levels varying from 1.1 % to 12.9 % depending on the heating temperature, ENDS delivery vehicle, Nic concentration, vaping topography, and other factors (see Son et al., 2018). Following oxidation, as ENDS liquids age, concentrations of β-nic increase to 8 % of Nic levels in non-aerosolized solutions and can reach 25 % of Nic levels in ENDS aerosols (Martinez et al., 2015).

β-Nic is a potent inhibitor of the hepatic enzyme cytochrome P450 (CYP450) subtypes that are responsible for Nic metabolism in mice and humans (CYP2A6 & 2A13; Denton et al., 2004; Kramlinger et al., 2012; Shigenaga et al., 1989), and can slow Nic metabolism in mice at high doses (70 mg/kg, i.p.; Stålhandske and Slanina, 1982). These findings prompted a nicotyrine hypothesis (Abramovitz et al., 2015) that β-Nic modulates ENDS abuse liability in humans by slowing Nic metabolism. Given that slower Nic metabolism is associated with greater smoking cessation success in adults see review by Jones et al. (2022), they suggest that β-Nic may facilitate ENDS use as a smoking cessation aid for humans. They also posit that higher levels of β-Nic in ENDS products should result in greater satisfaction than non-β-Nic containing ENDS, suggesting that β-Nic could contribute to ENDS abuse liability.

We have begun to evaluate the impact of β-Nic on the abuse liability of Nic in rats to evaluate this hypothesis. We recently showed that β-Nic slows Nic metabolism at a clinically relevant concentration (25 % of Nic), suggesting that β-Nic also inhibits the CYP2B1 enzyme that mediates Nic metabolism in rats (Smethells et al., 2025). This study also assessed whether β-Nic had interceptive (i.e., subjective) effects and if it altered the discriminability of Nic in a drug discrimination paradigm. Specifically, rats are pretreated with saline or Nic ± β-Nic and are reinforced with food for responding on the correct drug associated lever. β-Nic only weakly substituted for Nic in females at very high doses (2.5 mg/kg), suggesting it has little, if any, interoceptive effects of its own. This is consistent with our in vitro findings showing that β-Nic did not bind to several nicotinic acetylcholine receptors (nAChRs) including the α4β2 nAChR that plays a primary role in nicotine discrimination (Smith and Stolerman, 2009). However, when Nic was combined with β-Nic, Nic discrimination was enhanced at moderate doses under a 10-min pretreatment. A more robust effect was seen under a 60-min pretreatment, with β-nic producing a more robust enhancement of Nic’s discriminative stimulus effects. These findings demonstrate that β-Nic has minimal behavioral effects of its own but that it can enhance the potency and duration of Nic’s subjective and, possibly, other behavioral effects.

The present study aimed to investigate whether β-Nic enhances the reinforcing effects of nicotine (Nic) or possesses reinforcing properties of its own, using an intravenous (i.v.) self-administration paradigm. This paradigm is considered the gold-standard preclinical model for assessing abuse liability because it involves volitional drug taking as occurs in humans (Mello and Negus, 1996). In Experiment 1a, the ability of β-Nic to substitute for Nic was examined. Experiment 1b employed a behavioral economic approach to assess whether combinations of β-Nic and Nic are more reinforcing than Nic alone. Based on our prior findings (Smethells et al., 2025), we hypothesized that β-Nic would not itself be reinforcing but would increase the reinforcing efficacy of Nic by reducing demand elasticity compared to Nic alone. In a second experiment, we evaluated β-Nic effects in an intracranial self-stimulation (ICSS) model, which provides measures of a drug’s reinforcement-enhancing/hedonic effects at moderate doses and aversive/anhedonic effects at high doses (Fowler et al., 2011; Markou and Koob, 1992). β-Nic’s effects on ICSS were first compared with those of Nic and the minor tobacco alkaloid nornicotine (Nornic) (positive controls) (Exp. 2a). β-Nic’s ability to extend the duration of Nic’s effects was also assessed using 10- and 60-minute pretreatment intervals as in our prior nicotine discrimination study (Exp. 2b). Based on our prior findings (Smethells et al., 2025), we hypothesized that β-Nic would prolong the effects of Nic on ICSS without itself affecting ICSS.

2. Method

2.1. Animals

Male and female Sprague-Dawley (SD; Inotiv [previously Envigo], West Lafayette, IN) rats (275 g and 225 g at delivery, respectively; sample sizes are in experimental descriptions) were individually housed with free access to chow and water in a temperature- (22° C) and humidity-controlled room with a 12 hr reversed light cycle (lights off at 6 am). Rats were individually housed in either an operant chamber (Experiment 1) or in a plastic tub cage (Experiment 2) and provided free access to water and restricted access to 16 g or 18 g of food per day for female and male rats, respectively. Protocols were approved by the Hennepin Healthcare Research Institute’s Institutional Animal Care and Use Committee and were in accordance with NIH guidelines set forth in the Guide for the Care and Use of Laboratory Animals (National Reserach Council, 2011).

2.2. Apparatus

Drug self-administration operant chambers (Med-Associates, St. Albans, VT) were composed of aluminum and polycarbonate walls and a stainless-steel grid floor. The chamber had two response levers on the front panel, each with a white stimulus light located directly above and a food hopper between the levers (food pellets: BioServ #F0021). The back panel contained a house light mounted centrally at the top. Infusion pumps delivered drug i.v. (0.1 ml/kg) via tygon tubing connected to the rat via a swivel and spring tether. Chambers were contained in sound-attenuating boxes equipped with ventilation fans that provided masking noise. A computer (OS: Windows 10^®) running MED-PC V^® (Med Associates) orchestrated experimental sessions and recorded data.

2.3. Drugs

(−)-Nicotine base (Sigma Chemical Co., St. Louis, MO) was dissolved into saline to form a 30 mg/ml stock solution, with the concentration verified by gas chromatography with nitrogen phosphorous detection using our routine assay (LeSage et al., 2003). β-Nic (obtained as a neat oil from Cayman Chemical, Ann Arbor, MI) was diluted to 50 mg/ml in pure EtOH. Stock solutions varied by no more that ± 5 % (average < 1 %) from the target concentration. All drug doses, including Sal, contained 6 % v/v EtOH (w/v = 0.79 g/ml; a non-reinforcing/ICSS altering dose of 3.95 mg/kg Sinden and Le Magnen, 1982) to obtain target doses administered at a volume of 0.1 ml/kg. The addition of 6 % EtOH to Sal and all drug solutions, including drug combinations, was done to control for the highest EtOH concentration used (i.e., 6 % EtOH in 0.3 mg/ml Nic + 3.0 mg/ml β-Nic).

2.4. Experiment 1: drug self-administration, demand, extinction and reinstatement

2.4.1. Jugular and femoral catheterization surgery

Rats were anesthetized with isoflurane (0.4 – 0.8 L/min at 2.5 % oxygen concentration levels) and then implanted with a chronic indwelling catheter into the right jugular vein according to our standard procedures (LeSage et al., 2002, 2003). The catheters exited the body between the scapulae and attached to the vascular-access harness. Following four days of recovery, rats were given daily catheter flushes (0.1 ml; i.v.) of heparinized saline (30 units/ml) and ceftriaxone (5.25 mg/ml), and a s.c. injection of buprenorphine (0.05 mg/kg; first two days only) for analgesia. Infusions (0.1 ml; i.v.) of methohexital (50 mg/ml) were administered weekly to confirm catheter patency. If a catheter became occluded or lost patency, another catheter was implanted in the femoral vein or ipsilateral jugular vein.

2.4.2. Experiment 1a: β-Nic substitution following Nic self-administration training

Rats (M = 6, F = 7) were trained during 23-hr sessions (7-days/week) to respond on a lever for Nic (0.03 mg/kg/infusion) under a Fixed-Ratio (FR) 2 schedule. 23-hr self-administration sessions were used because this access duration more closely approximates clinically relevant Nic serum concentrations and daily patterns of use than shorter (e.g., 1 hr) sessions (Harris et al., 2008) and can have greater sensitivity to effects of non-nicotine tobacco constituents (Cohen et al., 2012). Nic self-administration stability was assessed after ≥ 10 sessions both visually (no apparent trend) and statistically by averaging the final three sessions and requiring ≥ 20 infusions and > 2:1 active to inactive lever responding (i.e., drug-associated vs general activity responding). Once stable, Nic was replaced with a dose of β-Nic (0, 0.075, 0.150, 0.300, 0.600 mg/kg), each for 7 sessions, with the dose order counter-balanced across subjects. The relative concentrations of Nic to β-Nic were chosen based on prior work (Smethells et al., 2025) showing they encompassed the relative concentrations that slow Nic metabolism (25 % of Nic) and alter Nic discrimination (25 % and higher). If fewer than an average of 10 infusions across the final 3 sessions were earned at a given β-Nic dose, then baseline Nic SA was re-established for 7 sessions using the Nic training dose (data not presented) before other β-Nic doses were examined.

2.4.3. Experiment 1b: acquisition of drug self-administration and exponential demand in rats responding for Nic + /− β-Nic

A separate group of rats (M = 17, F = 18) were randomly assigned to self-administer a mixture of Nic (0.03 mg/kg/inf) and β-Nic (0.0 [Nic alone], 0.0075, 0.03, 0.1 or 0.3 mg/kg/inf) under an FR 2 schedule in 23-hr sessions (7-days/week). Initially rats were assessed for acquisition of self-administration after 10 sessions using the same stability criteria from 2.4.2 (visual stable, 20 + infs, 2:1 active to inactive responses). All rats were assessed for a minimum of 10 sessions under FR 2 to assess acquisition differences between Nic alone and Nic+ β-Nic combinations. To test demand, a daily FR escalation (3, 4, 5, 6, 8, 10, 13, 16, 20, 25, 32, 40, 50, 63, 79, 100, 126, 158, 200, 241, 316, 398, etc.; ≈ 0.1 log unit scale separation) was conducted. Once 0 infusions were earned, rats were placed back on the FR 2 and assigned a different Nic + /− β-Nic combination dose for 7 sessions before the FR escalation protocol was repeated. This number of 23-hr sessions was sufficient to recapture baseline self-administration rates for all rats. To reduce variability in demand measures caused by ratio strain during their first FR escalation, the initial drug condition was repeated midway through the progression (i.e., the condition order was ABCADE) and only data from this replication (i.e., 2nd “A” condition) were used in the final demand analysis.

2.5. Experiment 2: intracranial self-stimulation

2.5.1. ICSS surgery and behavioral procedures

Surgery, apparatus, and training procedures used here are described in detail elsewhere (Harris et al., 2010, 2011, 2013; Swain et al., 2020). Briefly, animals were anesthetized using inhaled isoflurane (1–3 %) and implanted with an electrode in the medial forebrain bundle at the level of the lateral hypothalamus. Rats were later trained to respond for electrical brain stimulation using a modified version of the Kornetsky and Esposito(1979) discrete-trial current-threshold procedure (Markou and Koob, 1992), as we have described previously (Harris et al., 2010, 2011; Swain et al., 2020). Briefly, each trial was initiated with presentation of a non-contingent stimulation followed by a 7.5-s window, during which a positive response produced a second contingent stimulation identical to the first. Lack of responding during the 7.5-s window was considered a negative response. Each positive or negative response was followed by an inter-trial interval averaging 10 s, during which time additional responses delayed the onset of the subsequent trial by 12.5 s. Stimulus intensities were presented in four alternating descending and ascending series (step size, 5 μA), with five trials presented at each current intensity step. The current threshold for each series was defined as the midpoint between two consecutive intensity steps that yielded three or more positive responses and two consecutive intensity steps that yielded three or more negative responses. The overall ICSS threshold for the ≈ 45 min session was defined as the mean of the current thresholds from the four alternating series. To assess performance effects (e.g., motor disruption), response latencies (time between onset of the non-contingent stimulus and a positive response) were averaged across all trials in which a positive response was made

2.5.2. Experiment 2a: effects of nicotine, β-nicotyrine, and nornicotine alone on ICSS

Rats (M = 7, F = 6) were tested in daily ICSS sessions conducted Mon-Fri until thresholds were stable (less than 10 % coefficient of variation over a 5-day period and no apparent trend). To habituate animals to the injection procedure, s.c. saline was administered 10 min prior to ICSS testing twice per week (Tuesdays and Fridays) for at least 1 session and until ICSS thresholds were stable. Effects of Nic, β-nic, and Nornic were then tested. Nic was always tested first, while order of testing for β-nic and Nornic was counterbalanced across rats. A 2-week washout period and reestablishment of stable baseline ICSS thresholds occurred between each drug dose-response assessment. All drugs were administered s.c. 10 min prior to the session. To evaluate whether the 6 % EtOH present in saline and all Nic, β-nic, and Nornic doses (see Drugs) affected ICSS, effects of saline without EtOH were also tested as part of each drug dose-response determination. Nic was administered at doses of 0 (saline alone), 0 + EtOH (saline + 6 % ETOH), 0.125, 0.25, and 1.0 mg/kg. These Nic doses typically produce either reductions (0.125, 0.25 mg/kg) or elevations (1.0 mg/kg) in ICSS thresholds (Harris et al., 2015a, 2015b, 2018). β-nic and Nornic were administered at doses of 0 (saline alone), 0 + EtOH (saline + 6 % ETOH), 0.025, 0.1, 0.25, 1.0, 2.5, and 5.0 mg/kg. These Nornic doses bracket the range within which reductions or elevations in ICSS thresholds occur (Harris et al., 2015a, 2015b). Injections were typically administered on Tuesdays and Fridays, provided that ICSS thresholds were within baseline range on intervening days, and doses of each drug were administered in a counterbalanced order.

2.5.3. Experiment 2b: effects of nicotine + β-nicotyrine on ICSS

2.5.3.1. 10-min pretreatment interval.

Following completion of Experiment 2a, the same rats were tested under drug-free conditions for at least 2 weeks and until ICSS thresholds were stable. Effects of Nic alone and Nic combined with 0.25 mg/kg β-nic were subsequently tested using a 10-min pretreatment interval in a subset of rats (M = 3, F = 2). The order of formulations (Nic alone versus Nic + β-nic) was counterbalanced across rats, and rats were tested under drug-free conditions for at least two weeks and until ICSS thresholds were stable between each formulation. Nic alone doses were 0 (saline alone), 0 + EtOH (saline + 6 % EtOH), 0.03, 0.125, 0.25, or 1.0 mg/kg, administered s.c. 10 min prior to testing in a counterbalanced order every 2–3 days provided that ICSS thresholds were stable during intervening sessions. Nic + β-nic was administered in the same manner, except that the rats received cocktails of each of the above Nic doses combined with 0.25 mg/kg β-nic. We have used this same Nic dose range to evaluate effects of other isolated tobacco constituents and ENDS vehicles (e.g., propylene glycol) combined with Nic on ICSS (Harris et al., 2018, 2020, 2025). This β-nicotyrine dose was used because it is clinically relevant (i.e., 25 % of the highest nicotine dose (Martinez et al., 2015)) and because it enhanced nicotine’s discriminative stimulus effects under certain conditions in our previous study (Smethells et al., 2025).

2.5.3.2. 60-min pretreatment interval.

This experiment was tested in the same 5 rats that completed 10-min pretreatment testing, as well as the remaining 3 males and 4 females that had completed Experiment 2a (one male was lost to attrition prior to completing Experiment 2a, see Results). Procedures were the same as 10-min pretreatment testing, except that a 60 min pretreatment interval was used. This longer pretreatment interval was used to allow effects of β-nic on Nic metabolism to more clearly manifest (see Smethells et al., 2025). Effects of both 0.25 and 1.0 mg/kg β-nic on the Nic dose-response function were tested. The 1.0 mg/kg β-nic dose was included because it produced greater effects on Nic discrimination than 0.25 mg/kg β-nic in our prior study (Smethells et al., 2025) at a 60-min pretreatment interval, but not at a 10-min pretreatment interval. Order of formulations (i.e., Nic alone, Nic + 0.25 mg/kg β-nic, and Nic + 1.0 mg/kg β-nic) was counterbalanced across rats.

2.6. Data analysis

2.6.1. Experiment 1a: substitution of β-Nic for Nic iv self-administration

To examine if β-Nic substituted for Nic, the number of infusions earned were averaged across the final 3 sessions at each drug dose and then analyzed using a two-factor ANOVA (sex × β-Nic dose). Main effects were followed up with a one-factor ANOVA followed by Dunnet multiple comparisons tests to determine which doses differed from saline. Degrees of freedom for this and all other ANOVAs were adjusted using the method of Geisser and Greenhouse to correct for any violations of sphericity.

2.6.2. Experiment 1b: acquisition, exponential demand, extinction, and reinstatement

2.6.2.1. Acquisition.

The main dependent measure of acquisition was the number of sessions it took to reach the acquisition criteria, which was: 20 + infusions and 2:1 active to inactive lever response ratio averaged across the last three sessions prior to the first FR escalation (i.e., condition “A”). Differences in the number of sessions to meet acquisition criteria was compared across drug formulations using a one-way ANOVA.

2.6.2.2. Exponential demand quantification of the FR escalation.

To complete the demand function for individual demand fits (i.e., Table 1), the final unit price, where consumption was 0, was replaced with 0.01 since 0 is undefined on a log scale and the log of 0.01 (i.e., log 0.01 = −2) is the lowest log-unit value below the log of 1 infusion (i.e., log 0.03 = −1.52; see methods in Smethells et al., 2021). Additionally, to make average grouped fits of demand functions representative, 0 infusions (i.e., 0.01) were interpolated for each subject from the point where 0 infusions were earned to the highest unit price achieved within the group.

Table 1.

Individual rat exponential demand parameter fits for Nic + β-Nic dose combinations.

Female
β-Nic dose		0.0 mg/kg/inf			0.0075 mg/kg/inf			0.03 mg/kg/inf			0.10 mg/kg/inf			0.30 mg/kg/inf
ID#	k	Q0	α	R2	Q0	α	R2	Q0	α	R2	Q0	α	R2	Q0	α	R2
105	2.29	1.40	0.011	0.97	2.14	0.021	0.98	1.51	0.007	0.94	1.31	0.008	0.97	1.37	0.022	0.96
106	1.71	1.23	0.021	0.97	1.26	0.018	0.98	1.41	0.018	0.94	1.10	0.029	0.97	1.10	0.023	0.96
116	2.50	0.98	0.019	0.75	1.42	0.012	0.99	1.86	0.012	0.90	1.31	0.010	0.91	1.99	0.021	0.97
117	2.36	2.21	0.035	0.96	1.50	0.026	0.94	1.23	0.011	0.92	0.83	0.026	0.88	1.59	0.022	0.96
155	2.43	2.33	0.004	0.97	1.01	0.010	0.93	3.05	0.006	0.97	2.21	0.006	0.98	2.03	0.012	0.94
159	2.85	5.13	0.016	0.91	2.22	0.011	0.97	1.14	0.018	0.90	0.70	0.022	0.86	1.36	0.022	0.92
163	2.34	2.20	0.017	0.98	1.78	0.009	0.92	1.53	0.004	0.89	1.29	0.007	0.97	1.98	0.009	0.97
222	2.33	1.14	0.011	0.96	1.51	0.007	0.94	1.41	0.007	0.98	2.04	0.016	0.94	1.77	0.011	0.98
226	2.14	1.27	0.040	0.88	1.78	0.026	0.92	2.18	0.027	0.96	2.15	0.028	0.86	1.62	0.057	0.82
232	2.22	0.45	0.020	0.76	1.48	0.025	0.94	1.16	0.017	0.91	2.22	0.024	0.95	1.54	0.041	0.93
255	2.53	1.19	0.035	0.84	1.17	0.026	0.94	1.55	0.018	0.96	1.56	0.012	0.78	2.18	0.016	0.96
313	2.48	2.64	0.006	0.98	1.91	0.004	0.97	2.65	0.006	0.96	1.47	0.003	0.85	2.18	0.009	0.91
314	2.62	1.44	0.023	0.93	3.71	0.014	0.96	1.45	0.019	0.90	1.55	0.018	0.93	2.58	0.023	0.96
317	2.67	0.92	0.022	0.93	0.80	0.024	0.90	1.02	0.022	0.96	1.21	0.019	0.95	2.50	0.019	0.98
318	2.76	1.68	0.006	0.95	1.84	0.006	0.95	1.95	0.005	0.95	2.86	0.006	0.97	2.20	0.004	0.94
2079	2.52	1.00	0.035	0.92	2.58	0.014	0.97	1.39	0.020	0.97	1.32	0.026	0.96	3.10	0.023	0.97
7971	1.71	0.99	0.050	0.80	1.21	0.024	0.79	1.18	0.017	0.86	1.16	0.021	0.80	0.84	0.066	0.76
7972	2.91	0.93	0.005	0.93	1.67	0.004	0.97	1.69	0.005	0.94	1.60	0.006	0.96	1.52	0.005	0.95
Male
55	2.83	2.06	0.004	0.95	3.22	0.003	0.96	2.33	0.003	0.99	2.57	0.003	0.95	2.49	0.003	0.94
103	2.52	2.44	0.009	0.96	2.23	0.010	0.96	2.10	0.006	0.94	1.53	0.004	0.98	2.06	0.007	0.98
104	2.47	2.42	0.012	0.98	1.41	0.007	0.95	2.00	0.002	0.94	1.52	0.004	0.96	2.07	0.004	0.98
151	2.13	1.41	0.024	0.90	1.91	0.019	0.87	1.55	0.027	0.93	2.12	0.014	0.86	1.41	0.021	0.93
154	2.44	1.90	0.008	0.96	2.51	0.010	0.91	1.96	0.009	0.97	1.28	0.007	0.97	2.31	0.012	0.91
164	2.44	2.64	0.045	0.95	2.36	0.025	0.93	3.62	0.020	0.96	3.06	0.028	0.97	2.04	0.019	0.95
166	2.54	2.53	0.010	0.99	1.61	0.008	0.96	2.03	0.009	0.97	2.08	0.010	0.98	2.39	0.012	0.98
167	2.46	1.29	0.011	0.94	1.66	0.019	0.95	1.09	0.012	0.91	2.04	0.016	0.97	1.66	0.019	0.96
2881	2.67	0.98	0.008	0.90	1.74	0.011	0.99	1.27	0.009	0.97	1.45	0.009	0.95	0.94	0.012	0.92
2882	2.48	2.37	0.042	0.96	2.80	0.050	0.95	1.49	0.026	0.94	1.39	0.017	0.86	2.65	0.040	0.86
7973	2.64	1.67	0.007	0.96	1.59	0.008	0.92	2.00	0.007	0.98	1.36	0.007	0.95	1.73	0.007	0.95
7975	1.58	0.69	0.069	0.75	1.00	0.037	0.83	1.32	0.034	0.82	0.86	0.043	0.81	0.43	0.120	0.87
7984	2.48	1.44	0.020	0.96	2.28	0.015	0.97	2.44	0.042	0.95	2.34	0.037	0.90	2.30	0.020	0.97
7985	2.10	1.60	0.019	0.96	1.75	0.021	0.95	1.88	0.012	0.95	1.21	0.008	0.89	1.39	0.014	0.96
7987	2.61	4.08	0.036	0.91	0.55	0.059	0.90	1.83	0.021	0.88	2.39	0.021	0.94	2.13	0.030	0.90
7988	2.14	1.52	0.031	0.96	1.20	0.043	0.91	1.76	0.034	0.95	1.55	0.045	0.94	2.91	0.031	0.96
8000	2.21	1.57	0.079	0.94	1.47	0.055	0.95	1.60	0.015	0.97	1.30	0.020	0.91	1.54	0.037	0.93

X.XX (bolded and underlined) values denote significant (p < 0.0125) within-subject increase in abuse liability (lower α & higher Q₀) compared to Nic alone, whereas X.XX (bolded and italicized) denote a significant decrease in abuse liability metrics.

Initially, separate three-factor ANOVAs (Sex × β-Nic dose × FR value) was conducted to determine if lever responding (active vs. inactive) and Nic consumption differed across FR values according to β-Nic dose and/or sex. If no effect of sex was observed, then analyses were conducted with sex collapsed. To quantify demand elasticity and intensity across the Nic ± β-Nic dose-response, Nic group mean consumption in mg/kg was fit across FR values using the Hursh and Silberberg (2008) demand equation:

$$\mathrm{l}\mathrm{o}\mathrm{g}Q=\mathrm{l}\mathrm{o}\mathrm{g}{Q}_0+k\left(e^{-\alpha Q_{0}C}-1\right)$$ (1)

In this equation, $\mathrm{Q}$ is the quantity of a commodity consumed (mg/kg of Nic) and C is unit-price cost of the commodity (FR/infusion). ${\mathrm{Q}}_0$ and $\alpha$ are free parameters derived from the best-fit function and refer to maximal consumption at zero price (i.e., demand intensity) and rate of change in consumption across price (i.e., demand elasticity), respectively. Commodities that have larger α values have more elastic demand (i.e., rapid decrease in consumption) and less essential value (i.e., reinforcing efficacy), whereas those with smaller α values have more inelastic demand (i.e., slower decrease in consumption) and greater essential value. The scaling parameter, $k$, is a constant equal to the log unit range of mean consumption (group curve fits; Fig. 3) or individual rat consumption across β-Nic doses (individual rat curve fits; Table 1). Normalization of data using $k$ allows for comparisons of free parameter estimates for grouped demand functions (i.e., α and ${\mathrm{Q}}_0$) across the β-Nic dose-response. Group differences in abuse liability metrics (α and ${\mathrm{Q}}_0$) between Nic alone and each Nic + β-Nic combination dose were assessed using an Extra Sum-of-Squares F-Test using the 99 % confidence intervals (CIs) to determine if the group level fits significantly differed across β-Nic dose. To assess individual demand functions across β-Nic doses, different k values for individual rats provided the best curve fits and more accurate parameter estimates to describe individual demand. Thus, to compare reinforcing efficacy (α) across subjects with differing $k$ values, essential value (EV) was calculated using the individual fit parameters in the following equation from Hursh and Roma (2016):

$$EV=1/\left(100\mathrm{*}{\alpha }^{\mathrm{*}}{k}^{1.5}\right)$$

Fig. 3.

Nicotine consumption across β-Nic doses (mg/kg/inf; i.v.) is plotted as a function of unit price (FR/unit dose) in Female (n = 18) and Male (n = 17) rats (FR values increased daily: 3, 4, 5, 6, 8, 10, 13, 16, 20, 25, 32, 40, 50, 63, 79, 100, 126, 158, 200, 241, 316, 398, etc.; ≈ 0.1 log unit scale separation). The 99 % confidence interval (CI) of the Extra Sum-of-Squares F-test fit for the group-level α parameter (demand elasticity; lower values indicate more inelastic demand) from the exponential demand equation. Note that data for nicotine alone are the same in each panel. * indicates a significant difference (p < 0.05) in demand elasticity.

Essential Value (EV), P_max (price of maximum responding) and O_max (consumption at P_max) values were calculated (see Hursh, 2014 for equations) for each subject using the individually fit $k$, ${\mathrm{Q}}_0$ and α values. The mean of individually fit EV, P_max and O_max values were the primary measures to assess within-subject differences between demand functions using one-factor ANOVAs (log transforms were used to normalize data as needed). Data analysis was conducted in Prism v.10.4 and SPSS v. 21. For all analyses, a Bonferroni correction was applied to the p value for significance to adjust for the number of comparisons in each family of comparisons.

2.6.3. Experiment 2a: ICSS data analysis

Intracranial self-stimulation thresholds (in μA) and response latencies (in seconds) were expressed as a percentage of baseline (i.e., mean during last 5 sessions prior to each dose-response determination). In alone and combination testing, paired-samples t-tests comparing the saline alone and saline + 6 % EtOH conditions showed no significant differences. The mean of these conditions was therefore used as a single vehicle control condition for all analyses. In Experiment 2a, ICSS threshold and response latency data during Nic, β-nic, and Nornic dose-response testing were analyzed using separate two-factor ANOVAs with sex as a between-subject factor and dose as a within-subject factor. Because there were no effects of sex or sex × dose interactions for any drug, data at each dose were collapsed across sex and compared to vehicle using Dunnett’s post hoc tests.

2.6.4. Experiment 2b: ICSS data analysis

During combination testing, ICSS threshold and response latency data following Nic + β-nic and Nic alone at each pretreatment interval were analyzed using separate two-factor ANOVAs with formulation and Nic dose as within-subject factors. Sex was not included as a factor in these ANOVAs because only 2 females completed the protocol. After being collapsed across sex and formulation, data at each Nic dose were compared to 0 mg/kg Nic using Dunnett’s post hoc test. Due to attrition, Nic + 0.25 mg/kg β-nic data at the 60-min pretreatment interval were not available for one rat (see below). These data and data for the 1.0 mg/kg β-nic condition were therefore compared to Nic alone in separate ANOVAs rather than within a single analysis.

3. Results

3.1. Experiment 1a: β-Nic substitution for nicotine self-administration

Fig. 1 shows the infusions earned under a FR 2 schedule for Sal, Nic and across a range of β-Nic doses. Nic maintained responding that was significantly higher than Sal (t₁₀ = 9.12, p < 0.0001), but no dose of β-Nic did so. No differences between sexes were found.

Fig. 1.

Infusions earned (+/− SEM) under a FR 2 when Sal and a range of β-Nic doses were substituted for Nic (0.03 mg/kg/inf; i.v.) in female (n = 7) and male (n = 6) rats trained to self-administer nicotine in 23-hr sessions. The dashed lines indicate the 95 % CI of Sal.

3.2. β-Nic on the acquisition of nicotine self-administration

The mean number of sessions to meeting acquisition criteria were not significantly different across β-nic dose groups (12.31 ± 8.27, 15.89 ± 8.1, 13.63 ± 6.37, 11.38 ± 3.89, 13.38 ± 2.93, for the 0.0, 0.0075, 0.03, 0.1 & 0.3 mg/kg β-Nic doses, respectively, not shown). There was also no effect of β-Nic dose nor sex in the number of drug infusions earned across the first 10 sessions of acquisition. There was only a main effect of training session on number of drug infusions (F _3.127,131.3 = 5.378, p < 0.01), reflecting the increase in responding as rats learned to self-administer drug.

3.3. Demand assessment

3.3.1. Assessment of active and inactive responding

There was no significant main effect of sex or interaction with sex and thus, analyses were conducted with sex collapsed. Fig. 2 shows that active lever responding was significantly higher than inactive lever responding at all but the highest unit prices at each β-Nic dose (0.0: F_1,34 = 37.91, p < 0.0001, 0.0075: F_1,34 = 41.78, p < 0.0001, 0.03: F_1,34 = 26.34, p < 0.0001, 0.10: F_1,34 = 41.47, p < 0.0001; 0.30: F_1,34 = 31.16, p < 0.0001), and that active lever responding was an inverted U-shape function of unit price. There was a significant main effect of FR value (0.03 mg/kg β-Nic: F_22,748 = 22.57, p < 0.0001; 0.10 mg/kg β-Nic: F_22,748 = 38.17, p < 0.0001), β-Nic (0.03 mg/kg β-Nic: F_1,34 = 4.83, p < 0.05; 0.10 mg/kg β-Nic: F_1,34 = 6.94, p < 0.05) and an interaction of these factors (0.03 mg/kg β-Nic: F_22,748 = 1.75, p < 0.05; 0.10 mg/kg β-Nic: F_22,748 = 1.41, p < 0.05) on active lever responding, with significantly higher responding under Nic + β-Nic at moderate FR values (FR 20 – 60; Fig. 2). Active lever responding for mixtures of Nic + 0.03 and 0.10 mg/kg of β-Nic were significantly higher than those produced by Nic alone, especially at moderate FR values in Fig. 2B & C, respectively. However, responding under the Nic + 0.0075 and 0.30 β-Nic doses were not significantly different than Nic alone.

Fig. 2.

Active and inactive (inverted triangles) lever responses for i.v. infusions of 0.03 mg/kg/inf Nic alone (open symbols) and 0.03 mg/kg/inf Nic + x.xxxx mg/kg/inf β-Nic (closed symbols) are plotted as a function of unit price (FR/unit dose) with sexes combined (N = 35). Note that data for Nic alone are the same in each panel and B-Nic doses increase across panels (A: 0.0075, B: 0.0300, C: 0.1000 & D: 0.3000 mg/kg/inf) * Different from 0.03 mg/kg/inf nicotine alone at that FR, p < 0.05.

3.3.2. Elasticity and intensity of demand

A three (Sex × β-Nic dose × FR) factor ANOVA showed no main effects of sex on infusions earned at any β-Nic dose, however, there was a significant FR × Sex interaction at some unit doses (e.g., 0.03 mg/kg β-Nic dose [F_21,714 = 1.99, p < 0.01]). Sex differences in demand elasticity are therefore presented for each dose. Fig. 3 shows changes in Nic consumption across FR values and β-Nic doses with exponential demand curve fits for each sex. The mean associated demand parameters are presented in Fig. 4 (individual rat data are presented in Table 1). The main finding was that when moderate β-Nic doses (0.03 & 0.10 mg/kg) were added to Nic, it caused demand to become more inelastic and facilitated greater Nic consumption at higher unit prices in both sexes, as indicated by non-overlapping 99 % CIs for the group α parameter at the moderate β-Nic doses (See Fig. 3 for CI ranges). In addition, the mean α (F_4,136 = 7.92, p < 0.0001), P_max (F_4,136 = 5.48, p < 0.001) and O_max (F_4,136 = 7.94, p < 0.0001) of individual rats were significantly higher at moderate β-Nic doses, with the Nic + 0.03 and 0.10 mg/kg β-Nic doses higher than Nic alone condition (See Fig. 4 and bolded values in Table 1). There were no differences in demand intensity (${\mathrm{Q}}_0$) across β-Nic doses, indicating Nic consumption did not differ at any β-Nic dose at a theoretical unit price of 0.

Fig. 4.

Mean (± SEM) essential value, P_max and O_max values from individual rats derived from the exponential demand equation are presented as a function of β-Nic dose (0.0, 0.03, 0.10, & 0.30 mg/kg/inf). * denotes a significant difference (p < 0.05) from Nic alone (0.0 mg/kg/inf β-Nic).

3.4. Effects of nicotine alone, β-nicotyrine, and nornicotine on ICSS

3.4.1. Attrition and baseline measures

One male did not complete nornicotine dose-response testing in Experiment 2a due to loss of ICSS headcap (see Table 2, upper panel, for number of rats of each sex completing each drug dose-response determination). Mean baseline ICSS thresholds and ICSS response latencies did not differ between males and females for any drug dose-response determination (Table 2).

Table 2.

Mean ± SEM baseline ICSS thresholds (in μA) and response latencies (in sec) in males and females during each dose-response determination in Experiment 2a and 2b (top panel and bottom panel, respectively).

Experiment 2 a	Male			Female
	N	Threshold (μA)	Latency (sec)	N	Threshold (μA)	Latency (sec)
Nic	7	87.4 ± 17.4	2.7 ± 0.2	6	101.5 ± 13.9	2.7 ± 0.2
β-Nic	7	97.6 ± 13.8	2.7 ± 0.2	6	103.0 ± 15.4	2.6 ± 0.2
Nornic	6	89.3 ± 9.1	2.6 ± 0.2	6	106.6 ± 14.2	2.7 ± 0.2
Experiment 2b			Male	Female
	Pretreatment Interval (min)	N	Threshold (μA)	Latency (sec)
Nic + 0.00 mg/kg β-Nic	10	5	104.2 ± 14.5	2.8 ± 0.2
Nic + 0.25 mg/kg β-Nic	10	5	102.8 ± 11.8	2.8 ± 0.2
Nic + 0.00 mg/kg β-Nic	60	7	111.3 ± 8.1	2.9 ± 0.1
Nic + 0.25 mg/kg β-Nic	60	6	102.3 ± 8.8	2.7 ± 0.2*
Nic + 1.00 mg/kg β-Nic	60	7	121.2 ± 14.4	2.9 ± 0.2

Different from Nic + 0 mg/kg β-Nic

^*p < 0.05.

3.4.2. ICSS thresholds and response latencies

3.4.2.1. Nicotine.

There was a significant effect of Nic dose on ICSS thresholds (F_{1.3, 14.5} = 23.5, p < 0.0001), but no effect of sex or interaction between Nic dose and sex. Analysis of data collapsed across sex indicated that ICSS thresholds were reduced compared to vehicle at 0.125 and 0.25 mg/kg Nic (q = 3.6 and 3.4, p < 0.01 and 0.05, respectively) and elevated at 1.0 mg/kg Nic (q = 4.2, p < 0.01) (Fig. 5A). There was a significant effect of Nic dose on ICSS latencies (F_{1.6, 17.8} = 8.1, p < 0.01), but no effect of sex or interaction between Nic dose and sex. Comparison of data collapsed across sex indicated that ICSS latencies were reduced compared to vehicle at 0.125 and 0.25 mg/kg Nic (q = 3.1 and 3.0, respectively, p < 0.05) but did not differ from vehicle at 1.0 mg/kg Nic (Fig. 5B).

Fig. 5.

ICSS thresholds (% baseline, mean ± SEM) and ICSS response latencies (% baseline, mean ± SEM) following s.c. injection of Nic (A and B), β-Nic (C and D), and Nornic (E and F). Data are collapsed across sex for all drugs. Different from 0 mg/kg at that dose, *^,p < 0.05, ** p < 0.01.

3.4.2.2. β-nicotyrine.

There was no significant effect of β-Nic, sex, or interaction on ICSS thresholds (Fig. 5C) or response latencies (Fig. 5D).

3.4.2.3. Nornicotine.

There was a significant effect of Nornic dose on ICSS thresholds (F_{1.9, 18.9} = 7.5, p < 0.01), but no effect of sex or interaction between Nornic dose and sex. Analysis of data collapsed across sex indicated that ICSS thresholds were reduced compared to saline at 1.0 mg/kg Nornic (q = 3.7, p < 0.05) and elevated at 5.0 mg/kg Nornic (q = 3.0, p < 0.05) (Fig. 5E). There were no significant effects of Nornic dose, sex, or interaction on response latencies (Fig. 5F).

3.5. Experiment 2b: effects of nicotine + β-nicotyrine on ICSS

3.5.1. Attrition

Two males and three females were lost to attrition prior to completing dose-response testing for Nic alone and at least one of the two Nic + β-nic conditions (i.e., Nic + 0.25 mg/kg β-nic or Nic + 1.0 mg/kg β-nic) at the 60 min pretreatment interval (see bottom panel of Table 2 for the number of rats completing each dose-response determination during combination testing at the 10- and 60-minute pretreatment intervals).

3.5.2. 10-min pretreatment interval

Neither baseline ICSS thresholds or response latencies differed between formulations (i.e., Nic alone versus Nic + 0.25 mg/kg β-Nic; Table 2). Analysis of ICSS threshold data for Nic + 0.25 mg/kg β-nic and Nic alone at the 10-min pretreatment interval indicated a significant effect of Nic dose (F_{2.4, 37.5} = 51.0, p < 0.0001), but no effect of formulation (i.e., Nic alone versus Nic + 0.25 mg/kg β-Nic), or interaction between these variables. Analysis of data collapsed across formulation indicated that ICSS thresholds were reduced compared to 0 mg/kg at 0.125 and 0.25 mg/kg nicotine (q = 5.7 or 7.0, p < 0.05 or 0.01, respectively). The difference between 0 mg/kg and 1.0 mg/kg nicotine did not achieve statistical significance (p = 0.08) (Fig. 6A). There was no significant effect of Nic dose, formulation, or interaction between these variables on response latencies (Fig. 6B).

Fig. 6.

ICSS thresholds (% baseline, mean ± SEM) (A) and ICSS response latencies (% baseline, mean ± SEM) (B) following s.c. injection of Nic (0–1.0 mg/kg) + β-nic (0 or 0.25 mg/kg) at the 10-min pretreatment interval. ICSS threshold and response latency data following injection of Nic (0–1.0 mg/kg) + β-nic (0, 0.25, or 1.0 mg/kg) at the 60-min pretreatment interval are shown in Fig (C)-(F). All data are collapsed across sex. Significantly different from 0 mg/kg nicotine (collapsed across formulations), *p < 0.05, **p < 0.01.

3.5.3. 60 min pretreatment interval

Baseline ICSS thresholds did not differ between β-Nic dose formulations at the 60-min pretreatment interval. Baseline ICSS latencies were lower for Nic + 0.25 mg/kg β-nic compared to nicotine alone at the 60-min pretreatment interval (t = 8.7, p < 0.05), but did not differ between Nic + 1.0 mg/kg β-nic and Nic alone (Table 2).

Analysis of data for Nic + 0.25 mg/kg β-nic and Nic alone at the 60 min interval indicated no significant effects of Nic dose, formulation, or interaction between these variables for ICSS thresholds (Fig. 6C) or ICSS latencies (Fig. 6D). For Nic + 1.0 mg/kg β-nic and Nic alone, there was a significant effect of Nic dose (F_{1.5, 9} = 8.7, p < 0.05) on ICSS thresholds, but no effect of formulation or interaction. Analysis of data collapsed across formulations indicated that ICSS thresholds were reduced compared to 0 mg/kg at 0.125 mg/kg nicotine (q = 4.5, p < 0.05) but not at other nicotine doses (Fig. 6E). There was no significant effect of Nic dose, formulation, or interaction on ICSS response latencies at the 1.0 mg/kg β-nic dose (Fig. 6F).

4. Discussion

The present results show that β-Nic alone was not behaviorally active, as indicated by its failure to substitute for Nic during self-administration (Fig. 1) and lack of effects on ICSS thresholds (Fig. 5). However, β-Nic did alter the reinforcing effects of Nic when they were co-administered, as indicated by increased response output (Fig. 2B/C) and more inelastic demand for Nic (lowered α/higher EV, P_max and O_max; Table 1, Figs. 3 and 4). In contrast, β-Nic did not alter the effects of Nic on ICSS (Fig. 6). These findings are generally consistent with our recent work (Smethells et al., 2025) that found clinically relevant doses of β-Nic (25 % of Nic levels) alone did not have interoceptive (i.e. subjective) discriminative stimulus effects but slowed Nic metabolism and prolonged the discriminative stimulus effects of Nic. The primary difference is that the β-Nic doses that altered the reinforcing efficacy of Nic were higher (100 % & 300 % of Nic levels) than those measured in ENDS aerosols, which may limit the generality of present finding. Broadly, however, these novel findings demonstrate behavioral mechanisms by which β-Nic could influence the abuse liability of ENDS, which have important clinical and tobacco regulatory implications.

4.1. Effects of β-Nic on Nic demand elasticity

Moderate β-Nic doses caused Nic demand to be more inelastic (vs. Nic alone; i.e., lower α values, higher EV, P_max and O_max values), both when it was assessed on group means (Fig. 3) and within-subject on the individual demand fits (Table 1; Fig. 4), indicating that β-Nic increases the reinforcing efficacy of Nic. Demand intensity (${\mathrm{Q}}_0$) did not consistently differ from Nic alone on group fits, indicating that β-Nic does not theoretically increase consumption of Nic when unit price was zero (i.e., Nic was “free”). These findings suggest that β-Nic could contribute to the abuse liability of ENDS by increasing the reinforcing efficacy of Nic, making ENDS users less sensitive to increases in price. However, at the highest β-Nic dose (0.30 mg/kg), Nic consumption did not differ from the Nic alone baseline, indicating a bi-tonic dose-response relationship for β-Nic effects on Nic self-administration. That is, by slowing Nic metabolism and thereby increasing blood/brain Nic concentrations, high doses of β-Nic may begin to increase satiation and/or aversive effects of Nic. Because only a single Nic dose was used in the present self-administration study and β-Nic did not decrease Nic self-administration, further studies are needed to address this issue by examining β-Nic effects across a range of Nic doses to determine whether β-Nic produces a leftward shift in the Nic dose-response curve, as would be expected if β-Nic increases satiation/aversion at higher Nic doses.

4.2. Mechanism β-Nic’s effects

The lack of reinforcing effects for β-Nic during Nic self-administration (Fig. 1) and the absence of reinforcement-enhancing and aversive effects on ICSS (Figs. 5 and 6A; 10 min pretreatment conditions) are consistent with our recent work. Specifically, we found that β-Nic alone generally does not produce subjective effects during drug discrimination when substituted for Nic and that nicotine-naïve rats did not discriminate β-Nic alone from saline (Smethells et al., 2025). Furthermore, β-Nic alone did not bind to several nAChRs (α2β2, α3β4 & α4β2) in vitro in our prior work, which includes the α4β2 nAChR subtype that mediates nicotine’s discriminative stimulus, reinforcing, reinforcement-enhancing, and aversive effects (Papke et al., 2007). These findings suggest that the primary mechanism through which β-Nic enhances the reinforcing properties of Nic is through its ability to slow Nic metabolism rather than by a pharmacodynamic mechanism.

The metabolic effect of β-Nic on Nic abuse liability is relevant to prior research examining the behavioral effects of Nic during methoxsalen administration, another compound that slows Nic metabolism via CYP2A5/6 inhibition (Benowitz et al., 2004; Damaj et al., 2007). In humans with tobacco use disorder, methoxsalen pretreatment dose-dependently increased Nic plasma concentrations produced by oral Nic (i.e., > AUC) and increased the latency to smoking the first cigarette during a free-smoking period, but otherwise it did not alter metrics of Nic abuse liability (i.e., cigarettes smoked, total puffs, etc.; Sellers et al., 2000). In mice, methoxsalen dose-dependently slowed Nic metabolism (up to fourfold), increased Nic conditioned place preference, and increased severity of nicotine withdrawal (Bagdas et al., 2014). In rats, methoxsalen has much lower affinity at CYP2B1 (Damaj et al., 2007), the primary enzyme in the metabolic pathway of Nic in this species (Koenigs and Trager, 1998). However, it does slow cotinine formation (Ding et al., 2024), which is an allosteric modulator at nAChRs that enhances Nic’s reinforcing effects. Behaviorally, methoxsalen pretreatment reduced the acquisition but not maintenance of nicotine self-administration under a FR-1 schedule. Effects of methoxsalen on self-administration at higher FRs was not studied. Although β-Nic did not affect acquisition in the current study, its lack of effect on consumption at low FRs (e.g., ${\mathrm{Q}}_0$ was similar across β-Nic doses; Table 1) is consistent with the lack of effects of methoxsalen on maintenance in Ding et al. (2024). Another compound that slows Nic metabolism and alters the abuse liability of Nic is menthol. Work by Alsharari et al., (2015) and colleagues found that menthol (15–45 mg/kg) slows Nic metabolism in mice by 40 %. Mixtures of menthol and Nic cause a leftward shift in the Nic self-administration dose-effect curve, increasing intake of near-threshold reinforcing doses (Biswas et al., 2016). This indicates increased potency and abuse liability of Nic, which is consistent with the ability of β-Nic to increase the abuse liability of Nic in the present study. In humans, clinical surveys have shown that people who smoke mentholated cigarettes tend to show greater metrics of tobacco use disorder (i.e., time to first cigarette, cigarettes per day, etc.; Cohn et al., 2022, 2025; Leas et al., 2022). Collectively, these clinical and preclinical studies suggest that pharmacological manipulations that slow nicotine metabolism can both reduce and enhance abuse liability metrics of Nic depending on Nic use history (naïve vs. experienced) and the direct CNS effects of the inhibitor (i.e., menthol alone is reinforcing; Nesil et al., 2019).

Similarly, epidemiological research has found that adolescents with slower Nic metabolism have a greater likelihood of developing tobacco use disorder. In contrast, adult smokers with slower Nic metabolism tend to smoke fewer cigarettes per day, be less dependent, and be more successful at smoking cessation (see a review by Jones et al., 2022). The mechanism(s) accounting for this age difference in the impact of slower metabolism on Nic abuse liability are unclear, but may stem from developmental differences in nAChR expression (Placzek et al., 2009; Zeid et al., 2018), neuroplasticity (Rogers et al., 1998), or compensatory mechanisms in metabolic pathways (see a review by Colyer-Patel et al., 2023). Future studies on how these factors might mediate age differences in the relationship between Nic metabolism and tobacco use are warranted.

4.3. Inability of β-Nic to influence the effects of Nic on ICSS

Given that β-Nic can increase the duration of Nic’s discriminative stimulus effects (Smethells et al., 2025), the inability of β-Nic to prolong the effects of Nic on ICSS in the present study was unexpected. However, whereas there was a clear difference in nicotine’s discriminative stimulus effects between the 10- and 60-min pretreatment intervals in Smethells et al. (2025), effects of nicotine on ICSS did not differ between these timepoints in the current study (see Fig. 6). The use of a longer pretreatment interval that produces a clearer difference in nicotine’s effects on ICSS (e.g., 120 min) may be necessary for β-Nic’s effects on Nic pharmacokinetics to manifest in an ICSS model. Other factors that differ between the discrimination and ICSS models (e.g., chronic nicotine exposure in the discrimination model, differences in neurobiological mechanisms, etc.) could also account for this discrepancy between studies.

4.4. Implications for tobacco harm reduction and regulation

The present findings are a significant contribution to ENDS research and have important implications for the use of ENDS as a harm reduction approach. Switching smokers to ENDS has been proposed as a harm reduction strategy to decrease the mortality and morbidity associated with combusted tobacco use (Hatsukami et al., 2004). Our findings suggest that the reinforcing efficacy of ENDS and their ability to substitute for combustible cigarettes in adult smokers might be enhanced by increasing the β-Nic levels in ENDS liquid and aerosol (e.g., via oxygenation of ENDS liquids), consistent with the nicotyrine hypothesis proposed by Abramovitz et al., (2015). However, adolescents could be especially sensitive to the influence of β-Nic and thereby more vulnerable to initiation of ENDS use, as those with slower Nic metabolism are more vulnerable to develop a tobacco use disorder (Karp et al., 2006; Ray et al., 2009; Rubinstein et al., 2013; but see Tyndale and Sellers, 2001). These insights have translational relevance for understanding how tobacco product constituents that alter Nic clearance may exacerbate addiction risk, particularly in vulnerable populations like adolescents, and highlight the need for targeted regulatory strategies to mitigate such interactions in novel nicotine delivery systems.

The present findings also have important implications for FDA regulation of ENDS. First, they suggest that the FDA may want to consider classifying β-Nic as an addiction related Harmful and Potentially Harmful Constituent (HPHC). Briefly, a HPHC is any chemical compound (i.e. constituent) in tobacco products that has the potential to cause harm to smokers or nonsmokers (i.e., addiction, cancer, etc.). This designation would support the FDA requiring manufacturers to monitor, report and potentially limit β-Nic levels in ENDS solutions. Second, the FDA could regulate manufacturing processes to limit β-Nic formation, such as by setting limits on device temperature/wattage, restricting practices like deliberate oxygenation, aging, or direct addition, and establishing rules to prevent significant β-Nic accumulation in e-liquids.