This study used cigarette smoke exposure in an effort to capture some of the features of nicotine delivery to a cigarette smoker that are not provided by parenteral nicotine dosing. Rodent models of cigarette exposure have been used widely to study tobacco smoke toxicology, particularly carcinogenesis [13
]. Much less attention has been directed at using smoke exposure systems to study nicotine pharmacology or pharmacokinetics, and most such studies have not measured serum or tissue concentrations of nicotine to determine how they relate to the exposures of cigarette smokers, or carboxyhemoglobin concentrations to assure lack of toxicity from carbon monoxide exposure. An exception is a NSE paradigm exposing rats to 10 pulses of cigarette smoke over 10 min which was used to model smoking topography and which found nicotine absorption from smoke to be rapid, and nicotine elimination to be comparable to that of nicotine administered i.v. [15
]. Precise characterization of exposures may not be necessary for many types of smoke pharmacology studies, but is critically important when evaluating an intervention such as addiction vaccines which act by altering nicotine pharmacokinetics. The smoke exposure system used in this study has been characterized with regard to nicotine and carbon monoxide exposures over a range of clinically relevant smoke exposure conditions, allowing its use to more closely and quantitatively model the exposures of smokers [19
]. Clinically relevant nicotine exposures can be obtained without excessive carboxyhemoglobin accumulation or overt evidence of behavioral toxicity. This type of well-defined exposure should be useful in studying nicotine vaccines, as well as other aspects of tobacco smoke pharmacology and the modifying effects of non-nicotine components of tobacco smoke.
The main finding of this study is that vaccination with 3′-AmNic-rEPA substantially altered the distribution to brain of nicotine delivered via cigarette smoke inhalation, with a reduction of 90% after a single 10 min exposure modeling the smoking of 1 cigarette, and a reduction of 35% after a 2 hr exposure modeling a period of more sustained smoking. The smaller reduction after longer exposure is consistent with previously reported effects of this vaccine on nicotine administered i.v., which are greatest when the nicotine dose is low, when brain nicotine concentrations are measured shortly (1–25 min) after a nicotine dose, and when nicotine is delivered via bolus doses as compared to continuous infusion [36
The similar serum nicotine concentrations in control groups, immunized with the unconjugated carrier protein, after 10 min NSE or 10 min i.v. infusion confirms that the nicotine dose delivered by 10 min NSE was in the target range of 0.015 mg/kg, similar to mg/kg estimates of nicotine uptake from a single cigarette by a smoker [37
]. Vaccination substantially reduced nicotine distribution to brain whether nicotine was administered via smoke or by itself as an i.v. infusion, showing that the effects of vaccination on nicotine distribution are quite robust despite marked differences in route and manner of nicotine administration. These findings support the validity of the i.v. dosing paradigm in rats as a model of vaccine effects on cigarette smoking in humans. Several comparisons suggested that vaccination might be minimally more effective in the setting of NSE exposure than IV infusion exposure, but these were probably incidental because brain nicotine concentrations, the most important measure of efficacy, were comparable after these exposures.
The 2 hr WBE exposure data are of interest because this exposure more closely approximates nicotine intake in someone smoking repeatedly throughout the day. The greater efficacy of vaccination in rats receiving the smaller (10 min) nicotine exposures compared to the 2 hr WBE exposures, respectively achieving a brain reduction of 90% and 35%, can be put in perspective by estimating the relative molar ratios of drug to binding capacity of antibody. The total amount of nicotine-specific IgG in vaccinated rats can be estimated from the product of the measured serum concentration (mean of the vaccinated groups) and the reported steady state volume of distribution of IgG in rat (268 μg/ml × 125 ml/kg = 34 mg/kg = 0.22 μmol/kg), with 2 binding sites per IgG resulting in a binding capacity of 0.45 μmol/kg [38
]. Using this estimate, the 10 min i.v. exposure (15 μg/kg = 0.093 μmol/kg) was equivalent to 21% of the estimated binding capacity of nicotine-specific IgG present. In this setting, there was a large excess of binding capacity compared to the nicotine dose. The nicotine dose delivered by the 2 hr WBE is not known but in a previous study similar serum nicotine levels were produced by an i.v. nicotine infusion of 125 ug/kg/hr (1.5 μmol/kg) infused over 2 hr, which is equivalent to 330% of the estimated binding capacity of antibody [20
]. The reduction in nicotine distribution to brain after 2 hr WBE despite this excess of nicotine compared to antibody is of clinical interest because serum nicotine-specific IgG concentrations that have been achieved in vaccinated smokers in clinical trials (~50 μg/ml) are substantially lower than those routinely achieved in vaccinated rats (~250 μg/ml), yet vaccination appears to enhance smoking cessation rates [2
A potential difference between inhaled and parenteral i.v. nicotine dosing is that absorption of nicotine via inhalation could be impacted by binding to pulmonary mucosal antibody. If present in sufficient amount, such antibodies could reduce or slow nicotine uptake by providing a barrier to absorption into blood. In this study both nicotine-specific IgG titers and nicotine concentrations were increased in the BAL fluid of vaccinated rats, and their values were positively correlated, showing retention of nicotine in pulmonary mucosal fluid by antibody. The fraction of the nicotine dose retained in pulmonary mucosal fluid by immunization cannot be calculated because the extent of recovery of pulmonary mucosal antibody in BAL fluid is not known. The absolute amount of nicotine retained in BAL fluid by immunization (the difference between immunized and control groups) was very low, equivalent to <0.1% of the 10 min smoke exposure dose and considerably less for the 2 hr smoke exposures. Even assuming that recovery of pulmonary mucosal fluid in BAL fluid is inefficient, it is unlikely that nicotine binding in pulmonary mucosal fluid contributed appreciably to the kinetics of nicotine absorption from cigarette smoke.
The presence of nicotine-specific IgG in BAL fluid could reflect local production at the pulmonary mucosa or transfer of antibody from blood. Because nicotine was retained in BAL fluid to a similar extent by both vaccination and passive immunization with Nic311, nicotine-specific IgG must have been transferred from blood. Although nicotine-specific IgG was detected in BAL fluid after vaccination, Nic311 was not detected in BAL fluid after passive immunization. This is probably due to the lesser sensitivity of the ELISA assay for Nic311 compared to antibodies generated by vaccination, along with the somewhat lower levels of Nic311 in blood compared to nicotine-specific IgG in vaccinated rats.
The role of drug-specific IgA in mediating the effects of nicotine vaccines, or other addiction vaccines, is not known. In the current study nicotine-specific IgA was not detectable in BAL fluid after vaccination, indicating that there was no appreciable accumulation of nicotine-specific IgA in pulmonary mucosa fluid. This is not unexpected since parenteral vaccination is not typically a potent stimulus for mucosal antibody production [39
]. Intranasal vaccination of mice against nicotine, which delivers immunogen to the lung as well as the upper respiratory tract, has been shown to reduce nicotine distribution to brain in mice. Serum and saliva nicotine-specific IgA titers were detected, but BAL titers were not examined [40
]. Intranasal or direct pulmonary immunization might be of interest to further explore whether sufficient pulmonary mucosal nicotine-specific IgA levels can be produced in this manner to influence nicotine absorption [41
In contrast to the lack of measurable nicotine-specific IgA in BAL fluid, nicotine-specific IgA titers in serum after vaccination were substantial with values 21–25% as high as the nicotine-specific IgG titers. Serum IgA titers resulting from parenteral vaccination against nicotine have not been reported previously. Accurate estimates for the volume of distribution of IgA are not available to allow calculation of total body nicotine-specific-IgA content, but these appreciable titers suggest that nicotine-specific IgA could contribute to the binding of nicotine and effects of vaccination.
A limitation of this study is that passive inhalation of cigarette smoke by rats does not reproduce the pulsatile puffing and deep inhalation of cigarette smoking by humans. Pulsatile uptake of nicotine and delivery of nicotine to brain is widely regarded as important for the rewarding and reinforcing efficacy of nicotine, since more rapid delivery of drug to brain is in general more reinforcing than slow delivery [9
]. However a recent human imaging study found that brain uptake of 11
C-nicotine from spiked cigarettes is damped by transit through the lungs [44
]. Patterns of brain nicotine uptake were variable, with a pulsatile character clearly present in some subjects but less so or absent in others. Since all subjects were experienced smokers, puff-related pulsatile nicotine uptake into brain may not be essential to maintaining smoking behavior, or a necessary feature of a smoke inhalation model in rats.
In summary, vaccination with the nicotine immunogen 3′-AmNic-rEPA substantially reduced the distribution to brain of nicotine delivered via cigarette smoke inhalation. These data add validity to the i.v. nicotine dosing regimens used previously to model tobacco use in rats, and provide a means of studying the role of pulmonary antibody or other route-specific factors in modifying the pharmacokinetics of nicotine inhaled from cigarette smoke.