Animals.

Studies used Sprague Dawley (SD) or Holtzman Sprague Dawley (HSD) rats (Envigo, Madison, WI or Charles River, PA) depending on availability and were double housed under a 12/12-h standard light/dark cycle, free-fed, and testing occurred during the light phase. In self-administration studies SD rats were individually housed under a 12/12-h light/dark cycle, food was restricted to 18–20 g/day, and self-administration sessions occurred during the dark phase of the light cycle. Food restriction was imposed for self-administration rats because it is healthier than free-feeding for rats during long-term studies and can facilitate acquisition of drug self-administration [30,31]. Self-administration occurred during the dark phase because this is typical for studies using short-access sessions and it models the time of day when smoking (i.e., nicotine self-administration) occurs in humans [26,31]. Male and female rats were used in different experiments to increase heterogeneity. Female rats were used during initial studies when availability of monoclonal antibody was limited (Experiments 1 and 3). Both male and female rats were used for critical experiments (Experiments 2, 6 and 7). Experiment-specific details are provided below.

Donation of PBMCs from previously vaccinated individuals.

A clinical sample collection study designed to collect blood from subjects who had received repeated injections of a nicotine conjugate vaccine (3’AM-Nic-rEPA, consisting of the racemic hapten 3’-aminomethyl-nicotine (3’AM-Nic), the structure of which is published elsewhere [32], linked to the amino acid side-chain and terminal amino groups of Pseudomonas aeruginosa exoprotein A (rEPA) via a succinyl linker) for smoking cessation in the past 10 years was conducted by Optimal Research (Rockville, MD and surrounding areas). Subjects (male and female) between 21 and 75 years old who were immunocompetent, had hematological test results that were in the normal range, and in good general physical and mental health, were recruited as specified in the study protocol (ATI-1501). ATI-1501 is a blood sample collection clinical protocol that included routine procedures similar to standard blood donation practices as follows: After screening for eligibility and provision of informed consent, 450-mL blood samples were obtained from each of three participants and processed to isolate PBMCs. Personal identifiable information was inaccessible to all individuals who were not directly conducting the study.

Nic•mAb lead generation.

IgM+/IgD+ cells were depleted from the PBMC pools collected under protocol ATI-1501 to enrich the population of IgG+ B-cells. After seeding 200–250 IgG+ B-cells per well in a 96-well plate, the B-cells were transformed by infection with the Epstein-Barr virus and co-cultured in the presence of CD40L-expressing feeder cells for amplification. After 2 weeks, the culture supernatants were screened for the presence of secreted anti-nicotine antibodies in the ELISA protocol outlined below using racemic 3’-aminomethyl-nicotine-polyglutamic acid (3’AM-Nic-pGlu) as the coating antigen for this initial screening, indicating that at least one B-cell within the pool in the well was producing anti-nicotine antibodies. After identification of a positive B-cell pool, the corresponding cells were deposited on an ISAAC^™ (ImmunoSpot Array Assay on a Chip) microarray [33] coated with racemic 3’AM-Nic-pGlu conjugate at 10 μg/ml. The array consists of wells designed to accommodate a single lymphocyte in each well. After a 3 h incubation, cells secreting nicotine-specific antibodies were identified by staining with a fluorescently labelled secondary anti-IgG antibody (anti-human IgG(Fab’)₂-Cy3), which stains around the wells in which a B lymphocyte had secreted a nicotine conjugate-specific IgG. Single antibody-positive cells were retrieved by micromanipulation, and the DNA sequences for the IgG light- and heavy-chains in those cells were isolated by RT-PCR and cloned into expression vectors to produce recombinant IgG1 candidates via transient transfection of CHO-S cells and purification by Protein A affinity chromatography. The purified mAbs were then dialyzed into phosphate buffered saline (PBS; 20 mM phosphate, 150 mM NaCl, pH 7.4), and concentration determined by A₂₈₀ using an estimated extinction coefficient of 1.4 based on the protein sequence. In summary, 54 million total B-cells were screened, 146 B-cell pools were identified as 3’AM-Nic-pGlu binders in ELISA and single cells could be retrieved for 64 hits. These hits were then further qualified by competition ELISA to S-(–)-nicotine while using 3’AM-S-(–)-Nic-pGlu conjugate as the plating antigen in order to focus only on mAbs specific to S-(–)-nicotine and eliminate those specific to pGlu or an epitope covering the linker, and counter-screened in competition ELISA with S-(–)-cotinine, also using 3’AM-S-(–)-Nic-pGlu conjugate as the plating antigen. Finally, 32 novel antibodies were progressed to the production/purification phase.

Conjugate binding ELISA.

The conjugate used in the ELISA assay (except for the initial screening described above) was an enantiopure form of the racemic hapten used in the original vaccine, since we were interested in monitoring the binding to the S-(–) form of nicotine exclusively. Briefly, ELISA plates were coated with 1 ng/ml of 3’AM-S-(–)-Nic-pGlu conjugate diluted in coating buffer (0.1M sodium phosphate) overnight at 4°C. Plates were blocked for 1 h at room temperature with blocking and dilution buffer (1% Non-Fat Dry Milk (NFDM) in 1X PBS). Dilutions of sample (culture supernatant or purified mAbs) in PBS + 1% NFDM were added to the blocked wells and incubated for 90 min at 37°C. Plates were washed 5 times with washing buffer (1X PBS + 0.1% Tween 20) and incubated for 60 min at 37°C with Goat α-Human IgG-ɣ-HRP conjugate diluted 1/10,000 in dilution buffer. After washing the plates 5 times in washing buffer, TMB (3,3’,5,5’-tetramethylbenzidine) substrate (KPL) was added and plates incubated for 30 min at room temperature. Reaction was stopped by addition of 1 M H₃PO₄, and absorbance read at 450 nm. A qualitative competition assay was carried out using a similar protocol, but pre-incubating the samples with either free S-(–)-nicotine (1 μM) or S-(–)-cotinine (50 μM) as competitors prior to addition to the coated and blocked ELISA plates. Candidates displaying ≤60% competition with S-(–)-nicotine or >20% competition with cotinine were eliminated. This step was included to ensure elimination of hits specific to pGlu or an epitope covering atoms at the interface of S-(–)-nicotine, pGlu and the amide bond connecting both; as well as candidates cross-reacting with cotinine.

Nicotine assay.

Blood and brain were collected as described within each experiment. Trunk blood or venous blood were centrifuged at 2,000 x g to obtain serum. Brains were rinsed in water and both serum and brains were stored at -20 °C until measured via gas chromatography. Nicotine concentrations in serum or brain were measured using gas chromatography with nitrogen-phosphorus detection [34,35]. Concentrations that were below the 2 ng/mL limit of quantitation for the assay were treated as being 1 ng/mL for the purposes of analysis. Brain nicotine levels were corrected for brain-blood content as previously described [34].

Control mouse monoclonal antibody NIC9D9 preparation and in vitro characterization.

The murine nic•mAb NIC9D9 (licensed from The Scripps Research Institute, La Jolla, CA by Antidote Therapeutics, Inc.) was selected as a reference antibody and produced from a hybridoma cell line [36,37]. This nic•mAb has been previously characterized with high specificity for S-(–)-nicotine [38]. The clarified and filtered NIC9D9-containing supernatant was concentrated, and buffer exchanged by tangential flow filtration, and the solution titrated to pH 5.0 with acetic acid. The material was loaded onto a 20 mL Capto S cation exchange column (GE Life Sciences), and the antibody eluted in 1–2 column volumes of 200 mM sodium acetate, pH 5.5. Relevant fractions were pooled and dialyzed against a 200-fold volume of 1x PBS, and purified antibody concentrated to 10 mg/mL using Amicon Ultra centrifugal filter units (Millipore). Purity was >90% by SDS-PAGE, and functional quality control assays were conducted analyzing binding to a S-(–)-nicotine-BSA conjugate in ELISA format (essentially as described in Isomura et al. 2001 [37]).

Affinity of monoclonal antibodies for nicotine.

The affinity for free S-(–)-nicotine in solution of the Nic•mAb candidates displaying >60% competition with S-(–)-nicotine or ≤20% competition with S-(–)-cotinine was determined using the BiOptix 404pi enhanced SPR instrument. Antibodies were immobilized on a BiOptix CMD200m sensor chip using standard EDC/NHS amine coupling with blocking of the remaining active esters with ethanolamine. S-(–)-nicotine was injected at different concentrations (3x serial dilutions) in running buffer (10 mM HEPES (pH 7.4), 150 mM NaCl, 3 mM EDTA, 0.05% Tween-20), followed by buffer flow at 25°C to monitor dissociation. K_D values were calculated based on the on- and off-rate constants using the Scrubber2 software (BioLogic Software, Campbell, Australia).

To confirm the findings using SPR, a solution binding experiment was performed by equilibrium dialysis using ³H-nicotine on microtiter plates as previously described [39]. Briefly, duplicate samples of a given nic•mAb were diluted to a concentration of approximately 2.8 μg/mL and added to one side of the membrane in 96-well equilibrium dialysis plates. The dialysate side consisted of a range of nicotine concentrations (8–256 ng/mL) mixed with approximately 10⁴ DPM/0.1 mL radiolabeled nicotine. Plates were placed on a shaker for 72 h at room temperature. The radiolabeled nicotine was used to measure nicotine concentrations on both sides of the membrane. Nicotine binding affinity (represented by the K_D) was measured from One-site–Specific Binding using Prism 8.0 (GraphPad Software Inc., La Jolla CA).

Antibody selectivity.

The selectivity of the nic•mAbs 5G4, 7A8, 12F5, and 8D1 for S-(–)-nicotine was determined by competitive ELISA using 3’AM-S-(–)-Nic-pGlu-coated plates essentially as described above. Serial dilutions (in a blocking/diluent buffer of PBS + 1% non-fat dry milk) of competitors were added to the coated and blocked plates. Subsequently, the nic•mAb was added at a fixed concentration yielding a non-saturating A₄₅₀ level in control wells without competitor. After 45 min incubation at 37°C, the plates were washed, and detection of bound nic•mAb was done as previously described. Small molecule competitors included S-(–)-nicotine, acetylcholine chloride, β-nicotinamide adenine dinucleotide (NAD), bupropion, S-(–)-cotinine, cytisine, dopamine (3-hydroxytyramine) hydrochloride, mecamylamine hydrochloride, nicotinamide, (±)-norepinephrine (+)-bitartrate, (±)-nornicotine, serotonin hydrochloride, and varenicline tartrate (all from Sigma). These compounds include neurotransmitters, nicotine metabolites, structurally similar compounds, and smoking cessation drugs. The cross-reactivity was calculated from the ratio of the IC₅₀ values as follows: (IC₅₀, nicotine/IC₅₀, competitor) × 100%.

Fully human Nic•mAb preparation.

To reduce/eliminate Fc effector function (complement-dependent cytotoxicity or antibody-dependent cellular cytotoxicity), which is not needed for binding and neutralizing nicotine and could be a liability in the event of cross-reactivity with membrane targets, 5G4, 7A8, 12F5, and 8D1 were reformatted from the IgG1 to the IgG4 isotype [40] and named ATI-1010 through ATI-1013, respectively. Additional Fc sequence modifications were concurrently made to prevent Fab-arm exchange in vivo [41]. The fully human nic•mAbs ATI-1010, 1011, 1012, and 1013 were produced in sufficient quantity for in vivo testing using the ExpiCHO Expression System (Thermo-Fisher). Essentially, ExpiCHO cells were expanded at 37°C in 2L shake flasks. Cultures were adjusted to 6x10⁶ cells/mL and transfected per the vendor-supplied protocol. Cultures were fed on days 1 and 5 post-transfection, and the temperature was lowered to 32°C on day 1. Cultures were harvested on day 12 by centrifugation, 0.22 μm filtered, and stored at 4°C. Purification was performed on a Protein A affinity column (MAbSelect resin (GE Life Sciences)) with glycine pH 3.0 elution. After neutralization and pooling of the main peak fractions, the material was exchanged into PBS by dialysis. Purity was ≥95% by SDS-PAGE with Coomassie blue staining.

Stable cell pool generation and expression for production of ATI-1013.

For the larger quantities of ATI-1013 required for conducting the in vivo addiction and non-GLP toxicology studies, stable-pools of ATI-1013 were generated in CHO-DG44 cells (Aragen Bioservices). Following transfection by electroporation and selection for stable transfectants, mini-pools were generated and gene amplification was conducted using gradually increasing methotrexate (MTX) and G418 (Geneticin) selection over 3 weeks in a stepwise manner. The top 120 mini-pools were expanded and evaluated with normalized cell densities in a 24-well format to obtain a more accurate ranking. The 20 mini-pools with the highest titers from the 24-well plates were evaluated in shake flasks (50 mL Gibco CD OptiCHO + 8 mM L-glutamine; ThermoFisher Scientific), and mini-pool 117 gave the highest titer. SDS-PAGE resulted in expected band sizes. Further scale-up to 1.4L shake flask reached a peak cell density (18 x 10⁶ cells/ml) at day 11 with harvest at day 14 with 65% viability yielding 1.3 g/L titer at harvest. Mini-pool 117 was then scaled-up further to 16L in a WAVE^™ bioreactor (GE Life Sciences). Purification was conducted using Protein A affinity purification (MabSelect SuRe, GE Life Sciences), with elution in 0.1 M Glycine, pH 3.0, and neutralization using 1.0 M Tris, pH 8.0, followed by flow-through anion exchange chromatography (CaptoQ; GE Life Sciences). Concentration and buffer exchange into formulation buffer (25 mM L-Histidine, 250 mM Sucrose, pH 7.0) was performed using tangential flow filtration (Sartorius Vivaflow 200 30kDa). Final purity was >95% by SDS-PAGE with a monomeric purity of >99% as estimated by SEC.

Experiment 1: Effect of Nic•mAb on distribution of a single nicotine dose.

Adult female SD rats (n = 8/group) were obtained with a jugular venous catheter in place. ATI-1010, 1011, 1012, and 1013 were dosed at 20 mg/kg i.p. 2 h prior to i.v. dosing of 0.03 mg/kg nicotine via jugular catheter. NIC9D9, a mouse monoclonal antibody with high affinity for nicotine (K_D = 43 nM) [38] and known in vivo activity, was included as a positive control. PBS alone (vehicle) was administered 1 mL/kg as a negative control. Blood and brains were collected 3 min after drug administration, processed, and assayed as described above (see Nicotine assay).

Experiment 2: Effect of escalating Nic•mAb doses on distribution of a single nicotine dose.

To investigate dose effects of Nic•mAb on distribution of nicotine, an experiment identical to Experiment 1 was performed using age-matched groups of male and female adult SD rats (n = 8/group, 4M and 4F per group) except that rats received vehicle, 10, 20, or 40 mg/kg i.p. of ATI-1012 and ATI-1013 2 h prior to administration of 0.03 mg/kg nicotine i.v.. Rats that had less than 5 μg/mL serum antibody levels (indicating incomplete nic•mAb administration due to catheter malfunction), as measured by ELISA, were excluded from the analyses. The excluded animals had an average serum antibody level of 0.73 μg/mL. The number of excluded animals are as follows for ATI-1012 administration (parentheses indicate the ATI-1012 mg/kg dose): 2 (10), 2 (20), and 4 (40); and for ATI-1013 administration (parentheses indicate the ATI-1013 mg/kg dose): 2 (10), 3 (20), and 2 (40).

Experiment 3: Estimation of Nic•mAb pharmacokinetic parameters.

Adult female SD rats (n = 3/group) were obtained with a jugular venous catheter in place (Charles River, PA). Rats received 5 mg/kg i.v. ATI-1012 or ATI-1013 via tail vein. Blood (0.4 mL) was collected into serum separator tubes via the jugular catheter at pre-dose, at 10 and 30 min, and at 2, 6, 12, 24, 48, 72, and 120 h. Serum was isolated and stored at -20°C until analysis. The serum levels of ATI-1012 or ATI-1013 were determined using a concentration-based ELISA and quantitated against serial dilutions of the dosed mAb.

Experiment 4: Effect of Nic•mAb on distribution of repeated nicotine doses.

Adult HSD rats were anesthetized with 0.1 mg/kg fentanyl, 0.05 mg/kg dexmedetomidine i.m., and 100 mg/kg propofol i.p., and a jugular venous catheter was placed. Two groups of rats (n = 10/group, 5M and 5F per group) received ATI-1013 40 or 80 mg/kg i.v. via jugular catheter and one group received 80 mg/kg of non-specific human polyclonal IgG (Gammagard; Baxter Healthcare Corp., Westlake Village, CA) as a control. Twenty-four hours later, 0.03 mg/kg nicotine was dosed i.v. once every 10 min via jugular catheter for a total of 5 doses. Blood was sampled 3 min after the first dose of nicotine. Animals were sacrificed 3 min after the fifth nicotine dose, i.e., 43 min after the first nicotine dose and blood and brain collected. Three rats (1 female that received 40 mg/kg ATI-1013 and 1 female and 1 male that received 80 mg/kg ATI-1013) had high brain nicotine levels, suggesting contamination during measurement, were removed from analyses. Outliers were identified using ROUT (Q = 1%).

Experiment 5: Effect of Nic•mAb on self-administration of nicotine.

Adult male SD rats were initially trained for nicotine self-administration (NSA) using a unit nicotine dose of 0.03 mg/kg under a fixed-ratio (FR) 3 schedule during 2 h sessions using our routine protocol [42,43]. This unit dose is a common training dose for nicotine self-administration and was chosen to facilitate acquisition [31]. Sessions were conducted five days per week (Monday through Friday). After stable NSA was established (at least 5 infusions per session, an active:inactive lever ratio of at least 2:1, and no trend in the number of infusions over five consecutive sessions), the unit dose was reduced to 0.015 mg/kg, which results in serum nicotine concentrations during 2 h sessions that are more similar to smoking in humans [44,45]. After NSA stabilized at this unit dose (same criteria as above) on a Friday, rats (n = 7/group) began receiving twice-weekly i.v. infusions of 160 mg/kg ATI-1013 or 160 mg/kg Gammagard (control mAb) 30 min prior to the session on the following Mondays and Thursdays while rats continued NSA at the 0.015 mg/kg dose for 10 consecutive sessions (2 weeks). Then, the unit nicotine dose was reduced to 0.0075 mg/kg on a Monday for another 10 consecutive sessions (2 weeks) while the twice-weekly nic•mAb treatment continued. A higher dose of ATI-1013 was chosen for this experiment because the self-administered cumulative nicotine dose was higher (over twice in some rats) than the experimenter-administered nicotine doses in the other experiments.

Experiment 6: ATI-1013 repeat dose pharmacokinetics experiment.

Age-matched adult SD rats (n = 6/group, 3M and 3F per group) were obtained with femoral vein and jugular vein catheters in place (FVC and JVC, respectively; Charles River, PA) received i.v. doses via the FVC of 40 mg/kg ATI-1013 weekly for 4 weeks. Residual nic•mAb concentrations were measured in serum from blood draws via the JVC at various time points after i.v. dosing. The ELISA detection assay employed relies on nicotine conjugate binding to 3’AM-S-(–)-Nic-pGlu [29] and thus reflects functional nic•mAb levels binding S-(–)-nicotine in serum. The first 4 time-points after ATI-1013 administration (but before the second dose was given) were used to measure ATI-1013 half-life to validate the findings from Experiment 3. At the end of this study, rats were dosed with 0.03 mg/kg i.v. nicotine and sacrificed 3 min later and samples were analyzed to assess the amount of unbound nicotine before and after ultrafiltration [46].

Experiment 7: Toxicology assessment of high doses of ATI-1013.

The purpose of this study was to evaluate the repeat-dose tolerability of four (4) fixed i.v. doses of 200 mg/kg ATI-1013 dosed once weekly in the presence of 1 mg/kg/day of nicotine given continuously by s.c. for 28 days at Noble Life Sciences, beginning on study day 1.

Weight matched (225-300g) female and male SD rats were divided into four groups (n = 16/group, 8M and 8F per group). Animals were implanted with osmotic pumps delivering a continuous dose of nicotine or vehicle through the study period. ATI-1013 was delivered weekly by intravenous injection. Group 1 received saline, group 2 received nicotine only, group 3 received ATI-1013, and group 4 received ATI-1013 and nicotine.

Body weights were collected twice weekly and cage-side observations made daily. On study day 29, blood was collected via the jugular vein under anesthesia and analyzed for hematology, serum chemistry, and coagulation. Additional aliquots of blood were taken for assaying serum levels of ATI-1013. After blood collection animals were placed under anesthesia with isoflurane and sacrificed by thoracotomy. Tissues were harvested and weighed by tissue with kidneys, testis, and ovaries weighted as a pair.

Assessment of toxicity was based on mortality, clinical observations, and body weight during the 28 d study and at the end of study on organ weights, gross anatomic pathology, hematology, serum clinical chemistry, and coagulation. Tissue histopathology evaluations for heart, liver, lung, kidney, spleen, skeletal muscle, brain, colon, stomach, ovary, and testis have been conducted. Tissues were fixed immediately in formalin, embedding in paraffin, staining with H&E, and were reviewed by a veterinary pathologist.

Data analysis.

In the single-dose nicotine pharmacokinetic studies serum or brain nicotine concentrations were compared by Welch’s one-way ANOVA with Dunnett’s T3 multiple comparisons test. Non-parametric Kruskal-Wallis testing was also conducted to assess the robustness of ANOVA’s conclusions based on parametric assumptions for the relatively small sample size used in these experiments. Estimates of pharmacokinetic parameters (volume of distribution, clearance, terminal half-life) were obtained from serum concentrations using noncompartmental analysis (PKSolver) [47]. In the multiple dose nicotine pharmacokinetic studies, serum and brain nicotine concentrations were compared between the different doses of ATI-1013 and IgG groups using two-tailed unpaired t-tests with Welch’s correction for unequal variances. For nicotine self-administration studies, the primary endpoint was the mean number of infusions during the last 3 sessions (Wednesday—Friday) at baseline and the last 3 sessions (Wednesday—Friday) at each unit dose during nic•mAb treatment. Because variances were unequal between groups, means were compared between groups using independent-samples t-tests with Welch’s correction and a significance level of p = 0.017 to account for the three comparisons. Mean infusions within each group between baseline and at each unit nicotine dose post-mAb administration were similarly analyzed but using paired t-tests.