In Vivo Molecular Imaging Analysis of a Nasal Vaccine That Induces Protective Immunity against Botulism in Nonhuman Primates Free

Nasal administration is one of the most effective routes for the induction of Ag-specific protective immunity in both systemic and mucosal compartments. A nasally delivered, live attenuated influenza vaccine, FluMist, was recently approved in the United States (1); however, an inactivated nasal influenza vaccine with a heat-labile enterotoxin adjuvant was associated with facial paralysis in a study in Switzerland (2). Toxin-based adjuvants, composed of cholera toxin or cholera toxin B subunit, are redirected to the olfactory bulb (OB) in the CNS when administered via the nasal route in mouse (3). The results of this mouse study suggest that the deposition of vaccine or adjuvant, or both, in the CNS via the nasal route might cause adverse effects. Thus, a key issue to investigate during nasal vaccine development is whether a candidate vaccine, with or without adjuvant, could reach the CNS. Tracking of nasally administered vaccines should take into account the anatomical differences between mice and humans in the olfactory epithelium of nasal cavity. Mice and dogs harbor well-developed olfactory epithelium that covers 70–80% of the total surface area of the nasal cavity, whereas that of nonhuman primates and humans covers only ∼10% of the total surface area of the nasal cavity (4). Therefore, the evaluation of a nasal vaccine by safety pharmacology studies, such as absorption, distribution, metabolism, and excretion (ADME) studies, must be performed in multiple species, including nonhuman primates, before clinical trials in humans.

Traditionally, the evaluation of ADME in the development of biotechnology-derived pharmaceuticals has involved autoradiography of the whole body and the detection of radioactivity associated with dissected tissues by means of radioactive ¹²⁵I- or ¹¹¹In-labeled pharmaceuticals (5). However, recent molecular imaging technologies allow visualization and quantitative measurement of biological processes in living systems (6, 7). Radioisotope-based molecular imaging techniques, such as positron-emission tomography (PET), have been used for the noninvasive detection of pathological changes in cancer and to ascertain functional changes in the nervous system (8, 9). Hybrid technologies combining PET imaging with traditional structural imaging techniques, such as computerized tomography (CT) scanning and magnetic resonance imaging (MRI), have improved the sensitivity and specificity of these analyses (10).

PET imaging agents usually consist of small-molecule tracers, such as glucose analogs or neuroreceptor ligands (9), and rarely consist of proteins (11). In this study, we attempted to develop a PET imaging system that uses a protein tracer for evaluating nasal vaccine candidates in multiple species. For this approach, we chose to evaluate a recombinant form of the nontoxic fragment of the C-terminal half of the H chain of Clostridium botulinum type A neurotoxin (BoHc/A) as a subunit vaccine candidate against botulism (12). It has been shown that systemic immunization of primates with a similar preparation of type B (BoHc/B) resulted in the generation of neutralizing Ab response (13). Because of the potential for botulinum neurotoxin to be disseminated by bioterrorists via airborne or oral routes and to result in a high mortality rate, it was classified as a category A agent by the U.S. Centers for Disease Control and Prevention. Thus, the development of an effective nasal vaccine against botulism is considered a potent strategy to prevent inhalation and gastrointestinal botulism (14).

In this paper, we show that a nasal BoHc/A vaccine candidate provided full protective immunity against toxin-based neurologic complications in nonhuman primates. To assess the safety issue of redirection of the Ag to the CNS, we developed PET with ¹⁸F-labeled BoHc/A ([¹⁸F]-BoHc/A) and evaluated the ADME of the nasal BoHc/A vaccine in both mice and nonhuman primates. In contrast with traditional whole-body autoradiography, the use of PET in combination with CT or MRI allowed noninvasive and dynamic in vivo quantitative imaging of [¹⁸F]-BoHc/A. Both PET and direct tissue counting showed no evidence of CNS deposition of the vaccine Ag in the cerebrum or OB after nasal administration of [¹⁸F]-BoHc/A or the more traditional [¹¹¹In]-BoHc/A in mice and nonhuman primates.

C. botulinum neurotoxin type A (BoNT/A) from C. botulinum type A 62 was purified from the culture supernatant as previously described (15). The toxicity of purified BoNT/A (1.1 × 10⁸ mouse i.p. LD₅₀/mg protein) was assayed by time to death after i.v. injection into mice (16).

The gene coding the nontoxic C-terminal fragment of H chain of type A botulinum neurotoxin (strain 62A; aa 872–1296), BoHc/A, was synthesized by PCR with Phusion High-Fidelity polymerase (Daiichi Pure Chemicals, Tokyo, Japan) and specific primers (sense, 5′-CGTCGGGATCCAATATTATTAATACTTCTATATTG-3′; antisense, 5′-TAGTAGTCGACTTACAGTGGCCTTTCTCCCCATCC-3′; BamHI and SalI restriction enzyme sites shown by underlining). After digestion with BamHI and SalI, the gene fragment was inserted into the plasmid pGEX-6P3 (GE Healthcare, Piscataway, NJ). After DNA sequencing, the plasmid was used to transform Rosetta 2(DE3) pLysS cells (Novagen, Madison, WI), and the cells were treated with 0.1 mM isopropyl-β-d-thiogalactopyranoside for 4 h at 28°C to induce BoHc/A expression. The cells were then suspended in PBS and sonicated on ice. After centrifugation, the supernatant was loaded onto a DEAE Sepharose column (GE Healthcare) in PBS. To purify the GST-fused BoHc/A protein, we loaded the pass-through fraction onto a glutathione Sepharose column (GE Healthcare) in PBS. The fusion protein on the column was treated with PreScission Protease (GE Healthcare) overnight at 4°C to cleave between GST and BoHc/A, and the digested fraction containing BoHc/A was collected. Finally, purification by gel filtration on a Sephacryl S-100 (GE Healthcare) column was performed. Purified BoHc/A was confirmed by liquid chromatography/mass spectrometry (LC/MS-MS; Applied Biosystems, Foster City, CA) analysis after treatment with trypsin gold (Promega, Madison, WI). The level of LPS contamination in the purified BoHc/A (<10 endotoxin units/mg protein) was measured by using the Limulus J Single Test (Wako, Osaka, Japan).

[¹⁸F]-BoHc/A was labeled with N-succinimidyl-4-[¹⁸F]fluorobenzoate ([¹⁸F]SFB). In brief, [¹⁸F]fluoride was produced by the 18O(p, n)[¹⁸F] nuclear reaction by the use of a cyclotron (HM-18; Sumitomo Heavy Industries, Tokyo, Japan) at Hamamatsu Photonics PET Center. [¹⁸F]SFB was prepared from 4-(trimethylammonium triflate) benzoate (17) and [¹⁸F]KF/K[2,2,2] according to the procedure of Tang et al. (18) with some modifications. [¹⁸F]KF/K[2,2,2] complex was added to 5 mg of the precursor in 1.5 ml acetonitrile (CH₃CN) and reacted at 80°C for 10 min. After cooling the reaction mixture, 20 μl 1M tetra-n-propyl-ammonium hydroxide (Pr₄NOH) in 0.5 ml CH₃CN was added and hydrolyzed at 120°C for 5 min. N,N,N′,N′-tetramethyl-O-(N-succinimidyl)uranium tetrafluoroborate (15 mg) in 0.5 ml CH₃CN was added to the reaction mixture and reacted at 80°C for 5 min to produce [¹⁸F]SFB. The reaction mixture was diluted with 1.5 ml of 5% acetic acid (CH₃COOH) and transferred to an HPLC injector. An extra 1.5 ml of 5% CH₃COOH was used to rinse the reaction vessel. The combined crude product was purified by semipreparative HPLC (HPLC conditions: Inertsil ODS-3 column, CH₃CN:H₂O = 300:700, 6 ml/min, detection: 254 nm). The radioactive peak that eluted at 13.5 min was collected, diluted with 30 ml H₂O, and passed through a Sep-Pak C18 cartridge (Nihon Waters K.K., Tokyo, Japan). The [¹⁸F]SFB retained on the cartridge was released with 4 ml CH₂Cl₂ and recovered to a V-vial through a Sep-Pak Dry cartridge (Nihon Waters). The decay-corrected radiochemical yield of [¹⁸F]SFB was 28.1–38.3% (average, 33.5%; n = 4), based on [¹⁸F]fluoride, and the radiochemical purity was >99%. The [¹⁸F]SFB was then concentrated at 60°C under He flow (200 ml/min) and used directly for next step. [¹⁸F]SFB was added to a solution of 0.5 M borate buffer (pH 8.5, 200 μl), PBS (100 μl), BoHc/A (4.1 mg/ml, 200 μl), and allowed to react at room temperature for 30 min. Purified BoHc/A was successfully radiolabeled by conjugation with [¹⁸F]SFB, which reacted with free amino groups including an N-terminal and ε–Lys amino groups in the protein. The product was purified by gel-permeation chromatography (Superose 12, PBS, 1 ml/min), and the radioactive peak eluted at 13.5 min was collected. The 373–698 MBq (average, 548 MBq; n = 4) [¹⁸F]-BoHc/A was obtained in 140 min from the end of bombardment (EOB). The radiochemical purity from EOB was 100%, and the decay-corrected radiochemical yield from EOB was 2.8–4.5% (average, 3.7%; n = 4). The sp. act. was 1.91–2.85 MBq/μg protein (average, 2.46 MBq/μg protein; n = 4). Cholera toxin (List Biological Laboratories, Campbell, CA) was similarly radiolabeled by conjugation with [¹⁸F]SFB and purified by the HPLC. The 126 MBq [¹⁸F]-cholera toxin was obtained in 150 min from EOB. The radiochemical purity from the EOB was 100%, and the decay-corrected radiochemical yield from EOB was 0.85%. The sp. act. of the [¹⁸F]-cholera toxin was 1.42 MBq/μg protein.

BoHc/A and cholera toxin (List Biological Laboratories) were labeled with indium chloride ([¹¹¹In], half-life 2.805 d; Nihon Medi-Physics, Tokyo, Japan) anhydride (Dojindo, Kumamoto, Japan) via N-terminal and ε–Lys amino groups using DTPA as described previously (19). The sp. act. of [¹¹¹In]-BoHc/A and [¹¹¹In]-cholera toxin were 638,000–915,000 (average, 775,000 cpm/μg protein; n = 4) and 825,000–909,000 cpm/μg protein (average, 867,000 cpm/μg protein; n = 2), respectively.

Female BALB/c mice (4–7 wk old) purchased from Japan SLC (Shizuoka, Japan) were used for the direct radioactive counting in tissues after nasal administration of [¹¹¹In]-BoHc/A at the Institute of Medical Science of The University of Tokyo, and for PET/CT imaging and direct radioactive counting in tissues after nasal administration of [¹⁸F]-BoHc/A at PET Center of Hamamatsu Photonics K.K. For the primate study, two groups of monkey housed in two different research centers were used. Four female naive cynomolgus macaques (Macaca fascicularis, 5 y old, ∼3 kg) were used for the immunization/challenge study and were maintained at the Tsukuba Primate Research Center for Medical Science in National Institute of Biomedical Innovation (NIBIO, Ibaraki, Japan). In the separate experiment, two male and naive rhesus macaques (Macaca mulatta, 5–6 y old, ∼5 kg) were used for the PET imaging study at PET Center of Hamamatsu Photonics K.K. All experiments were performed according to the guidelines provided by the Animal Care and Use Committee of The University of Tokyo, Hamamatsu Photonics K.K., and NIBIO.

After nasal administration of 0.90 MBq (0.34 μg) per 20 μl [¹⁸F]-BoHc/A (10 μl, each nostril), mice were anesthetized by i.p. injection of chloral hydrate (300 mg/kg i.p.) and placed in a small-animal PET scanner (Clairvivo-PET; Shimadzu, Kyoto, Japan) (20). Scans were conducted for 10 h. A CT scan was obtained by using a small-animal CT scanner (Clairvivo-CT; Shimadzu) after the PET scan to identify organs and tissues.

Mice (n = 2) were fixed on an acrylic plate with thread and surgical tape (two animals on each acrylic plate) after nasal administration of 0.90 MBq (0.34 μg) per 20 μl [¹⁸F]-BoHc/A. Mice with an acrylic plate were placed at the center position between two planar positron whole-body imaging system (PPIS) detectors and scans with 10-min time frame were performed for 10 h with 60 × 10-min frame using a planar positron imaging system (PPIS-4800; Hamamatsu Photonics, Hamamatsu, Japan) (21). PPIS data were analyzed by using ImageJ software and expressed as counts per pixel. Time-activity curves (TACs) of planar images were expressed as total counts in region of interest.

After nasal administration of 0.90 MBq (0.34 μg) per 20 μl [¹⁸F]-BoHc/A or 0.56 MBq (0.39 μg) per 20 μl [¹⁸F]-cholera toxin to mice (n = 3 at indicated time), the radioisotope counts in various tissues were directly measured by a gamma counter (1480 Wizard; PerkinElmer, Waltham, MA), and the values were adjusted for tissue weight to provide standardized uptake values (SUVs), at the times indicated in Fig. 2A–C. A similar procedure was carried out after nasal administration of mice (n = 3 at indicated time) with 1 × 10⁶ cpm (1.1 μg) per 18 μl [¹¹¹In]-BoHc/A or 1 × 10⁶ cpm (1.1 μg) per 20 μl [¹¹¹In]-cholera toxin, at the times indicated in Fig. 2D–F.

The cynomolgus macaques were nasally immunized with 500 μg per 500 μl BoHc/A (250 μl, each nostril), on four different occasions, at 2-wk intervals. Before immunization and 7 d postimmunization, serum and nasal wash samples were collected for the determination of BoNT/A-specific Ab titers.

BoNT/A (5 μg/ml, 100 μl) was coated onto 96-well microtiter plates overnight at 4°C. Nonspecific binding to the plates was blocked with 1% BSA, and then 2-fold serial dilutions of samples were added and incubated for 2 h at room temperature. After washing, HRP-conjugated goat anti-monkey IgG (Nordic Immunological Laboratory, Tilburg, Netherlands) or HRP-conjugated goat anti-monkey IgA (Cortex Biochem, San Leandro, CA) diluted 1:1000 were added and incubated for 2 h at RT together with 1% normal goat serum (Vector, Burlingame, CA). The reaction was finally developed with the use of the TMB Microwell Peroxidase Substrate System (XPL, Gaithersburg, MD). End-point titers were expressed as the reciprocal log2 of the last dilution, which gave an OD450 of 0.1 greater than the negative control.

After nasal administration of [¹⁸F]-BoHc/A (34.5 MBq [18.1 μg] per 315 μl and 54.7 MBq [28.6 μg] per 500 μl) to male rhesus macaques (n = 2), each macaque’s head was tilted back for 10 min and then scanned in an upright position after anesthesia. PET scans were conducted for 345 min with frames of 25 × 3 min, followed by 27 × 10 min, with the use of a high-resolution animal PET scanner (SHR-7700; Hamamatsu Photonics, Shizuoka, Japan) (22). An MRI scan was obtained by using a MAGNETOM Allegra (3T; Siemens, Munich, Germany) to identify cerebrum regions. After PET imaging, the [¹⁸F] radioactivity within each dissected organ of the macaque, including brain and OB, was directly measured by a gamma counter (1480 Wizard).

PET data were analyzed by means of the PMOD software package (PMOD Technologies, Zurich, Switzerland). Each PET image was superimposed on the corresponding MRI or CT data to identify the volume of interest. SUVs were calculated from the level of radioactivity in the volume of interest. PET images were presented as a maximum intensity projection (MIP) image. TACs of PET images were based on SUVs.

[¹⁸F]-BoHc/A was synthesized after expression of recombinant BoHc/A in Escherichia coli and purification by chromatography (Supplemental Figs. 1 and 2). Mice were nasally administered 0.9 MBq per 20 μl [¹⁸F]-BoHc/A (10 μl per nostril) and were then subjected to anesthesia after 5 min to allow immediate analysis by PET together with CT (PET/CT) or by means of a PPIS. The MIP images of PET data, which were obtained from 10 min to 10 h after nasal administration, were coregistered with a CT surface rendering image, and sites of [¹⁸F]-BoHc/A accumulation were identified by image fusion analysis (Fig. 1A). PET data obtained 50–60 min after nasal administration of [¹⁸F]-BoHc/A were also visualized as an MIP image coregistered with a CT image to produce a three-dimensional view of a mouse (Supplemental Video 1). Quantitative TACs of [¹⁸F]-BoHc/A from the PET data showed that the nasal cavity and urinary bladder were major sites of accumulation of radioactive material after 10 h, whereas the throat, stomach, and intestine were temporary sites of accumulation of radioactive material (Fig. 1B). The whole-body imaging by PPIS (Supplemental Video 2 and Fig. 3) showed dynamic imaging of [¹⁸F]-BoHc/A in the nasal cavity, throat, stomach, intestine, and urinary bladder in mice from 10 min to 10 h after nasal administration. [¹⁸F]-BoHc/A was detected in the nasal cavity throughout the entire examination period.

When the tissue distribution of [¹⁸F]-BoHc/A was quantified by direct counting of radioactivity (Fig. 2A), [¹⁸F]-BoHc/A was found to be preferentially located in urine and in the nasal cavity and, in decreasing order of abundance, in the esophagus, stomach, intestine, feces, nasopharynx-associated lymphoid tissue, and kidney. These data are consistent with the quantitative TACs of [¹⁸F]-BoHc/A in mice obtained by PET/CT imaging (Fig. 1B). Furthermore, we could not detect [¹⁸F]-BoHc/A in the CNS, including the OB and cerebrum (Fig. 2B). The accumulated [¹⁸F]-BoHc/A in the cerebrum and OB were lower than 0.01 and 0.04 SUV within the 6-h period after nasal administration of 0.90 MBq per 20 μl [¹⁸F]-BoHc/A to each mouse, respectively. In contrast, the positive control [¹⁸F]-cholera toxin was detected in the OB at 6 h after nasal administration of 0.56 MBq per 20 μl [¹⁸F]-cholera toxin (Fig. 2C). When the radioactive samples extracted from nasal cavity and esophagus were further examined by size-exclusion chromatography on prepacked disposable (PD-10) columns (GE Healthcare UK, Buckinghamshire, U.K.), [¹⁸F]-BoHc/A was found in the high m.w. fractions. The radioactivity detected in samples from the digestive tract, which included stomach, small intestine, and feces, was present in both high and low m.w. forms, whereas that in the urine sample was present only in low m.w. forms (Fig. 3A).

We used the traditional [¹¹¹In]-radiolabel method to examine the persistence of the nasal BoHc/A vaccine in mouse tissues. Because the half-life of [¹¹¹In] is 67.3 h, we counted the radioactivity in various tissues from 10 min to 72 h after nasal administration of [¹¹¹In]-BoHc/A. The [¹¹¹In]-BoHc/A was detected in the nasal cavity as early as 10 min and remained for at least 24 h after administration. The highest level of radioactivity was found in feces sampled 6 h after nasal administration (Fig. 2D). No evidence for the deposition of [¹¹¹In]-BoHc/A in the CNS was found. The accumulated BoHc/A in the cerebrum and OB were <0.01 and <0.1 SUV for the entire examination period from 10 min to 48 h after nasal administration of 1 × 10⁶ cpm per 18 μl (9 μl, each nostril) [¹¹¹In]-BoHc/A per mouse, respectively (Fig. 2E). On the contrary, [¹¹¹In]-cholera toxin was detected in the OB from 6 h after nasal administration of 1 × 10⁶ cpm per 20 μl (10 μl, each nostril) [¹¹¹In]-cholera toxin per mouse (Fig. 2F). This finding is in complete agreement with the data obtained with [¹⁸F]-BoHc/A (Fig. 2B). Radioactive material in samples extracted from the nasal mucosa 10 min to 24 h after administration and subjected to size-exclusion chromatography analysis by using PD-10 columns was found mostly in the high m.w. fractions at 6 h and then gradually appeared in the low m.w. fractions (Fig. 3B). This finding suggests that intact BoHc/A molecules were attached to the nasal mucosa for at least the first 6 h after administration. Radioactivity was found in the low m.w. fractions of urine and feces samples at 6 h after nasal administration, suggesting that the vaccine Ags were degraded.

Nasal immunization of mice, but not macaques, with BoHc/A can induce protective immunity against BoNT/A (12). To investigate whether the recombinant BoHc/A nasal vaccine induces protective immunity, we nasally immunized three cynomolgus macaques with 500 μg BoHc/A, and we administered to one cynomolgus macaque PBS as a control (Fig. 4A). Ag-specific Ab responses were assessed by using BoNT/A as a coating Ag for ELISAs. After nasal immunization, the BoNT/A-specific serum IgG Ab titers increased in all BoHc/A-vaccinated macaques, but not in the control macaque administered PBS. In addition, the toxin-specific secretory IgA Ab titers in nasal washes increased in all the nasally immunized macaques, but not in the control macaque (Fig. 4B). These results indicated that BoHc/A is a highly immunogenic nasal vaccine capable of inducing toxin-specific Abs not only in systemic compartments but in mucosal compartments in cynomolgus macaques.

Next, a challenge test with 25 μg (2.7 × 10⁶ LD₅₀) BoNT/A per kilogram body weight was performed on all three BoHc/A-vaccinated macaques and on the one control macaque; the dosage required was estimated from the neutralizing activity in serum from the immunized macaques (Supplemental Fig. 4). The BoHc/A-vaccinated macaques, but not the control, were completely protected against the high lethal dose of i.p. injected BoNT/A (Fig. 4C). There were no clinical signs of toxin-associated disease over the 30-d observation period after the BoNT/A challenge test. The control macaque was euthanized 4 h after the challenge because of the development of several clinical signs of botulism. After 30 d, all BoHc/A-vaccinated macaques that had undergone the challenge test were sacrificed. Detailed examinations revealed the absence of any clinical signs of organ abnormality or botulism. Taken together, these results demonstrate that the nasal BoHc/A vaccine is a safe and effective mucosal vaccine against C. botulinum neurotoxin.

Because we confirmed that the nasal immunization of macaques with BoHc/A induced protective immunity, we then investigated whether the nasal BoHc/A vaccine is transported to the CNS via the OB in nonhuman primates using two naive rhesus macaques on the different experiment. To ensure that the [¹⁸F]-BoHc/A completely penetrated the nasal cavity and the olfactory epithelium, we laid two macaques (No. 5 and 6) on their backs for 10 min after the nasal administration of 34.5 and 54.7 MBq [¹⁸F]-BoHc/A in 315 and 500 μl PBS, respectively. After anesthesia, the macaque’s head was placed into the PET scanner system and real-time imaging was performed for 6 h. To confirm the exact position of the cerebrum, we performed an MRI scan and then superimposed the PET images onto the MRI images. Real-time PET images of macaque No. 6, from 15 min to 6 h after nasal administration, are shown in Fig. 5A. No CNS deposition of vaccine Ags in cerebrum was detected within 6 h after nasal administration of [¹⁸F]-BoHc/A in either macaque. To confirm the absence of [¹⁸F]-BoHc/A in the cerebrum of the two macaques, we directly counted radioactivity in each tissue, including cerebrum and OB, at 6 h after immunization (Fig. 5B, 5C). We found that the accumulated [¹⁸F]-BoHc/A in the cerebrum and OB were <0.01 and <0.04 SUV, respectively. Taken together, these results demonstrated that there was no CNS deposition of nasally administered BoHc/A vaccine in the rhesus macaques. Six hours after nasal administration of [¹⁸F]-BoHc/A, radioactivity was detected in the urine, stomach, kidney, small intestine, and nasal mucosa (Fig. 5B). Among these samples, the highest level of radioactivity was found in the urine. The radioactivity in urine was found in the low m.w. fractions after size-exclusion chromatography analysis by using a PD-10 column (Fig. 3A), suggesting that degradation of [¹⁸F]-BoHc/A had occurred.

Radio- or luminescence-labeling techniques have the capacity to track vaccine Ags and adjuvants (3, 23). However, these classical methods are generally end-point approaches because they require the use of multiple animals for diachronic analysis. In this paper, we showed that quantitative real-time analysis of the tissue distribution and degradation of nasal vaccines could be achieved in both rodents and nonhuman primates. To evaluate the safety of the nasal vaccine candidate against C. botulinum neurotoxin, BoHc/A, we developed an in vivo molecular imaging system for ADME by using [¹⁸F]-BoHc/A and PET together with CT or MRI and the use of PPIS. Our data from in vivo molecular imaging with [¹⁸F]-BoHc/A was consistent with the results obtained by classical direct counting to measure the radioactivity in various tissues in mice and nonhuman primates. Thus, our in vivo molecular imaging system is a potentially important strategy for the precise analysis of nasal vaccines and adjuvants for preclinical safety testing in individual animals.

After the nasal administration of [¹⁸F]-BoHc/A to mice, a major portion of [¹⁸F]-BoHc/A immediately adhered to the nasal mucosa and, simultaneously, some Ag was cleared from the nasal cavity by the mucociliary clearance mechanism (4) and was transported to the throat, esophagus, gastrointestinal tract (including the stomach and intestine), and urinary bladder in order of decreasing abundance (Fig. 1A, Supplemental Videos 1 and 2, and Supplemental Fig. 3). We used our PET imaging method to identify the sites of accumulated [¹⁸F]-BoHc/A in real time after nasal administration and to quantify the amounts of accumulated [¹⁸F]-BoHc/A (Fig. 1B). We confirmed the accuracy of this approach by comparing these data with those derived by the classical direct counting assay on various dissected tissues (Fig. 2A). Because radioactive material with a low m.w. was found in the stomach and intestine, as well as in the feces, we suggest that the portion of nasal [¹⁸F]-BoHc/A or [¹¹¹In]-BoHc/A vaccine that reached the digestive tract was degraded and then absorbed in the intestines (Fig. 3A, 3B). Radioactive material with a low m.w. was also found in the urine. Because [¹⁸F]-BoHc/A and [¹¹¹In]-BoHc/A were synthesized by labeling N-terminal and ε–Lys amino groups on BoHc/A, nasally administered [¹⁸F]-BoHc/A or [¹¹¹In]-BoHc/A might be degraded in the gastrointestinal tract and converted to [¹⁸F]-ε –Lys or [¹¹¹In]-ε –Lys, which undergo fast blood clearance via the kidneys. Many [¹¹¹In]-glycoproteins administered to rodents are degraded and converted to [¹¹¹In]-ε –Lys, which is then recovered in urine and feces (24). Most of the low m.w. compounds labeled with [¹¹¹In] were detected in feces at 6 h after nasal administration (Fig. 2D), whereas most of the low m.w. compounds labeled with [¹⁸F] were present in urine (Fig. 2A). This finding suggests that the [¹⁸F]-compounds are absorbed from the gastrointestinal tract and cleared by the kidneys into the urine at a faster rate than is the case for the [¹¹¹In] compounds. In addition, we found no indication of accumulated radioactivity in bones in mice nasally administered [¹⁸F]-BoHc/A (Fig. 2A). Because [¹⁸F] fluoride accumulates in bone (25), our observation suggests that degradation of [¹⁸F]-BoHc/A to [¹⁸F] fluoride did not occur.

The radioactive material detected in the nasal mucosa that was derived from [¹⁸F]-BoHc/A or [¹¹¹In]-BoHc/A was high m.w., suggesting the maintenance of the intact nasal vaccine for as long as 6 h after nasal administration (Fig. 3A, 3B). Although BoHc/A degraded gradually on the surface of the nasal epithelium and/or within the nasal cavity beginning 6 h after administration (Fig. 3B) by contact with proteolytic enzymes (26, 27), intact BoHc/A can bind and penetrate the mucosal epithelial cells by a transcytosis mechanism mediated by an as yet unknown receptor (28). Therefore, we propose that some of the BoHc/A taken up by and released from the nasal epithelium might stimulate the airway immune system and lead to the induction of Ag-specific immune responses. In fact, we demonstrated the induction of toxin-specific systemic IgG and mucosal IgA Ab responses with strong Ag neutralizing activity when macaques were nasally immunized with 500 μg BoHc/A (Fig. 4B). In addition, we performed a toxin challenge test to provide proof of efficacy of the nasal BoHc/A vaccine in nonhuman primates (Fig. 4C). All the macaques that received the nasal BoHc/A vaccine were protected from an i.p. challenge with a dose of BoNT/A (25 μg [2.7 × 10⁶ LD₅₀]/kg), which is ∼35,000 times the lethal dose of ∼0.7 ng/kg (29).

Recent nasal vaccine studies have raised concerns about the deposition and accumulation of candidate vaccine Ags in the CNS through direct transport from nasal mucosa to the cerebrum via the OB (1, 3). For this reason, we investigated the deposition and accumulation of the nasal BoHc/A vaccine by direct tissue counting analyses of [¹⁸F]-BoHc/A or [¹¹¹In]-BoHc/A in mice. We showed that there was no OB deposition or accumulation of BoHc/A in the CNS within 6 h (≤0.04 SUV by [¹⁸F] counts) or 48 h (≤0.1 SUV by [¹¹¹In] counts) after nasal administration (Fig. 2B, 2D). These results demonstrate that the nasal BoHc/A vaccine did not reach the mouse CNS, despite the large surface area of the nasal cavity containing olfactory epithelium that is connected to the CNS (4), and thereby confirm the safety of the vaccine. On the basis of these promising findings for the efficacy of the nasal BoHc/A vaccine in mice, we used rhesus macaques as a nonhuman primate model to perform the preliminary experiments necessary to validate the approach for future human trials. Using the [¹⁸F]-BoHc/A–PET in vivo molecular imaging method, we showed that there was no [¹⁸F]-BoHc/A deposited in the cerebrum of rhesus macaques 6 h after nasal administration of 57.4 MBq [¹⁸F]-BoHc/A (Fig. 5A). Furthermore, 6 h after nasal administration of [¹⁸F]-BoHc/A, no [¹⁸F] radioactivity was detected in the cerebrum or OB of the macaque by direct tissue counting (Fig. 5C). Thus, we showed that there is no deposition of the Ag in the CNS in nonhuman primates. This result confirms our results in mice and shows that nasal delivery of BoHc/A is safe with respect to its ADME profile. In addition, the radioactivity in various tissues of the macaques after nasal administration of [¹⁸F]-BoHc/A were in complete agreement with the data obtained in mice (Figs. 2A, 5B). Degraded radioactive material with a low m.w. was also found in the urine of macaques (Fig. 3A). Taken together, our findings suggest that after nasal administration of [¹⁸F]-BoHc/A to macaques, the [¹⁸F]-compound is absorbed from the gastrointestinal tract and cleared by the kidneys into the urine.

In summary, we developed an in vivo molecular imaging system that combines PET with CT or MRI and is suitable as an ADME profiling system for biotechnology-derived pharmaceuticals, including nasal vaccines. Using this system, we demonstrated the safety and efficiency of a nasal vaccine candidate against botulism in mice and nonhuman primates. Our in vivo molecular imaging system can be used to replace the current whole-body autoradiography method for the preclinical ADME evaluation and may be suitable for use in human clinical studies.

Disclosures The authors have no financial conflicts of interest.