All animals were kept at the experimental animal facility at IDT Biologika in accordance with the prevailing guidelines and the studies were performed according to European guidelines on animal welfare, clinical endpoints, and care of the Federation of European Laboratory Animal Science Associations (FELASA). The required permits for the animal studies were obtained from the appropriate veterinary authorities in the federal state of Saxony Anhalt, Germany (Landesverwaltungsamt Sachsen – Anhalt, Referat Verbraucherschutz, Veterinärangelegenheiten) (See Table S1 for approval numbers).

The vaccine construct SPBN GASGAS is derived from SAD L16, a cDNA clone of the oral rabies lyssavirus (RABV) vaccine strain SAD B19 (4). SPBN GASGAS lacks the pseudogene (ᴪ) and has been genetically modified by site-directed mutagenesis. To abolish residual pathogenicity and reduce the risk of the emergence of a less attenuated revertant by back mutation, mutations have been incorporated at multiple sites; amino acid positions 194 (AAT [Asn] → TCC [Ser]) and 333 (AGA [Arg] → GAG [Glu]) of the glycoprotein (5). Furthermore, the construct contains a second identical glycoprotein gene with modifications as described (6, 7). The vaccine virus was prepared according to Vos et al (8). Material for the overdose studies was concentrated via tangential flow filtration using ultrafiltration flat sheet cassettes with a Molecular Weight Cut Off (MWCO) of 300 kDa.

For detection of rabies virus in the brain, samples are tested by the Fluorescent Antibody Test (FAT) (9). To detect replication competent vaccine virus in tissues samples other than brain material the Rabies Tissue Culture Infection Test (RTCIT) was used (10). As this test involves 3 serial passages in cell culture until a sample is considered negative, samples were first screened by real-time RT-PCR for the presence of viral RNA as described by Hofmann et al (11). If positive in real-time RT-PCR, the sample was subsequently tested for infectious virus by RTCIT. To detect and quantify antibodies against rabies virus in vaccinated and contact animals, three different assays were used. For the detection of VNA, the Rapid Fluorescence Focus Infection Test (RFFIT) (12) with the modifications as described by Cox & Schneider (13) and the Fluorescent Antibody Viral Neutralization (FAVN) (14) were used. For the detection of binding antibodies against rabies virus a blocking ELISA was used according to the instructions of the manufacturer (BioPro, Rabies ELISA, BioPro, Prague,Czech Republic) (15). Primarily, the RFFIT was used but if the obtained results were inconclusive due to for example cell toxicity or unspecific reactions, a retention sample was tested by FAVN and/or ELISA. Initially, the FAVN-assay was used for these retention samples, but later replaced by the ELISA as this assay was better suited for qualifying samples unambiguous seropositive or seronegative (15).

General information on the different animal studies can be found in Table 1 and Table S1. The vaccine was administered by direct oral instillation, except for the study in pregnant vixens. Blood samples were collected and examined for the presence of virus neutralizing antibodies (VNA) prior to vaccine administration, at the end of the study and sometimes on different occasions during the observation period of all animals, except rodents (Table 1). In case animals were sedated for vaccine administration or sampling, a mixture of Xylazine and Ketamine was administered i.m.; for the domestic pigs, methadone was used instead of ketamine. The dosage administered was adjusted to body weight. Euthanasia was performed by an intracardial injection of Release® (300 mg/ml Pentobarbital-natrium) per sedated animal.

For the overdose studies in the target and non-target species, as required at least 10x the maximum dose most likely to be contained in a vaccine bait was administered and the animals were observed for at least 180 days (1). At the end of the study, brain tissue (pooled samples of the medulla oblongata, cerebrum [near location of cornu ammonis], cerebellum) of all animals were tested for the presence of vaccine virus using the FAT. Additionally, the gonads of the male raccoon dogs were tested for the presence of vaccine virus RNA by using real time RT-PCR.

Twelve animals of each target species received 1.7 ml SPBN GASGAS (10^7.5 FFU/ml) by direct oral instillation. Furthermore, 2 naïve contact animals of each target species were housed with one or more treated animals. On 1, 2, 3, 5, 7 and 10 days post vaccine administration 2 foxes or 2 raccoon dogs were euthanized and the following tissue and organ samples were taken and examined for the presence of the vaccine virus; brain (Cornu ammonis, Medulla oblongata, Cerebellum, and Cerebrum), Lnn. mandibularis and – retropharyngeales [only foxes], upper and lower mucosa of the oral cavity, tongue, velum palatinum, tonsils (Tonsila sublingualis, T. lingualis, T. palatine, T. pharyngica), lung, kidney, bladder, heart, colon, ileum, and jejunum. The two contact foxes and raccoon dogs were sacrificed at day 84 and 85, respectively. Twelve dogs received 3.5 ml SPBN GASGAS (10^7.9 FFU/ml) and the same sampling scheme as for the target species was used. For the domestic pigs, a different sampling schedule was applied; on day 1, 2, 4, 8 and 56 post vaccine administration two animals were euthanized. Besides brain samples the following organs/tissues were sampled; tongue, lung, kidney, liver, heart, bladder, jejunal lymph nodes, colon lymph nodes, colon, jejunum, ileum, Tonsilla pharyngea, Lnn. mediastinales, and Lnn. tracheobronchales. Furthermore, several samples were taken from tissues that are used for meat products; venter fat tissue, venter muscle tissue, femoral fat tissue, femoral muscle tissue, Crus fat tissue, and Crus muscle tissue. Also, on day 56 both contact pigs were sacrificed, and besides a blood sample also brain tissue was collected and examined for the presence of vaccine virus.

Twelve juvenile foxes and raccoon dogs were housed in 4 groups of 3 animals each and observed for 28 and 31 days, respectively. At day 0, 3 and 7, two animals in each group received 1.0 ml SPBN GASGAS (10^7.5 FFU/ml) by direct oral instillation. The remaining third animal in each group did not receive the vaccine construct and was kept as a naïve contact animal to determine possible horizontal transmission.

Three, pregnant foxes were offered a vaccine bait containing a blister filled with 1.7 ml SPBN GASGAS (10^7.2 FFU/ml) 8–12 days after mating. Two vixens refused the baits and were subsequently offered the vaccine by placing it in a drinking bowl. One vixen drank 14 ml and the other animal 17 ml of the vaccine construct (10^7.2 FFU/ml). As blood sampling would require anesthesia and these activities could lead to abortion, no blood sample was taken prior to vaccine administration. However, all three animals had not been vaccinated against rabies prior to study begin. The vixens were housed in individual cages. To examine if the vaccine virus was transferred from the vaccinated vixens to their offspring, blood and saliva were collected from the cubs in week 2, 4, 7 and 12 post parturition. The vixens and cubs were observed for 83 and 87 days post parturition, respectively. At the end of the observation period all animals were euthanized and subsequently brain and salivary glands were examined for the presence of the vaccine virus.

Both common vole (Microtus arvalis) and house mice (Mus musculus) received 0.03 ml SPBN GASGAS (10 ^9.1 FFU/ml) by direct oral instillation and were placed in cages together with at least one naïve contact animal. As only male guinea pigs (Cavia porcellus) were available, the animals could not be housed together and were kept individually; hence no direct contact between treated (0.5 ml SPBN GASGAS [10 ^9.1 FFU/ml], d.o.a.) and control animals was possible. However, the animals were kept in the same room and the nest material from the vaccinated animals was mixed every day with the nest material in the cages of the control animals. For the application of the vaccine the unsedated animals were fixed at the neck and kept nearly in vertical position; this straightened the oesophagus, allowing an easier passage of the substance. The fixation did not affect the swallowing reflex of the rodents. After fixation the pipette dip was placed 2–3 mm deep into the mouth and the substance was administered slowly. The naïve contact house and field mice were ear-punched for identification purposes.

All animals tested sero-negative prior to vaccine administration. After administration all foxes and raccoon dogs developed an immune response (>0.5 IU/ml), indicating successful vaccine uptake. One dog and one cat did not develop a detectable immune response in any of the samples collected post vaccine administration. Also, the 4 naïve contact cats and dogs remained sero-negative during the observation period, indicating no horizontal transmission of the vaccine virus. During the observation period, it was necessary to exchange dogs between cages to avoid injuries as a result of aggressive behaviour. The contact animals though were always kept in a cage with at least one treated animal. All vaccinated and contact animals survived the 6 months observation period, except for two vaccinated raccoon dogs that were found dead on day 107 and 108 post vaccine administration. One animal died of bacterial pneumonia with formation of lung abscesses and the other showed leucocytosis, inflammation-related lung changes and bacteraemia. Interdigital abscesses were suspected to be the cause of a bacterial sepsis which resulted in the sudden death of the animals. Furthermore, several foxes and raccoon dogs showed symptoms of impaired health for short periods during the observation period. These symptoms were unrelated to rabies and all animals recovered. Furthermore, no rabies virus was detectable in the brains of these and all other animals at the end of the study (FAT). Also, no viral RNA was detected in the gonads of the raccoon dogs at the end of the study.

None of the foxes and raccoon dog had rabies virus antibodies prior to vaccine administration. All but one fox remained healthy during the 28 day observation period: One fox was euthanized at day 23 post first vaccine administration due to an injured front leg. No RABV antigen was detected in the brains of the vaccinated and contact animals (FAT). Unfortunately, one contact fox accidentally received a vaccine dose on the day that the 3rd dose was administered; thus, the animal cannot be considered a naïve contact animal. All animals that received the vaccine had seroconverted at the end of the study, including the contact animal that was given mistakenly a single dose on day 7. Also, 2 of 3 contact foxes seroconverted (2.67 and 0.64 IU/ml) and one contact raccoon dog had VNA levels > 0.5 IU/ml (0.86 IU/ml), indicating contact with the vaccine.

All foxes, raccoon dogs, dogs and pigs tested sero-negative prior to vaccine administration. Viral RNA was detected in the palatine tonsils in all foxes that received the vaccine construct (PCR), except for one fox euthanized on day 1. Also, in 9 of the 12 palatine tonsils collected from the raccoon dogs, viral RNA was detected: 3 palatine tonsils collected from animals euthanized at day 2, 5 and 7 post vaccine administration tested negative. However, viable virus was re-isolated from the PCR-positive palatine tonsils only during the first 3 days. From day 5 onwards, these samples tested negative in RTCIT indicating that no viable virus was present anymore. In dogs, viral RNA was detected in 6 palatine tonsils, but only in a sample taken from a dog on day 3 post vaccine administration viable virus was isolated (Table 2). In another dog, sampled on day 5 post vaccine administration, besides the palatine tonsil also the retropharyngeal lymph nodes and pharyngeal tonsils tested PCR-positive and subsequently RTCIT-negative.

All other tissues and organ samples from foxes tested negative for RABV RNA, irrespective of the day of sampling. In raccoon dogs, 3 additional samples tested positive for viral RNA; both velum palatinum samples and a lung tissue sample collected on day 3 and 7, respectively. Only, in one velum palatinum sample infectious virus was re-isolated. In the domestic pigs, all samples tested PCR-negative although no sample of the Tonsilla pharyngea was collected from the animals sacrificed on day 1 and 2 post vaccine administration. No virus antigen was detected in the brain samples collected from all vaccinated and contact animals (FAT).

Only the 4 foxes sampled on day 7 and 10 post vaccine administration seroconverted (>0.5 IU/ml), the foxes euthanized earlier did not yet develop a detectable immune response. A similar situation was observed in the raccoon dogs. Both raccoon dogs euthanized at day 10 post vaccine administration seroconverted; 1.77 and 1.57 IU/ml. All blood samples collected from the other vaccinated animals were below the threshold of 0.5 IU/ml, although one sample taken from a raccoon dog on day 7 post vaccine administration was very close to this threshold (0.49 IU/ml). The result of the ELISA confirmed this by giving a positive signal (>40% inhibition). Only one dog euthanized 10 days post vaccine administration seroconverted (8.18 IU/ml). Both pigs euthanized 56 days post vaccine administration seroconverted. All contact animals from the different species did not develop an immune response. Also, no viral RNA was detected in the samples collected from these contact animals, indicating that horizontal transmission of the vaccine virus did occur during these studies.

The 3 vixens gave birth to 14 cubs on day 39 and 43 post vaccine administration; two litters of 5 cubs and one litter with 4 cubs. The cubs showed evidence of decreasing low levels of maternally derived antibodies at day 15 (<0.4 IU/ml) and 87 after birth (<0.1 IU/ml) (Table 3). The level of antibodies in the vixens was at all sampling points > 0.5 IU/ml. In none of the cubs vaccine virus was found in saliva, salivary glands or brain. No adverse or critical event (such as abortions) referring to general health was recorded in the vixens and their cubs.

In house mice and guinea pigs, no adverse reaction was observed in any of the animals during the entire observation period and subsequently no virus antigen was detected in the brain of the control animals (FAT). Seven common voles died during the observation period, but none of these animals showed any clinical signs of rabies as confirmed by negative FAT; 3 treated and 4 contact animals. Also, no vaccine virus was detected in the brains of the remaining contact voles at the end of the study (FAT).