ChAdOx1 nCoV-19 protection against SARS-CoV-2 in rhesus macaque and ferret challenge models

Vaccines against SARS-CoV-2 are urgently required, but early development of vaccines against SARS-CoV-1 resulted in enhanced disease after vaccination. Careful assessment of this phenomena is warranted for vaccine development against SARS CoV-2. Here we report detailed immune profiling after ChAdOx1 nCoV-19 (AZD1222) and subsequent high dose challenge in two animal models of SARS-CoV-2 mediated disease. We demonstrate in rhesus macaques the lung pathology caused by SARS-CoV-2 mediated pneumonia is reduced by prior vaccination with ChAdOx1 nCoV-19 which induced neutralising antibody responses after a single intramuscular administration. In a second animal model, ferrets, ChAdOx1 nCoV-19 reduced both virus shedding and lung pathology. Antibody titre were boosted by a second dose. Data from these challenge models on the absence of enhanced disease and the detailed immune profiling, support the continued clinical evaluation of ChAdOx1 nCoV-19.


Introduction
In response to the COVID-19 pandemic, multiple candidate vaccines have entered preclinical and clinical development, and clinical e cacy has now been demonstrated for two vaccines (1,2). Inactivated (3,4), adenoviral-vectored (5, 6) RNA (7) and DNA vaccines (8,9) have demonstrated protection against SARS-CoV-2 challenge in rhesus macaques, and SARS-CoV-2 infection has been shown to protect against rechallenge in this species (10,11). Non-human primate challenge studies following vaccination with candidate vaccines are principally used to assess vaccine safety, ruling out evidence of vaccine enhanced disease after vaccination and challenge. Here we report on studies in two different animal models following vaccination with ChAdOx1 nCoV-19 (AZD1222) and subsequent SARS-CoV-2 challenge.
In rhesus macaques we show, using computerised tomography (CT) scanning, that the changes in the lungs induced after SARS-CoV-2 challenge are similar to the changes in human lung tissue during COVID-19, and can be prevented by intramuscular vaccination with ChAdOx1 nCoV- 19. Importantly it has been demonstrated in the ferret model system that vaccine enhanced disease can be transiently induced by a formalin inactivated alum-adjuvanted SARS-CoV-2 vaccine (FIV), resulting in increased lung pathology after vaccination and challenge when compared to unvaccinated animals (12). Here we demonstrate that in the same model, lung pathology is reduced following ChAdOx1 nCoV-19 vaccination and challenge when compared to animals vaccinated with vaccine expressing an irrelevant antigen (green uorescent protein; GFP). Importantly, this work demonstrates both safety and a reduction in disease pathology in two animal models using a new methodology to detect disease. Further evidence of a Th1 bias after vaccination with ChAdOx1 nCoV-19 is demonstrated and the immune response post vaccination with ChAdOx1 nCoV-19 and after challenge with SAS CoV-2 is described.

Results
Immune response to ChAdOx1 nCoV-19 vaccination in rhesus macaques and ferrets. ChAdOx1 nCoV-19 is a replication-de cient simian adenoviral vector expressing a codon-optimised fulllength SARS-CoV-2 spike protein that has been shown to prevent SARS-CoV-2 pneumonia in rhesus macaques at a dose of 2.5 x 10 10 vp (6) and is immunogenic with an acceptable safety pro le in humans at a dose of 5 x 10 10 vp (13). Here, six adult rhesus macaques (three male, three female) were vaccinated with a single dose of 2.5 x 10 10 vp, with an equivalent control group receiving saline injections. Humoral immunogenicity was assessed at 14 and 27 days after vaccination by ELISA, with IgG ( Fig. 1A), IgM and IgA (Fig. S1A) spike-speci c antibodies induced in all the vaccinated animals. Neutralising antibodies were assessed in a PRNT 50 assay, determining the antibody titre required for a 50% reduction in viral plaque formation in susceptible cells. All six vaccinated animals produced neutralising antibodies with a median titre of 74.5 (sd 76.6) at day 14 and 95 (sd 131) at day 27. (Fig. 1A). Neutralising antibodies were also assessed in a pseudo neutralization assay (Fig. 1A), with a correlation (r 2 =0.4032, p=0.0265) between the two neutralisation assays (Fig. S1A right).
In all ferrets immunised with a single dose of 2.5 x 10 10 vp ChAdOx1 nCoV-19, spike-speci c IgG (Fig. 1B), IgM and IgA (Fig. S1B) antibodies were increased relative to animals vaccinated with control ChAdOx1 GFP vector, expressing green uorescent protein as the vaccine antigen. PRNT 50 titres reached 1118 (sd 478) at day 14 and 1708 (sd 809) at day 28 (median of 11 animals) (Fig. 1B). In six ferrets receiving a second vaccine dose at day 28, titres increased from 1379 (sd 699) at day 28 to 3867 (sd 1645) at day 35 (median of 6 animals). These titres were not signi cantly higher than a single dose vaccination (Fig. 1B).
T cell responses to SARS-CoV-2 spike were also assessed by interferon-gamma ELISpot in both rhesus macaques and ferrets. A signi cant increase in the total spike-speci c T cell response was observed at day 14 in rhesus macaques ( Fig. 2A). Across the peptide pools spanning the spike protein T cell responses were measured against all regions ( Fig. 2A); however responses were typically higher to S1 peptide pools when compared to S2, ( Fig. 2A). Measurement of cytokine production in the supernatant of PMBCs stimulated with spike peptides spanning the dominant S1 region showed a log increase in IL2 levels in vaccinated animals when compared to PBS control animals (Fig. S2A). In addition, an increase in IFNg was measured. No change in IL1b, IL8, IL6 or IL10 was measured between ChAdOx1 nCoV-19 vaccinated animals and PBS control animals. Simultaneous measurement of IFNg, IL5 or IL13 by FLUROSpot assay on day 27 rhesus macaque PBMCs, stimulated with spike peptides, demonstrated Th1 bias of the antigen speci c response with a higher number of antigen speci c IFNg producing cells observed, compared to cells producing either IL5 or IL13. (Fig. S2B).
A statistically signi cant (p=0.001) increase in spike speci c T cells was observed in ferrets from day 14 onwards when compared to the day of vaccination (Fig. 2B), and a small but non-signi cant increase in IFNg producing T cells was observed after boosting. Mapping of T cell responses across peptides spanning the spike protein, showed responses to all peptide pools, which were predominantly directed against the S1 peptide pools (pool 1 and pool 2) (Fig. 2B), with proportional responses to individual pools in each animal not changing over time (Fig. S3A). At day 28 post-vaccination, antigen speci c IFNg + CD8 + T cell responses were detected by ow cytometry (Fig. 2B). No statistically signi cant difference in IFNg + , TNFa + or IL4 + CD4 + T cells was observed between ChAdOx1 nCoV-19 and ChAdOx1-GFP vaccinated animals a on day 28 ( Fig. 2B) (data not shown). Comparison of IFNg detected by ELISpot or ICS demonstrated that CD8 + T cells were the predominant population producing IFNg at day 28 postvaccination (Fig. 2B).
CT assessment of SARS-CoV-2 disease in vaccinated and non-vaccinated rhesus macaques after SARS-CoV-2 challenge.
Twenty seven days after vaccination all twelve rhesus macaques were challenged with a total of 5 x 10 6 pfu SARS-CoV-2 administered via both intratracheal and intranasal routes. CT scans were performed on all animals 15 days prior to challenge and on day 5 after challenge. Two animals per group were euthanized at day 7 for necropsy, and CT scans were performed again on day 12 in the remaining animals. Representative examples of the CT scans are shown in (Fig. 3A and S4A.). A scoring system (Table S1) was used to quantitate disease pattern and distribution ( Fig. S4B and Table S2) (Supplementary Methods) which was combined to produce a total score (Fig. 3A) (14). The CT scans con rm that ve days after direct instillation of SARS-CoV-2 into the trachea and nose, lung tissue became infected resulting in pathological ndings similar to mild clinical cases of human COVID-19 (15). These changes were seen in four out of six saline control and two out of six ChAdOx1 nCoV-19 vaccinated macaques, with reduced disease scores in the two vaccinated animals who demonstrated pulmonary changes (Fig. 3A).
As in human disease, higher incidence of abnormalities in the lungs was observed in males than in females. Some abnormalities were detected in all male animals on either day 5 or day 12 post challenge, but only 50% of females (Table S1). Where abnormalities were reported they were at low levels with less than 25% of the lung involved, indicating that rhesus macaques experience mild disease in this challenge model similar to mild clinical cases of human disease (Fig. S4C). There were fewer abnormal ndings in vaccinated than control animals at day 5 post challenge, with equivalent amounts at day 12 (Fig. S4C). Systemic monitoring of animal health showed signi cantly more weight loss in control animals compared to ChAdOx1 nCoV-19 vaccinated animals (p=0.0108) (Fig. S4D). Over the entire post-challenge period, no statistically signi cant difference in body temperature between groups was observed, however at day 1 post challenge control animals had a higher median body temperature (39.4, sd 0.423) compared to vaccinees (median 38.8, sd 0.480) (Fig. S4E). Similarly, a small increase in body temperature in controls animals (median 39.050, sd 0.204) compared to ChAdOx1 nCoV-19 vaccinated animals (38.45, sd 0.383) was also observed at day 3 post challenge.
Detection of viral RNA and histopathology following challenge of rhesus macaques Bronchoalveolar lavage was performed at necropsy in two animals per group on days 7, 13 and 14 postchallenge. Viral RNA was only detected in bronchoalveolar lavage uid (BALF) from control animals (Fig.  3B). Viral RNA was also quantitated in nasal wash samples and throat swabs with similar results in both groups (Fig. 3B).
Viral RNA was detected in staining of lung tissue sections in both control animals on day 7 postchallenge and three out of four control animals on day 13/14, but only one vaccinated animal, on day 7 (Fig. 3C). Lesions consistent with infection with SARS-CoV-2 were observed in the lungs of animals from both the control and vaccinated groups (Fig. 3D), with a considerably greater severity in one of the control animals. These lesions included diffuse alveolar damage, alveolar hyperplasia, perivascular and peribronchiolar lymphoid in ltrates and bronchial/bronchiolar necrosis and exudates (Fig. 3D). No signi cant changes were observed in any other tissues examined. At 13/14 days post-challenge, multifocal areas of lung pathology, as described at 7 days post-challenge, together with signs of lesion resolution, were noted at reduced severity in three out of the four control animals; in the remaining animal, lesion severity had not reduced. Minimal lesions were also noted in three out of four vaccinated animals; however, in one animal, mild, multifocal interstitial pneumonia and perivascular cu ng was observed.
Detection of virus and histopathology following challenge of ferrets Ferrets were challenged with 5 x 10 6 pfu SARS-CoV-2 administered intranasally 28 days after the last vaccination, and the duration of challenge was 14 days. Challenges were staggered and took place for these groups initially (ChAdOx1 nCoV-19 and ChAdOx1 GFP prime only) followed by two further groups (ChAdOx1 nCoV-19 and ChAdOx1 GFP prime boost). Viral RNA was detected in all groups in nasal wash samples two days after challenge, with reductions in the ChAdOx1 nCoV-19 vaccinated groups by day 4 and all vaccinated animals except for one were negative at day 6 ( Fig. 4A). Viral RNA in nasal washes over the total challenge period tended to be lower in the prime boost group than prime only (Fig. 4A). In contrast, in the ChAdOx1 GFP control groups (after one or two doses of ChAdOx1 GFP) the viral RNA in the nasal washes remained above the limit of quanti cation until day 6, (Fig. 4A). Minimal viral RNA was detected in throat swabs or BALF samples in any of the groups (Fig S5A), with no virus above baseline detected in the lung of any animal (data not shown).
Histopathology was performed on two animals per group at day 6/7 post infection and the remainder at days 13/14, with scores summarized in Fig. 4B and detailed ndings included in the supplementary information. Animals vaccinated with one dose of ChAdOx1 nCoV-19 did not show any remarkable change in the lungs. On day 6/7 days post-infection in the group vaccinated with only one dose of ChAdOx1 GFP, one animal had mild lesions compatible with subacute bronchopneumonia and the other had occasional minimal bronchiolar in ltrates. In ferrets receiving one dose of ChAdOx1 GFP, histopathological changes at 13/14 days post-infection were reduced compared to 6 days post-infection. Minimal changes were observed in the lungs of animals receiving 2 doses of ChAdOx1 nCoV-19 at either 6 or 13/14 days post-infection, and a signi cant increase was measured in the histology score (minimal to mild changes) in the group receiving 2 doses of ChAdOx1 GFP when compared to the two dose ChAdOx1 nCoV-19 group.

Anamnestic responses following challenge
Responses were assessed in a virus IC 50 neutralisation assay showing an increased neutralisation titre from day 3 post challenge to day 7 or day 13/14 in vaccinated non-human primates, and days 7 to 13/14 in PBS controls (Fig. 5A left). Antigen-speci c cellular immune responses were measured in PBMCs stimulated with overlapping 15-mer SARS-CoV-2 spike protein peptide pools, using an ex vivo IFN-g ELISpot assay and also showed an increase in antigen speci c responses post-challenge in both ChAdOx1 nCoV-19 vaccinated and PBS control rhesus macaques after day 3 (Fig. 5A right). The median response in ChAdOx1 nCoV-19 vaccinated animals was higher than that measured in the PBS control animals on day 7 and day 13/14 post-challenge (Fig. 5A right).
Immunophenotyping ow cytometry assays were applied to whole blood samples collected immediately prior to ( In ferrets, neutralising antibodies did not dramatically increase in ChAdOx1 nCoV-19 vaccinated animals following SARS CoV-2 challenge, with both groups having similar titres of neutralising antibodies by the end of the study (Fig. 5D). In ChAdOx1 nCoV-19 vaccinated animals, the IgG antibody titres on the day of challenge, measured by ELISA, neutralisation or pseudo neutralisation assay, inversely correlated with the peak level of viraemia measured in each animal (Fig. 5E). There was no relationship between IFNg ELISpot and peak viraemia. Overall the data would suggest that vaccine induced protection in ferrets was not associated with the level of spike speci c T cells measured in these assays, but was associated with the humoral response to SARS-CoV-2 spike protein.

Discussion
A safe and effective vaccine is expected to be an essential requirement to effectively control the COVID-19 pandemic. Early development of a vaccine against Feline Infectious Peritonitis, which is also caused by a coronavirus, resulted in enhanced disease in vaccinated and then challenged animals (16), a phenomenon also seen in early development of vaccines against SARS-CoV-1 (17,18). Vaccine enhanced disease results in an increase in disease severity when vaccinated subjects are subsequently exposed or challenged with natural virus. Immunopathology in coronavirus vaccinated and challenged animals has been associated with increased levels of the Th2 cytokines IL5 and IL13 and altered ratios of IgG antibody subclasses (8,9,(11)(12)(13)15). This is similar to the vaccine enhanced disease observed with early vaccine development against respiratory syncytial virus (RSV); pathology was associated with relatively high titers of non-neutralising antibodies, a role for neutrophils, eosinophils and a predominantly a Th2biased response was described (19)(20)(21)(22)(23).
Preclinical studies of vaccines against SARS CoV-2 must therefore determine whether enhanced disease occurs in vaccinated animals once exposed to SARS-CoV-2 virus. Multiple studies of SARS-CoV-2 vaccines (3-8) have now been conducted in rhesus macaques without demonstrating enhanced disease.
Using CT scanning, we show that the lung pathology associated with infection with 5x10 6 pfu of SARS-CoV-2 in rhesus macaques closely mirrors that seen in humans with mild pneumonia caused by COVID-19, and is reduced in animals vaccinated with a single dose of ChAdOx1 nCoV-19 during the rst week post-infection. Histopathology performed on the lung tissues also indicated that the lesions observed in vaccinated animals are less severe than in controls at 7 days post-challenge, but also at 13/14 days postchallenge. The presence of viral RNA (using a probe that does not allow determination of viral replication) by ISH is also less frequent in vaccinated animals.
We demonstrate in the ferret model that virus shedding early after challenge with 5x10 6 pfu of SARS-CoV-2 was reduced in ChAdOx1 nCoV-19 vaccinated animals. A second vaccination with ChAdOx1 nCoV-19 transiently increased antibody titres in ferrets, which showed a negative correlation with the total virus shed in nasal washes. A range of antibody titres against the spike protein have been demonstrated in individuals who have had severe disease requiring hospitalisation, mild to moderate disease and asymptomatic infection. It is unclear what level of antibody titres against the viral spike protein are required to prevent infection or avert disease but it is generally accepted that high-titre neutralising antibodies are required.
Both animal models in this study con rmed the safety of vaccination with ChAdOx1 nCoV-19 when the respiratory tract is exposed to large quantities (5x10 6 pfu) of SARS-CoV-2 virus, and demonstrated reduced lung pathology as well as reduced nasal virus shedding in ferrets which correlated negatively with the neutralizing antibody titre induced by vaccination. Here, rhesus macaques were challenged by simultaneous virus instillation to both the upper and lower respiratory tract, as in some, but not all other vaccination and challenge studies (5)(6)(7)(8). Ferrets were challenged by the intranasal route only, but with the animal held vertically allowing some of the inoculum to enter the lungs. Both methods therefore result in immediate exposure of the lungs to SARS-CoV-2, whereas unless exposed to an extremely high concentration of virus the majority of human infections are likely to infect the upper respiratory tract initially, moving to the lungs if the infection is not rapidly controlled.
Currently there are no de ned correlates of protection against COVID-19 infection in humans, and the immunological thresholds required for vaccine e cacy remain unde ned (10). In a rhesus macaque SARS-CoV-2 infection model, protection against re-challenge was associated with immunologicallymediated control of infection with both neutralising and non-neutralising antibody as well as cellular responses increasing after secondary viral exposure (13). It is therefore speculated that high titre neutralising antibodies with a robust cytotoxic CD8 + T cell response and Th1 biased CD4 + effector response will be optimal for protective immunity following SARS-CoV-2 exposure, as demonstrated here. Viral vectored vaccines have been demonstrated to induce strong immune responses in older adults and immunocompromised individuals and have been used in repeat vaccinations, subsequently inducing strong cellular and humoral immunity (24)(25)(26)(27). ChAdOx1 nCoV-19 vaccination has previously been demonstrated to prevent SARS-CoV-2 mediated pneumonia in rhesus macaques (6), and this work is further supported and extended by the studies presented here.

Methods
Animals Twelve rhesus macaques of Indian origin (Macaca mulatta) were used in this study. Study groups comprised three males and three females and all were adults aged 4 years and weighing between 4.30 and 8.24kg at time of challenge. Before the start of the experiment, socially compatible animals were randomly assigned to challenge groups, to minimise bias. foraging. Following challenge, animals were transferred to ACDP Level three and housed in banks of cages of similar construction placed in directional air ow containment systems that allowed group housing and environmental control whilst providing a continuous, standardised inward ow of fully conditioned fresh air identical for all groups. Additional environmental enrichment was afforded by the provision of toys, swings, feeding puzzles and DVDs for visual stimulation. In addition to ad libitum access to water and standard old-world primate pellets, diet was supplemented with a selection of fresh vegetables and fruit. All experimental work was conducted under the authority of a UK Home O ce approved project license that had been subject to local ethical review at PHE Porton Down by the Animal Welfare and Ethical Review Body (AWERB) and approved as required by the Home O ce Animals (Scienti c Procedures) Act 1986. Animals were sedated by intramuscular (IM) injection with ketamine hydrochloride (Ketaset, 100mg/ml, Fort Dodge Animal Health Ltd, Southampton, UK; 10mg/kg) for procedures requiring removal from their housing. None of the animals had been used previously for experimental procedures. Twenty-eight healthy, female ferrets (Mustela putorius furo) aged 5-7 months were obtained from a UK Home O ce accredited supplier (Highgate Farm, UK). The mean weight at the time of challenge was 973 g/ferret (range 825 to 1129g). Animals were housed as described previously (bioRxiv 2020.05.29.123810; doi: https://doi.org/10.1101/2020.05.29.123810). All experimental work was conducted under the authority of a UK Home O ce approved project licence that had been subject to local ethical review at PHE Porton Down by the Animal Welfare and Ethical Review Body (AWERB). One animal in the ChAdOx1 nCoV-19 prime only group steadily lost weight from arrival at the facility (5 days prior to vaccination) and throughout the post-vaccination follow-up and was sacri ced on welfare grounds at day 14 of the study. As the weight loss was observed from arrival it was not deemed vaccine related, therefore all immunological data from this animal has been excluded from the analysis.
Vaccinations Rhesus macaques received 2.5x10 10 vp ChAdOx1 nCoV-19 administered in 100ml intramuscularly or received 100ml of phosphate buffered saline intramuscularly and were challenged with SARS-CoV-2 twenty-seven days later. Ferrets were randomly assigned to ChAdOx1 nCoV-19 and ChAdOx1 GFP vaccinated groups. An identi er chip (Bio-Thermo Identichip, Animalcare Ltd, UK) was inserted subcutaneously into the dorsal cervical region of each animal. Ferrets were immunised with 2.5 x 10 10 virus particles of ChAdOx1 nCoV-19 or ChAdOx1 GFP intramuscularly administered as a 100ml volume into the hind leg. Twenty-eight days after vaccination, half of the vaccinated animals were challenged with SARS-CoV-2, while the other half received a booster dose of ChAdOx1 nCoV-19 of ChAdOx1 GFP and were challenged with SARS-CoV-2 a further twenty-eight days later.
Enzyme-linked immunosorbent assay Maxisorp plates (Nunc) were coated overnight at 4°C with 250 ng/well spike protein in PBS, prior to blocking with 100 µl of casein in PBS (Thermo Fisher) for 1hr at RT. NHP serum was serially diluted 2x in casein in PBS was incubated at RT for 1hr. Antibodies were detected using a nity-puri ed polyclonal antibody alkaline phosphatase-labelled goat-anti-monkey IgG (Rocklands Laboratories), anti-monkey IgM (Rockland Laboratories) or anti-monkey IgA (Rockland Laboratories) in casein and developed with NPP-substrate (Sigma) and read at 405 nm. All wells were washed at least 3x with PBST 0.05% tween in between steps. Endpoint titers were calculated as follows: the log 10 OD against log 10 sample dilution was plotted and a regression analysis of the linear part of this curve allowed calculation of the endpoint titer with an OD of three times the background. Ferret serum was diluted in casein and incubated at RT for 2hr. Antibodies were detected using a nity-puri ed polyclonal antibody HRP-labelled goat-anti-ferret IgG (Abcam) in casein and TMB highest sensitivity (Abcam), developed for 12 minutes, and reaction was stopped using H 2 SO 4 and read at 450 nm. Antispike IgM or IgA antibodies were detected with alkaline Phosphatase conjugated anti-ferret IgM (Rockland Laboratories) or anti-ferret IgA (Sigma), development with NPP-substrate and read at 405 nm. All wells were washed at least 3x with PBST 0.05% tween in between steps. Ferret samples were run against a standard positive pool of serum generated from ChAdOx1 nCoV-19 vaccinated ferrets with high endpoint titre. Due to high levels of non-speci c responses, background was de ned as the mean + 2x stdev of all animals at day 0.

85020206), European Collection of Authenticated Cell Cultures, UK] monolayers in 24-well plates
(Nunc, ThermoFisher Scienti c, Loughborough, UK) under MEM (Life Technologies, California, USA) containing 1.5% carboxymethylcellulose (Sigma), 5% (v/v) foetal calf serum (Life Technologies) and 25mM HEPES buffer (Sigma). After incubation, at 37°C for 96 hours, plates were xed overnight with 20% (w/v) formalin/PBS, washed with tap water and stained with methyl crystal violet solution (0.2% v/v) (Sigma). The neutralising antibody titres were de ned as the serum dilutions resulting in a 50% reduction relative to the total number of plaques counted without antibody by using Probit analysis written in R programming language for statistical computing and graphics. An internal positive control for the PRNT assay was run using a sample of human MERS convalescent serum known to neutralise SARS-CoV-2 (National Institute for Biological Standards and Control, UK)
ELISpot PBMCs from rhesus macaques and ferrets were isolated from whole blood by layering over Lymphoprep (density 1.077g) and centrifugation for 30 minutes at 1000g. PBMCs were collected from the interface, washed with Hanks Balanced Salt Solution (HBSS) prior to resuspension in complete media (RPMI supplemented with 10% FCS, Pent-Strep, L-Glut and Hepes). IFNg ELISpot assay was performed using NHP IFNg (Mabtech) or Ferret IFNg ELISpot BASIC Kit according to the manufacturer's protocol (MABtech). PBMCs were plated at a concentration of 250 000 cells per well (NHPs) or 100 000 cells per well (Ferrets) and were stimulated overnight (18 to 20 hours) with four contiguous peptide pools spanning the length of the SARS-CoV-2 spike protein sequence at a concentration of 2µg/mL per peptide (Mimotopes) ( Table S7). Spots were counted and analysed on an AID ELISpot Reader (AID). Spot forming units (SFU) per 1.0x10 6 PBMCs were summed across the 4 peptide pools for each animal after subtraction of background response (media and PBMC only wells). Simultaneous production of IFNg, spike peptides (S1-pool 1 and pool 2 or S2-pool 3 and pool 4) at a nal concentration of 2µg/ml or ConA in the presence of golgi-stop (BD) and golgi-plug (BD). Cells were surface stained with anti-mouse/rat/human CD3 Alexa 405 (Clone PC3/188A) (Santa Cruz Biotechnology), anti-human CD8 APCCy7 (Clone OKT8) (Thermo sher) and live-dead aqua (Thermo sher), xed with Fix-Perm solution prior to intracellular staining with anti-bovine IFNg PE (Clone CC302) (Abserotec) and antimouse TNFa A647 (Clone MP6-XT22) (D28 samples). Data was acquired on a BD Fortessa and analysed in FlowJo version 9 or above. Data is presented total spike response, by summing together the frequency of cytokine positive cells detected in S1 and S2 stimulated wells after background subtraction of media stimulated cells.
Computed Tomography (CT) Radiology of NHPs CT scans were collected from sedated macaques using a 16 slice Lightspeed CT scanner (General Electric Healthcare, Milwaukee, WI, USA) in the prone and supine position. The change in position assists differentiation between pulmonary changes due to gravity dependant atelectasis from ground glass opacity at the lung bases caused by COVID. All axial scans were performed at 120KVp, with Auto mA (ranging between 10 and 120) and were acquired using a small scan eld of view. Rotation speed was 0.8s. Images were displayed as an 11cm eld of view. To facilitate full examination of the cardiac / pulmonary vasculature, lymph nodes and extrapulmonary tissues, Niopam 300 (Bracco, Milan, Italy), a non-ionic, iodinated contrast medium, was administered intravenously (IV) at 2ml/kg body weight and scans collected immediately after injection and ninety seconds from the mid-point of injection. Scans were evaluated by an expert thoracic radiologist, blinded to the animal's treatment and clinical status for the presence of COVID disease features: ground glass opacity (GGO), consolidation, crazy paving, nodules, peri-lobular consolidation; distribution -upper, middle, lower, central 2/3, peripheral, bronchocentric) and for pulmonary embolus.
The extent of lung involvement was estimated (<25%, 25-50%, 51-75%, 76-100%) and quanti ed using a scoring system developed for COVID disease, as follows: . This results in 12 zones in total. Measures: COVID pattern score = Nodule score + GGO score + consolidation score. Distribution (Zone) score = number of zones with disease, maximum score 12. Total CT score = COVID pattern score + Distribution (zone) score Whole Blood Immunophenotyping Assays were performed using 50µl of heparinised blood incubated for 30 minutes at room temperature with optimal dilutions of the following antibodies: anti-CD3-AF700, anti-CD4-APC-H7, anti-CD8-PerCP-Cy5.5, anti-CD95-Pe-Cy7, anti-CD14-PE, anti-HLA-DR-BUV395, anti-CD25-FITC (all from BD Biosciences, Oxford, UK); anti-γδ-TCR-BV421, anti-CD16-BV786, anti-CD20-PE-Dazzle (all from BioLegend); and amine reactive xable viability stain red (Life Technologies); all prepared in brilliant stain buffer (BD Biosciences). Red blood cell contamination was removed using a Utilyse reagent kit as per the manufacturer's instructions (Agilent). BD Compbeads (BD Biosciences) were labelled with the above uorochromes for use as compensation controls. Following antibody labelling, cells and beads were xed in a nal concentration of 4% paraformaldehyde solution (Sigma Aldrich, Gillingham, UK) prior to ow cytometric acquisition. Cells were analysed using a ve laser LSRII Fortessa instrument (BD Biosciences) and data were analysed using FlowJo (version 9.7.6, BD Biosciences). Immediately prior to ow cytometric acquisition, 50 µl of Truecount bead solution (Beckman Coulter) was added to each sample. Leukocyte populations were identi ed using a forward scatter-height (FSC-H) versus side scatterarea (SSC-A) dot plot to identify the lymphocyte, monocyte and granulocyte populations, to which appropriate gating strategies were applied to exclude doublet events and non-viable cells. Lymphocyte 2) with quanti cation between 1 x 10 1 and 1 x 10 6 copies/µl. Positive samples detected below the limit of quanti cation (LOQ) were assigned the value of 5 copies/µl, whilst undetected samples were assigned the value of < 2.3 copies/µl, equivalent to the assay's lower limit of detection (LLOD).
Histopathology NHPs: Each animal was assigned a histology number for blinding purposes. The following samples from each animal was xed in 10% neutral-buffered formalin, processed to para n wax and 4 µm thick sections cut and stained with haematoxylin and eosin (H&E); respiratory tract (left cranial and caudal lung lobes), trachea, larynx, tonsil, liver, kidney, spleen, mediastinal lymph node, and small and large intestine. Tissue sections were examined by light microscopy and evaluated subjectively and semi-quantitatively using a scoring system. Pathologists were blinded to treatment and group details and the slides randomised prior to examination in order to prevent bias (blind evaluation).
The slides were reviewed independently by three board-certi ed veterinary pathologists. For the lung, three sections from each left lung lobe were sampled from different locations: proximal, medial and distal to the primary lobar bronchus. The scoring system was applied using the following parameters and  Spike-speci c T cell response in rhesus macaques a. and ferrets b. monitored by IFNy ELISpot following vaccination and ICS (ferrets only). Response from ChAdOx1 nCoV-19 vaccinated NHPs was analysed with a Friedman one-way anova and post-hoc test. Response in ferrets was analysed with a nonparametric one-way anova (Kruskal Wallis) and post-hoc Dunn's multiple comparison test. A signi cant increase in the response compared to Day 0 was observed from day 14 onwards, with no statistically signi cant increase in the T cell response following booster vaccination. T cell responses in ferrets were measured by intracellular cytokine staining on day 28 post-vaccination and compared to responses measured by IFNy ELISpot.  Immune responses following challenge with SARS-CoV-2 a. Immune responses following challenge of rhesus macaques with SARS-CoV-2 was measured in virus neutralisation assays and by IFNγ ELISpot. b. Quanti cation of CD4+ and CD8+ T cells expressing HLA-DR and PD-1 prior to (day 0) and at days 3, 6-7 (7) and 13-14 post SARS-CoV-2 challenge of NHPs. c. Quanti cation of NHP monocyte sub-populations determined by expression of CD14 and CD16 by whole blood immunophenotyping ow cytometry assay.