key: cord-0982341-i6u0yh7s
authors: Luo, Shengxue; Zhang, Panli; Liu, Bochao; Yang, Chan; Liang, Chaolan; Wang, Qi; Zhang, Ling; Tang, Xi; Li, Jinfeng; Hou, Shuiping; Zeng, Jinfeng; Fu, Yongshui; Allain, Jean-Pierre; Li, Tingting; Zhang, Yuming; Li, Chengyao
title: Prime-boost vaccination of mice and Rhesus macaques with two novel adenovirus vectored COVID-19 vaccine candidates
date: 2020-09-29
journal: bioRxiv
DOI: 10.1101/2020.09.28.311480
sha: d9da113df1ac759568388f0d46fc21850b98ac2d
doc_id: 982341
cord_uid: i6u0yh7s

COVID-19 vaccines are being developed urgently worldwide, among which single-shot adenovirus vectored vaccines represent a major approach. Here, we constructed two novel adenovirus vectored COVID-19 vaccine candidates on simian adenovirus serotype 23 (Sad23L) and human adenovirus serotype 49 vectors (Ad49L) carrying the full-length gene of SARS-CoV-2 spike protein (S), designated Sad23L-nCoV-S and Ad49L-nCoV-S vaccines, respectively. The immunogenicity elicited by these two vaccine strains was individually evaluated in mice. Specific humoral and cellular immune responses were proportionally observed in a dose-dependent manner, and stronger response was obtained by boosting. Furthermore, five rhesus macaques were intramuscularly injected with a dose of 5×109 PFU Sad23L-nCoV-S vaccine for prime vaccination, followed by boosting with 5×109 PFU of Ad49L-nCoV-S vaccine at 4-week interval. Three macaques were injected with Sad23L-GFP and Ad49L-GFP vectorial viruses as negative controls. Both mice and macaques tolerated well the vaccine inoculations without detectable clinical or pathologic changes. In macaques, prime-boost vaccination regimen induced high titers of 103.16 S-binding antibody (S-BAb), 102.75 cell receptor binding domain (RBD)-BAb and 102.38 neutralizing antibody (NAb) to pseudovirus a week after boosting injection, followed by sustained high levels over 10 weeks of observation. Robust IFN-γ secreting T-cell response (712.6 SFCs/106 cells), IL-2 secreting T-cell response (334 SFCs/106 cells) and intracellular IFN-γ expressing CD4+/CD8+ T cell response (0.39%/0.55%) to S peptides were detected in the vaccinated macaques. It was concluded that prime-boost immunization with Sad23L-nCoV-S and Ad49L-nCoV-S vaccines can safely elicit strong immunity in animals in preparation of clinical phase 1/2 trials.

Ad49L-nCoV-S immunized mice, but not in the control groups (P<0.001, Fig. 3 , 165 A-D). A dose of 10 9 PFU Sad23L-nCoV-S vaccine induced antibody titers of 10 4.27 166 S-BAb and 10 3.98 RBD-BAb (Fig. 3, A and B) , while the same dose of 167

Ad49L-nCoV-S vaccine induced titers of 10 3.02 S-BAb and 10 2.42 RBD-BAb, 168 respectively (Fig. 3, C and D) . 169

The neutralizing antibody (NAb) titers to SARS-CoV-2 were titrated in two 170 individual vaccine immunized mice by surrogate virus based NAb test (sVNT) and 171

pseudovirus-based NAb test (pVNT) at 50% inhibitory dilution (ID50), respectively 172 ( Fig. 3 , E-H). A dose of 10 9 PFU Sad23L-nCoV-S vaccine induced NAb titers of 173 10 2.41 sVNT(ID50) and 10 2.79 pVNT (ID50) (Fig. 3 , E and F), and Ad49L-nCoV-S 174 vaccine induced titers of 10 1.38 sVNT (ID50) and 10 1.57 pVNT(ID50), respectively (Fig.  175 

Specific T-cell response of isolated splenocytes was examined in vaccinated and 177 sham mice after stimulation with S peptides, S protein and RBD protein, respectively 178 intracellular levels of IFN-γ, TNF-α and IL-2 expressing CD4 + or CD8 + T cell 186 response to S peptides were measured by ICS, in which significantly higher frequency 187 of IFN-γ and TNF-α but not IL-2 expressing CD4 + /CD8 + T-cells was found in both 188

Sad23L-nCoV-S and Ad49L-nCoV-S vaccinated mice compared to sham mice 189 (P<0.001, Fig. 3, M and N; Fig. S3, Overall, individual Sad23L-nCoV-S or Ad49L-nCoV-S vaccinated mice 191 developed specific BAb and NAb antibodies and T cell responses to S protein, RBD 192 protein or S peptides of SARS-CoV-2 in a dose-dependent fashion (Fig. 3) , 193 suggesting strong immunogenicity of the two novel adenovirus vectored Prime-boost immunization of mice with Sad23L-nCoV-S and Ad49L-nCoV- S 197 vaccines 198 To further improve reactivity and longevity of immune response to vaccines, the 199 prime-boost vaccination regimen was utilized to immunize C57BL/6 and BALB/c 200 mice (n=5/group) with a prime dose of 10 9 PFU Sad23L-nCoV-S on day 0 and a 201 boost dose of 10 9 PFU Ad49L-nCoV-S on day 28 (Fig. 4A ). In comparison with 202 prime immunization with Sad23L-nCoV-S, a booster with Ad49L-nCoV-S 203 significantly increased BAb and NAb titers in C57BL/6 and BALB/c mice (P<0.001; 204 Fig. 4, , and titers were maintained at high level for 10 weeks of monitoring (Fig. 205 4, D and G) . Profiling of IgG subclasses showed a predominant serum IgG2a to RBD 206 protein associating with Th1 response in prime-boost vaccinated mice (Fig. S4) . 207

Regarding specific T cell response, the boosting with Ad49L-sCoV-S on 208

Sad23L-nCoV-S prime immunization enhanced or stabilized specific IFN-γ-secretion 209 T cell responses to S peptides, S protein or RBD protein (Fig. 4H) , as well as levels of 210 intracellular cytokines IFN-γ and TNF-α but not of IL-2 response to S peptides 211 compared with single dose of vaccine vaccinated or sham control C57BL/6 and 212 BALB/c mice ( Fig. 4I; Fig. S5 ). 213

Taken together, the results suggest that prime-boost vaccination of two species of 214 mice with Sad23L-nCoV-S followed by Ad49L-nCoV-S enhanced specific immune 215 response to SARS-CoV-2 when compared with prime vaccination only with a 216 single-shot of Sad23L-nCoV-S or Ad49L-nCoV-S vaccine. 217 218 Rhesus macaques' specific immune response to prime-boost vaccination with 219 combined Sad23L-nCoV-S and Ad49L-nCoV-S vaccines 220

Eight rhesus macaques aged 11-14 years were selected and tested for the baseline 221 values of antibody and T-cell response to SARS-CoV-2 S from blood samples in 222 pre-vaccination at week -1 or week 0 ( Fig. 5A ; Table S1 ). The pre-existing NAb titer 223 to Sad23L, Ad49L or Ad5 was detected < 1:10 in serum samples from all eight 224 animals (Table. S1). 225

Five rhesus macaques were intramuscularly injected first with a dose of 5×10 9 226 PFU Sad23L-nCoV-S vaccine, then with a boost injection of 5×10 9 PFU 227

Ad49L-nCoV-S vaccine 4 weeks later, while three sham rhesus macaques controls 228 were injected with equal doses of Sad23L-GFP and Ad49L-GFP viruses, respectively 229 ( Fig. 5A ). Blood samples were collected weekly from these two groups of animals. 230

Both BAb (S-BAb and RBD-BAb) and NAb (sVNT and pVNT) levels were titrated 231 in the vaccinated group but not in sham group. BAb and NAb reactivity increased at 232

week 2 post prime-immunization, at week 4 (W0 for boost-immunization), at weeks 5 233 and 6 and then stayed at high levels up to week 10 ( PBMCs were isolated from whole blood of pre-and post-immunized monkeys 241

for evaluation of T cell responses to S peptides, S and RBD protein by ELISpot and 242 ICS (Fig. 6) . Prime vaccination with a dose of Sad23L-nCoV-S vaccine induced an 243 increase of IFN-γ secreting T cell response to S peptides (406.6-526.3 SFCs/million 244 cells) at weeks 2 and 4 post prime-vaccination, and then boosting with a dose of 245

Ad49L-nCoV-S vaccines enhanced the IFN-γ secreting T cell response (583.9-712.6 246 SFCs/million cells) at weeks 5 to 8 (Fig. 6A ). IFN-γ secreting T cell reaction to S and 247 RBD proteins stayed at high levels after prime-boost immunizations (Fig. 6 , B and C). 248

Relatively high IL-2 secreting T cell response to S peptides, S and RBD proteins was 249 observed (Fig. 6 , D-F), but weak IL-4 secreting T cell response by ELISpot (Fig. S6) . 250

Frequency of intracellular IFN-γ expressing CD4 + /CD8 + T-cell responses to S 251 peptides was observed in vaccinated macaques at weeks 2 to 8 (Fig. 6 , G-J), 252 significantly higher than observed in pre-vaccination and sham controls (P<0.01). 253

Frequency of intracellular TNF-α expressing CD4 + T cell response was also found 254 significantly different between vaccinated and sham monkeys (P<0.05; Fig. S7 , A and 255 B), but intracellular TNFα + CD8 + or IL-2 + CD4 + /CD8 + T cell response to S peptides 256 was not statistically different between groups (P>0.05; Fig. S7 , C-F). 257

In summary, prime-boost vaccination with Sad23L-nCoV-S and Ad49L-nCoV-S 258 vaccines at an interval of 4 weeks elicited higher levels of specific antibody and T-cell 259 responses against SARS-CoV-2 in rhesus macaques, which was recommended as 260 COVID-19 vaccine candidates for clinical trials in humans. 261 262 Biodistribution of Sad23L-nCoV-S and Ad49L-nCoV-S vaccine strains in the 263

Neutralizing antibody titers to Sad23L and Ad49L vectors were determined in rhesus 265 macaques and mice after prime only and prime-boost vaccinations (Fig. 7, A and B; 266 Fig. S8) . A high NAb reactivity to an individual adenoviral vector of either Sad23L or 267 Ad49L in macaques (1:1280 or 1:640) and mice (1:576 or 1:352) was induced post 268 prime immunization, but not enhanced post boost vaccination with a heterologous 269 vector (Fig. 7, A and B) , which might limit the homologous boosting of adenovirus 270 vectored vaccines. The lung, spleen, liver and muscle tissues (at intramuscular 271 injection site and para-tissues) from prime only or prime-boost immunized C57BL/6 272 mice were examined 4 weeks after inoculation of vaccines in order to assess vaccine 273 delivery by amplifying specific adenoviral hexon sequences of both Sad23L and 274

Ad49L vectors. The predicted 500bp PCR band was detected in lung, spleen and liver 275 tissues after Sad23L-nCoV-S vaccination, while the PCR band was observed in lung 276 and liver tissues but not found in the spleen after Ad49L-nCoV-S vaccination (Fig. 277 7C) . The expression of S antigen was observed in splenocytes and hepatocytes from 278 prime-only Sad23L-nCoV-S or Ad49L-nCoV-S and prime-boost vaccines inoculated 279 C57BL/6 mice by immunofluorescence staining (Fig. 7D ), but not found in lung and 280 muscle tissues (Fig. S9) . In this study, we generated two novel adenovirus vectored COVID-19 vaccines 299 encoding the full-length S gene of SARS-CoV-2. The intact S glycoprotein rather 300 than the shorter S or RBD proteins was shown to be the most effective antigen 301 eliciting protective immunity against SARS-CoV-2 infection in DNA vaccines and 302 Ad26 vectored vaccines (12, 13, 26 S vaccines (8-12,14,21) , three attractive aspects emphasized in this 306 study are highlighted below. 307

Firstly, there is a low-seroprevalence of pre-existing antibodies to Sad23L and 308

Ad49L vectors in humans. According to the investigation of NAb to three types of 309 adenoviruses in Chinese population, the prevalence of both Sad23L and Ad49L NAb 310 was below 10%, while the prevalence of Ad5 NAb was over 75% (Fig. 1D ). However, 311 the pre-existing anti-Ad5 immunity might partly limit vaccine effectiveness, 312 especially for populations aged over 50 (8,9,24,25) , while the low seroprevalence of 313 antibodies to novel adenoviral vectors such as ChAdox1 (11), Ad26 (12, 14, 21, 27) , 314 and Sad23L and Ad49L used in this study might avoid a negative impact on vaccine 315 efficacy (22) (23) (24) . 316

Secondly, the prime-boost vaccination regimen with two heterologous 317 adenoviruses vectored Sad23L-nCoV-S and Ad49L-nCoV-S vaccines examined in 318 mice and rhesus macaques suggested that the boost with Ad49L-nCoV-S vaccine 319 significantly enhanced the levels of neutralizing antibody and IFN-γ expressing 320 CD4 + /CD8 + T cell responses following prime-immunization with Sad23L-nCoV-S 321 vaccine ( Fig. 4-6 ). In addition, specific IFN-γ secretion T cell response was prolonged 322 at high level after boosting vaccination. Two recent publications compared prime only 323 with a single-dose of ChAdOx1 nCoV-19 and homologous boost with a second dose 324 of ChAdOx1 nCoV-19 vaccine in mice and pigs, and human clinical trials, showed 325 that boost vaccination significantly increased and prolonged the level of binding or 326 neutralizing antibody response to 20) . By boosting with Ad5-S 327 vaccine the Ad26-S vaccine primed immunization, a higher level of immunity was 328 observed in humans as reported in a recent Russian study (21). Compared with 329 homologous boosting of ChAdOx1 nCoV-19 or boosting of 21) , 330 heterologous prime-boost vaccinations with Sad23L-nCoV-S and Ad49L-nCoV-S 331 vaccines have the advantage of avoiding vector's immunity interfering with prime 332 vaccination or enhancing pre-existing Ad5 immunity (22) (23) (24) (25) . 333

Thirdly, in order to achieve an effective vaccine immunity, a low dose of two 334 heterologous adenovirus vectored vaccines (<5×10 10 vp) with prime-boost 335 immunization regimen should theoretically reduce severe adverse reaction induced by 336 a high dose of adenovirus vectored vaccine (>5×10 10 vp) with prime only 337 immunization in clinical trials (8, 9, 11, 14) . In this study, a relatively low dose of 338 Sad23L-nCoV-S and Ad49L-nCoV-S vaccines elicited a robust immunity in both 339 young mice and older rhesus macaques (aged 11-14 years), but no obvious clinical 340 symptoms or histopathological changes were observed ( 

Eight healthy outbred male rhesus macaques (Macaca mulatta) aged 11-14 years were 371 randomly allocated to this study (Table S1 ). Experimentation and sample collection 372 were ethically approved and carried out by the Huazheng Laboratory Animal 373

Breeding Centre, Guangzhou, China. All animal care and experimental procedures 374 (NFYYLASOP-037) were in accordance with national and institutional policies for 375 animal health and wellbeing. 376

The welfare issues (housing, feeding, environmental enrichment, etc.) were in 377 accordance with the recommendations of the Weatherall report 378 (https://acmedsci.ac.uk/more/news/the-use-of-non-human-primates-in-research). 379

Animals were individually housed in spacious cages and were provided with 380 commercial food pellets supplemented with appropriate treats. Drinking water was 381 provided ad libitum from an automatic watering system. Animals were monitored 382 daily for health and discomfort. Blood samples were obtained using sterilized needle 383 and syringe from the venous vessels of animal legs. Ad49L-GFP vectorial viruses by green fluorescent activity assay as previously 445 described (22, 29) . 446 447

Female C57BL/6 mice (5-6 weeks, n=5 each group) were individually inoculated 449 intramuscularly (i.m.) with a dose of 10 7 , 10 8 and 10 9 PFU Sad23L-nCoV-S or 450

Ad49L-nCoV-S vaccine, respectively. A dose of 10 9 PFU Sad23L-GFP or 451

Ad49L-GFP vectorial virus and an equivalent volume of PBS were used as sham 452 control. 453 Female C57BL/6 and BALB/c mice (5-6 weeks, n=5 each group) were prime 454 inoculated intramuscularly with a dose of 10 9 PFU Sad23L-nCoV-S vaccine, and then 455 at 4 week interval were boosted with a dose of 10 9 PFU Ad49L-nCoV-S vaccine. A 456 dose of 10 9 PFU Sad23L-GFP and a dose of 10 9 PFU Ad49L-GFP were used as sham 457 control. 458

Five rhesus macaques aged 11 to 14 years (Table S1) were first injected 459 intramuscularly with a dose of 5×10 9 PFU Sad23L-nCoV-S vaccine, and then at 4 460

week interval received a second dose of 5×10 9 PFU Ad49L-nCoV-S vaccine. Three 461 rhesus macaques aged 11 to 13 years (Table S1) China) and goat anti-human IgG-HRP conjugates (Sigma, USA), respectively. 472

Endpoint titers were defined as the highest reciprocal serum dilution that yielded an 473 absorbance > 0.2 and a ratio of signal than cutoff (S/CO) > 1. Log10 Pseudoviruses expressing a luciferase reporter gene were generated for measuring of 493

NAb to SARS-CoV-2 as previously described (26) 

Data are analyzed with unpaired two-tailed t test, one-way ANOVA. Neutralizing 536 antibody titer data were log transformed before analysis. Neutralizing antibody titer 537 data generated by the RBD-BAb and sVNT or S-BAb and pVNT assays were 538 compared using Spearman nonparametric correlation. Statistically significant 539 differences are indicated with asterisks (* P <0.05; ** P <0.01 and *** P <0.001). Ad49L-nCoV-S immunized mice 659 Table S1 . Basic information for rhesus macaques pre-vaccination 660 Table S1 . Basic information for rhesus macaques pre-vaccination Table S2 : Measuring of hematological and biochemistry indexes of rhesus macaques in the course of pre-and post-vaccination with Sad23L-nCoV-S and Ad49L-nCoV-S vaccines Table S3 . Peptides derived from amino acid sequences of SARS-CoV-2 S protein used in ELISpot and ICS Table S4 . Nested-PCR primers specific for hexon of Sad23L or Ad49L vector

Ad49L-nCoV-S vaccines inoculated C57BL/6 mice. Histopathological examination was carried out for brain, lung, heart, liver, kidney and muscle tissues (at intramuscular injection site and para-tissues) of mice in 4 weeks post prime only or prime-boost immunization with these two vaccines. Tissues were stained with hematoxylin and eosin. 

Examination of weight and body temperature of mice and rhesus 707 macaques inoculated with Sad23L-nCoV-S and Ad49L-nCoV-S vaccines

Animals were intramuscularly immunized by prime only with either Sad23L-nCoV-S 709

or Ad49L-nCoV-S vaccine at three different doses, or by prime-boost with first 710

Body 711 weight and temperature were monitored during the course up to 60 days for mice or 8 712 weeks for monkeys. (A) C57BL/6 mice (n=5/group

10 8 or 10 9 PFU Sad23L-nCoV-S vaccine, 10 9 PFU Sad23L-GFP and an equal

of PBS controls (top panel); a dose of 10 7 , 10 8 or 10 9 PFU Ad49L-nCoV-S vaccine

10 9 PFU Ad49L-GFP and an equal volume of PBS controls (middle panel); a dose of 716 37

PFU Sad23L-nCoV-S followed by a dose of 10 9 PFU Ad49L-nCoV-S, or 10

Ad49L-GFP vectorial viruses and an equal volume of PBS 718 controls (low panel). (B) Rhesus monkeys (11-14y) inoculated with a dose of 5×10 9

PFU Sad23L-nCoV-S and a dose of 5×10 9 PFU Ad49L-nCoV-S vaccines (R1, R2

R7 and R8), or 5×10 9 PFU Sad23L-GFP and a dose of 5×10 9 PFU Ad49L-GFP 721 controls (R3, R4 and R6)

test (pVNT). (I-L) IFN-γ secreting T cell response (spot forming cells [SFCs]/million 733 cells) of splenocytes to S peptides and S protein from

Ad49L-nCoV-S immunized mice was measured by ELISpot, respectively

Frequency of IFN-γ or TNF-α expressing CD4 + and CD8 + T cell response of 736 splenocytes to S peptides from Sad23L-nCoV-S or Ad49L-nCoV-S immunized mice 737 was obtained by ICS. Data is shown as mean ± SEM (standard errors of means). P 738 values are analyzed by one-way ANOVA with 2-fold Bonferroni adjustment

Total protein

Alkaline phosphatase

Glu, Glucose; BUN, Blood urea nitrogen