key: cord-0426824-sgi2097b
authors: Zhou, Runhong; Wang, Pui; Wong, Yik-Chun; Xu, Haoran; Lau, Siu-Ying; Liu, Li; Mok, Bobo Wing-Yee; Peng, Qiaoli; Liu, Na; Woo, Kin-Fai; Deng, Shaofeng; Tam, Rachel Chun-Yee; Huang, Haode; Zhang, Anna Jinxia; Zhou, Dongyan; Zhou, Biao; Chan, Chun-Yin; Du, Zhenglong; Yang, Dawei; Au, Ka-Kit; Yuen, Kwok-Yung; Chen, Honglin; Chen, Zhiwei
title: Nasal prevention of SARS-CoV-2 infection by intranasal influenza-based boost vaccination
date: 2021-10-22
journal: bioRxiv
DOI: 10.1101/2021.10.21.465252
sha: 9121c7b0ae5f05ae986de98eeddb74bf6be73aa0
doc_id: 426824
cord_uid: sgi2097b

Background Vaccines in emergency use are efficacious against COVID-19, yet vaccine-induced prevention against nasal SARS-CoV-2 infection remains suboptimal. Methods Since mucosal immunity is critical for nasal prevention, we investigated an intramuscular PD1-based receptor-binding domain (RBD) DNA vaccine (PD1-RBD-DNA) and intranasal live attenuated influenza-based vaccines (LAIV-CA4-RBD and LAIV-HK68-RBD) against SARS-CoV-2. Findings Substantially higher systemic and mucosal immune responses, including bronchoalveolar lavage IgA/IgG and lung polyfunctional memory CD8 T cells, were induced by the heterologous PD1-RBD-DNA/LAIV-HK68-RBD as compared with other regimens. When vaccinated animals were challenged at the memory phase, prevention of robust SARS-CoV-2 infection in nasal turbinate was achieved primarily by the heterologous regimen besides consistent protection in lungs. The regimen-induced antibodies cross-neutralized variants of concerns. Furthermore, LAIV-CA4-RBD could boost the BioNTech vaccine for improved mucosal immunity. Interpretation Our results demonstrated that intranasal influenza-based boost vaccination is required for inducing mucosal and systemic immunity for effective SARS-CoV-2 prevention in both upper and lower respiratory systems. Funding This study was supported by the Research Grants Council Collaborative Research Fund (C7156-20G, C1134-20G and C5110-20G), General Research Fund (17107019) and Health and Medical Research Fund (19181052 and 19181012) in Hong Kong; Outbreak Response to Novel Coronavirus (COVID-19) by the Coalition for Epidemic Preparedness Innovations; Shenzhen Science and Technology Program (JSGG20200225151410198); the Health@InnoHK, Innovation and Technology Commission of Hong Kong; and National Program on Key Research Project of China (2020YFC0860600, 2020YFA0707500 and 2020YFA0707504); and donations from the Friends of Hope Education Fund. Z.C.’s team was also partly supported by the Theme-Based Research Scheme (T11-706/18-N).

10 and NA) in the A/CA/04/2009 (H1N1) DelNS1 backbone by a reverse genetic procedure (24) . 284

The SARS-CoV-2 RBD and LAIV NP proteins were stably expressed in MDCK cells by 285

Western blot after 5 times of viral passages in MDCK cells (Fig. S1E) . 286 287

We then chose the immune competent SARS-CoV-2/BALB/c mouse model (25), which was 289 based on available antibody reagents to understand the potential correlate of immune protection. 290

As compared with the vector control group (v1) that received 50 µg intramuscular 291 electroporation (i.m./EP) pVAX plasmid prime plus i.n. 10 6 PFU LAIV-68 vector boost (Fig. 292 1A), we tested groups treated by the homologous 50 µg i.m./EP PD1-RBD-DNA twice (v2), a 293 heterologous i.n. 10 6 PFU LAIV-CA4-RBD prime plus i.n. 10 6 PFU LAIV-HK68-RBD boost 294 (v3), a heterologous 50 µg i.m./EP PD1-RBD-DNA prime plus i.n. 10 6 PFU LAIV-68-RBD 295 boost (v4) and a heterologous i.n. 10 6 PFU LAIV-CA4-RBD prime plus 50 µg i.m./EP PD1-296 RBD-DNA boost (v5) at 3-week intervals, consistent with COVID-19 vaccines under emergency 297 use. Vaccine-induced antibody responses were determined at day 9 (acute phase) and day 69 298 (memory phase) post the second vaccination. Both peripheral blood and bronchoalveolar lavage 299 (BAL) samples were collected from vaccinated mice for antibody detection by ELISA and 300 pseudovirus neutralizing assays. We found that the PD1-RBD-DNA/LAIV-HK68-RBD regimen 301 (v4) elicited and significantly sustained the highest amounts of RBD-specific IgG (acute: mean 302 5.2, range 4.86-5.45 logs AUC; memory: mean 4.62, range 4.54-4.69 logs AUC) and NAbs 303 (acute: mean 4.19, 4.04-4.34 logs IC 50 ; memory: mean 2.89, 2.4-3.23 logs IC 50 ) in sera during 304 both acute and memory phases as compared with other groups (Fig. 1B-C) . The RBD-specific 305

IgG titer and NAb IC 50 values were correlated positively (Fig. 1D ), similar to COVID-19 306 patients' sera (8). Moreover, v4 animals also developed and sustained significantly higher 307 amounts of RBD-specific mucosal IgG and IgA in BAL during both acute and memory phases as 308 compared with other groups (Fig. 1E-F) . The amount of BAL NAbs in v4 mice were at mean 309 2.80 (range 2.22-2.87) and mean 2.59 (range 2.02-3.15) logs IC 50 at the acute and memory phase, 310 respectively (Fig. 1G ). In contrast, v2 and v5 elicited similar amounts of NAb to v4 at the acute 311

phase, yet these responses did not sustain into the memory phase. Moreover, despite 312 heterologous intranasal immunizations twice, the v3 regimen did not induce equally potent and 313 sustained mucosal IgG and IgA as well as NAb responses, as compared with the v4 group ( Fig.  314 11 1G). Both BAL IgG and IgA titers correlated positively with the BAL NAb values (Fig. 1H ) 315 despite of the higher amount and better acute/memory of BAL IgG than BAL IgA. Interestingly, 316 serum NAb IC 50 values were correlated positively with the BAL NAbs IC 50 values at the memory 317 phase, but not correlated at the acute phase (Fig. 1I) . These results indicated that the v4 regimen 318 is likely unique for inducing potent and sustained systemic and mucosal memory IgG/IgA NAb 319 responses. 320 321

Since SARS-CoV-2-specific T cell responses are essential for control and resolution of viral 323 infection (12, 33), we sacrificed five groups of animals to measure vaccine-induced T cell 324 immune responses at day 9 (acute phase) and day 69 (memory phase) post the second 325 vaccination. Lymphocytes were isolated from both lungs (effector site) and spleens (secondary 326 lymph organ) of vaccinated mice for comparison. We found that the v4 regimen elicited and 327 sustained significantly higher frequencies of RBD-specific IFN-γ + CD8 T cells in lungs and 328 spleens during both acute ( Fig. 2A-B, Fig. S2A -B) and memory ( Fig. 2D-E, Fig. S2D -E) phases 329 as compared with other groups. Similar trends were found with IFN-γ + CD4 T cells elicited in 330 the v4 regimen but at lower frequencies. At the acute phase, the v4 regimen elicited the highest 331 mean frequency of RBD-specific IFN-γ + CD8 T cells (mean 28.83%, range 22-34.8%) in lungs 332 ( Fig. 2B) , which was even higher than that in spleens (mean 5.54%, range 3.63-7.06%) (Fig.  333 S2B). These cells included the highest frequencies of polyfunctional CTLs with a capacity of 334 releasing two (mean 21.09%, range 13.87-25.31%) or three (mean 2.27%, range 1.02-3%) 335 cytokines ( Fig. 2C ), which was also higher than those in splenic CD8 T cells releasing two 336 (mean 6.68%, range 4.68-9.17%) or three cytokines (mean 0.92%, range 0.7-1.36%) (Fig. S2C) . 337

At the memory phase, the v4 regimen sustained the highest mean frequency of RBD-specific 338 IFN-γ + CD8 T cells in lungs (mean 6.11%, range 2.05-9.7%) (Fig. 2E ) and spleens (mean 2.35%, 339 range 0.7-4.78%) (Fig. S2E ) as compared with other groups. These cells included the highest 340 frequencies of polyfunctional CTLs with a capacity of releasing two (mean 8.66%, range 2.61-341 12.41%) or three (mean 2.2, range 0.66-3.09%) cytokines (Fig. 2F) , which was higher than those 342 in splenic CD8 T cells releasing two (mean 3.59%, range 1.28-6.83%) or three cytokines (mean 343 0.69%, range 0.31-1.17%) (Fig. S2F) . These results demonstrated that besides Nabs, the v4 344 regimen also induced potent and polyfunctional memory CD8 T cell responses, especially in 345 12 lungs. Since overall immune responses induced by the heterologous v3 regimen were much 346 weaker than those by the v4 regimen, we also exanimated T cell responses against influenza 347 immunodominant nucleoprotein (NP) (34). At acute phase (Fig. S3A) , the v3 regimen induced 348 the highest frequencies of CD8 T cell response against influenza NP in lungs (mean 19.55%, 349 range 16.2-21.7%) as compared with v1 (mean 8.3%, range 7.77-8.9%) and v5 (mean 3.99%, 350 range 2.55-5.91%). Similar results were observed at the memory phase (Fig. S3B ). In contrast, 351 the v4 regimen induced significantly lower influenza NP-specific T cell response at both acute 352 (mean 2.26%, range 1.68-2.58%) and memory (mean 0.77%, range 0.49-1.06%) phases. The 353 heterologous prime using PD1-RBD-DNA instead of LAIV-CA4-RBD, therefore, offered an 354 advantage in promoting the RBD immunodominance likely by avoiding anti-vector immune 355 responses. 356 357

To investigate the efficacy of various vaccine regimens against the live intranasal SARS-CoV-2 359 challenge, we subsequently immunized additional groups of BALB/c mice (n=6 per group) using 360 the same doses and time interval as described above (Fig. 3A) . We did not include v5 due to low 361 mucosal immunogenicity and limited space in our animal P3 facility. Sera were collected at day 362 9 and day 28 after the 2 nd immunization to monitor anti-RBD IgG (Fig. 3B ) and neutralization 363 ( Fig. 3C) . Consistently, the highest IgG (acute: mean 4.79, range 4.7-4.89 logs AUC; memory: 364 mean 2.6, range 2.03-2.81 logs AUC) and neutralizing (acute: mean 3.9, range 3.69-4.14 logs 365 IC 50 ; memory: mean 2.94, range 2.34-3.33 logs IC 50 ) titers were induced in mice by the v4 366 regimen at both acute and memory phases. Immunized mice were then transduced with Ad5-367 hACE2 at the memory phase, 29 days post the boost vaccination for expressing human ACE2 in 368 nasal turbinate and lung (Fig. 3D) , followed by the intranasal SARS-CoV-2 challenge 6 days 369 later as previously described (25). At day 4 after the viral challenge, mice were sacrificed for 370 analysis. Lung specimen was harvested to quantify infectious viruses by plaque assay, viral load 371 by real-time PCR (RT-PCR) and infected cells by immunofluorescence staining (IF). We found 372 that all vaccinated animals had decreased infectious plaque-forming units (PFU) to the limit of 373 detection (10 PFU/mL) in lungs (Fig. 3E) . The v4 regimen, however, resulted in the most 374 significant genomic RdRp (gRdRp) drop in lungs by an average of 2.31 logs compared with 1.81 375 logs in v2 mice and 1.62 logs in v3 mice (Fig. 3F) . A similar observation was found in the 376 13 measurement of nucleocapsid protein (NP) subgenomic RNA (sgNP) (Fig. 3G ). These findings 377 demonstrated that immune responses induced by v2, v3 and v4 regimens had achieved 378 significant protection in lungs. To determine viral infection in both upper and lower respiratory 379 systems, we further performed immunofluorescence staining of SARS-CoV-2 NP antigen in both 380 lung (Fig. 3H ) and NT (Fig. 3I ) tissues. Since murine NT was too small to be sliced for viral load 381 tests, we only used it for the NP staining to maintain the necessary tissue structure. While 382 significantly reduced NP + cells were observed in lungs of v2 and v3 mice, infected cells were 383 barely found in lungs of v4 mice. Furthermore, no significantly reduced NP + cells were found in 384 NT of v2 and v3 mice as compared with v1 mice, but only a few NP + cells were seen in v4 mice. 385

Our results demonstrated that while protection was consistently found in lungs of vaccinated 386 animals, significant prevention of robust SARS-CoV-2 infection in NT was only achieved by the 387 v4 regimen. 388 389

Although we were not allowed to bring tissue specimens for measuring T cell immunity outside 391 the animal P3 laboratory after the SARS-CoV-2 challenge, we managed to determine if viral 392 infection recalled vaccine-induced NAbs for viral neutralization and clearance (35). By testing 393 RBD-specific IgG and NAb at day 28 (before challenge) and day 39 (also 4 dpi) after the 2 nd 394 vaccination, we found that SARS-CoV-2 infection indeed recalled significantly anti-RBD IgG 395 responses in all v2, v3 and v4 animals ( 

To further determine the vaccine efficacy, we tested the PD1-RBD-DNA/LAIV-HK68-RBD 410 regimen in K18-hACE2 transgenic mice, one of commonly used animal models for studying 411 COVID-19 (37-39). 8-week-old K18-hACE2 mice were vaccinated with various regimens using 412 the doses and time interval as described above (Fig. 4A ). The titer of RBD-specific IgG ( vaccine-induced protection (19, 44) . For example, the Novavax vaccine is effective against the 493 wildtype SARS-CoV-2 (95.6%) but provides reduced protection against the variants Alpha 494 (85.6%) and Beta (60%) (45). We, therefore, compared the neutralizing activity of immune sera 495 from PD1-RBD-DNA/BioNTech/Sinovac prime and LAIV-CA4-RBD boosted mice (Fig. 6 ) 496 against pseudotyped viruses that contain the D614G, the Alpha, the Beta and the Delta variants 497 ( Fig. 7A-B ) as described previously (19). Compared to the D614G viral strain, v6 sera showed 498 slightly enhanced neutralizing activity against the Alpha variant, while the sera of other groups 499 exhibited reduced neutralization against the Alpha, Beta and Delta variants (Fig. 7B) . In line 500 17 with a recent studies (19), the Beta and Delta variants were more resistant to neutralization by 501 sera from all vaccine regimens with an average fold reduction of 1.5-1.77 (v6), 3.00-3.60 (v7), 502 1.98-2.03 (v8), 1.56-2.45 (v9) and 1.26-1.59 (v10) as compared to the D614G strain, respectively 503 ( Fig. 7B) . Although animals in the LAIV-CA4-RBD boost regimens (v6, v8 and v10) showed 504 the mean 1.69-fold reduction against Beta or Delta variants, their mean NAb IC 50 titer of 1: 3633 505 (v6, range 1:1586-1:8250), 1:4370 (v8, range 1:2305-1:9068) and 1:3929 (v10, range 1:189-506 14393) remained high, respectively, which was superior or comparable to those of homologous-507 vaccinated mice as well as to the results of clinical vaccines against the wild type virus in murine 508 models (6, 46, 47). Importantly, the BAL from LAIV-CA4-RBD boost groups (v6, v8 and v10) 509

were still able to neutralize the Beta and Delta variants (v6: mean 453, range 117-762; v8: mean 510 280, range 93-465; v10: mean 281, range 50-760) although they showed the average fold 511 reduction of 2.64-2.90 (v6), 1.7-4.3 (v8) and 1.26-2.28 (v10) as compared to the D614G strain, 512 respectively ( Fig. 7C-D) . These results demonstrated that the systemic prime/LAIV boost 513 regimen-induced high amounts of systemic and mucosal NAbs may confer cross-protection 514 against the variants of concern before the tailor-made vaccines become available. 515 and lower respiratory tracts was achieved by the AD26 vaccine encoding the full S protein 552 against SARS-CoV-2 in rhesus macaques in another study (56) . Both studies, however, used the 553 macaque model that requires inoculation of 10 5 TCID 50 SARS-CoV-2 into each nare and 554 intratracheal for effective infection, which is not natural and is in great contrast to the robust 555 nature of NT infection in humans. Using a DNA vaccine encoding the full S immunogen, we 556 recently reported that there was significant protection in the lung but not in NT against SARS-557

CoV-2 in Syrian hamsters even though the vaccinated animals developed pseudovirus 558 neutralizing IC 50 titer larger than 1:1000 (17) . In this study, we consistently found that significant 559 

Temporal profiles of 626 viral load in posterior oropharyngeal saliva samples and serum antibody responses during 627 infection by SARS-CoV-2: an observational cohort study

A Novel Coronavirus from 629 Patients with Pneumonia in China

Safety 631 and Immunogenicity of Two RNA-Based Covid-19 Vaccine Candidates

Immunogenic 634 BNT162b vaccines protect rhesus macaques from SARS-CoV-2

Safety and Immunogenicity of SARS-CoV-2 mRNA-1273 Vaccine in Older Adults

SARS-CoV-2 mRNA vaccine design enabled by prototype pathogen preparedness

Port 642 JR, et al. ChAdOx1 nCoV-19 vaccine prevents SARS-CoV-2 pneumonia in rhesus macaques. 643

High neutralizing 645 antibody titer in intensive care unit patients with COVID-19

Clinical progression 648 and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a 649 prospective study

Improved Molecular 651 Diagnosis of COVID-19 by the Novel, Highly Sensitive and Specific COVID-19-RdRp/Hel 652 Real-Time Reverse Transcription-PCR Assay Validated In Vitro and with Clinical Specimens

SARS-CoV-2 shedding 655 and seroconversion among passengers quarantined after disembarking a cruise ship: a case 656 series

Natural transmission 658 of bat-like SARS-CoV-2PRRA variants in COVID-19 patients

Safety and 660 Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Efficacy and 662 Safety of the mRNA-1273 SARS-CoV-2 Vaccine

Safety 664 and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim 665 analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Neutralizing Antibody LY-CoV555 in Outpatients with Covid-19

Robust SARS-CoV-2 infection in 670 nasal turbinates after treatment with systemic neutralizing antibodies

Learning the language of viral evolution and 672 escape

CoV-2 Variants B.1.351 and B.1.1.7. Nature. 2021

Safety and immunogenicity of the ChAdOx1 nCoV-19 vaccine against SARS-CoV-2: a 677 preliminary report of a phase 1/2, single-blind, randomised controlled trial

Vaccine Breakthrough Infections Reported to CDC -United States

Anti-spike IgG causes severe acute 683 lung injury by skewing macrophage responses during acute SARS-CoV infection

PD1-based DNA vaccine 686 amplifies HIV-1 GAG-specific CD8+ T cells in mice

An A14U Substitution in the 688 3' Noncoding Region of the M Segment of Viral RNA Supports Replication of Influenza Virus 689 with an NS1 Deletion by Modulating Alternative Splicing of M Segment mRNAs

Generation of a Broadly Useful 692 Model for COVID-19 Pathogenesis, Vaccination, and Treatment

Tandem bispecific neutralizing 694 antibody eliminates HIV-1 infection in humanized mice

696 Evaluation of the mRNA-1273 Vaccine against SARS-CoV-2 in Nonhuman Primates

Antibody responses against SARS 699 coronavirus are correlated with disease outcome of infected individuals

Longitudinally profiling neutralizing antibody response to SARS coronavirus with pseudotypes

An infectivity-705 enhancing site on the SARS-CoV-2 spike protein targeted by antibodies

Afucosylated IgG characterizes enveloped viral responses and correlates with COVID-19 709 severity

Generation of DelNS1 711 Influenza Viruses: a Strategy for Optimizing Live Attenuated Influenza Vaccines

Adaptive immunity to SARS-CoV-2 and COVID-19

Memory T 715 cells established by seasonal human influenza A infection cross-react with avian influenza A 716 (H5N1) in healthy individuals

SARS-CoV-2 immunity: review and 718 applications to phase 3 vaccine candidates

COVID-19 720 vaccine BNT162b1 elicits human antibody and TH1 T cell responses

SARS-CoV-2 723 infection of human ACE2-transgenic mice causes severe lung inflammation and impaired 724 function

COVID-19 726 treatments and pathogenesis including anosmia in K18-hACE2 mice

Intranasal ChAd Vaccine Protects Upper and Lower Respiratory Tracts against SARS-CoV-2

Anti-viral protective capacity of tissue resident memory T cells

SARS-CoV-2 vaccines in development

T-cell homing specificity and plasticity: new concepts and 735 future challenges

BNT162b 737 vaccines protect rhesus macaques from SARS-CoV-2

Neutralization of 739 SARS-CoV-2 lineage B.1.1.7 pseudovirus by BNT162b2 vaccine-elicited human sera

Covid-19: Novavax vaccine efficacy is 86% against UK variant and 60% 742 against South African variant

SARS-CoV-2 spike 744 glycoprotein vaccine candidate NVX-CoV2373 immunogenicity in baboons and protection in 745 mice

Evaluation of the immunogenicity of prime-boost vaccination with the replication-deficient 748 viral vectored COVID-19 vaccine candidate ChAdOx1 nCoV-19

Imperfect vaccines and the evolution of 750 pathogen virulence

DNA prime/MVTT 752 boost regimen with HIV-1 mosaic Gag enhances the potency of antigen-specific immune 753 responses

Heterologous MVA-S prime Ad5-S boost 755 regimen induces high and persistent levels of neutralizing antibody response against SARS 756 coronavirus

Spatiotemporal interplay of 758 severe acute respiratory syndrome coronavirus and respiratory mucosal cells drives viral 759 dissemination in rhesus macaques

Inactivated Vaccine Candidate, BBIBP-CorV, with Potent Protection against SARS-CoV-2

Development of an inactivated 764 vaccine candidate for SARS-CoV-2

A single-dose mRNA vaccine 766 provides a long-term protection for hACE2 transgenic mice from SARS-CoV-2

A modified vaccinia Ankara vector-based vaccine protects macaques from SARS-CoV-2 770 infection, immune pathology, and dysfunction in the lungs

Ad26 vaccine protects against SARS-CoV-2 in rhesus macaques

Intranasal 776 administration of RSV antigen-expressing MCMV elicits robust tissue-resident effector and 777 effector memory CD8+T cells in the lung

Safety and 779 Immunogenicity of a DNA SARS-CoV-2 vaccine (ZyCoV-D): Results of an open-label, non-780 randomized phase I part of phase I/II clinical study by intradermal route in healthy subjects in 781 India

India's DNA Covid Vaccine Is a World First -More Are Coming

Safety and 785 immunogenicity of INO-4800 DNA vaccine against SARS-CoV-2: A preliminary report of an 786 open-label, Phase 1 clinical trial

Co-infection by severe acute 788 respiratory syndrome coronavirus 2 and influenza A(H1N1)pdm09 virus enhances the severity of

 RBD-specific IgG (square) indicate the standard error of the mean. Statistics were generated using one-way ANOVA 808 followed by Tukey's multiple comparisons test. *p<0.05; **p<0.01. 809